Experimental simulation and modeling of sock-to-skin

friction during running

Eglantine Baussan

To cite this version:

Eglantine Baussan. Experimental simulation and modeling of sock-to-skin friction during running.
Other. Université de Haute Alsace - Mulhouse, 2010. English. �NNT : 2010MULH3179�. �tel-00871091�

HAL Id: tel-00871091

[Link]
Submitted on 8 Oct 2013

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
ANNÉE 2010 N° 2010MULH33

UNIVERSITÉ DE HAUTE ALSACE

ÉCOLE NATIONALE SUPÉRIEURE D’INGÉNIEURS SUD ALSACE
LABORATOIRE DE PHYSIQUE ET DE MÉCANIQUE TEXTILES
EMPA LABORATORY FOR PROTECTION AND PHYSIOLOGY

THÈSE
Pour l’obtention du titre de

DOCTEUR DE L’UNIVERSITÉ DE HAUTE ALSACE

DISCIPLINE : MÉCANIQUE
Par

Églantine BAUSSAN

Étude tribologique et modélisation du contact peau - chaussette.

Application à la course à pied.

Soutenue publiquement le 6 décembre 2010 devant le jury suivant :

Professeur Guy MONTEIL ENSMM Besançon Rapporteur

D. R. CNRS Yves BERTHIER INSA de Lyon Rapporteur
Professeur Betty LEMAIRE-SEMAIL Polytech'Lille Examinatrice
Docteur René ROSSI EMPA St Gallen Co-encadrant
Docteur Siegfried DERLER EMPA St Gallen Co-encadrant
Professeur Marie-Ange BUENO Université de Haute Alsace Directrice

1
 REMERCIEMENTS

J’aimerais en premier lieu exprimer ma reconnaissance à Marie-Ange pour avoir initié

ce projet de recherche en codirection avec l’EMPA St Gallen et me l’avoir proposé. En plus
de m’avoir convaincue d’entreprendre ce travail, elle m’a accompagnée au quotidien par ses
conseils et son soutien et je l’en remercie.
Mes pensées vont également à Siegfried et René qui ont supervisé mon travail et
m’ont chaleureusement accueillie dans leur groupe de recherche lors de mes travaux à St
Gallen.

Je suis aussi très reconnaissante aux professeurs Guy Monteil, Yves Berthier et Betty
Lemaire-Semail d’avoir accepté de rapporter et d’examiner mon travail. J’ai apprécié leurs
conseils et leurs remarques constructives lors de ma soutenance.
Merci à Dominique Malfait, PDG de Labonal, pour sa disponibilité, pour nous avoir
fourni notre base de travail et nous avoir suivi tout au long de ce projet.
Je tiens de même à remercier Alexandre Watzky de l’Université Paris XII pour ses
conseils quant au modèle que nous avons développé.
Ce travail n’aurait pas été réalisé sans le financement de la région Alsace, de
l’Université de Haute Alsace et de l’EMPA Swiss Federal Laboratories for Materials Science
and Technology St Gallen, qu’ils soient assurés de ma reconnaissance.

Je tiens tout particulièrement à remercier Brigitte qui en plus de m’avoir formée et

accompagnée tout au long de mon monitorat et de mes années d’ATER, a toujours été
disponible pour moi et m’a aidé pour l’exploitation informatique de mes résultats et sur la
partie modélisation qu’elle continue d’ailleurs d’améliorer. Son soutien m’a été très précieux
lors des passages difficiles de ma thèse, je regretterai notre collaboration et nos sorties
sportives.
Un grand merci aussi à Michel pour son amitié et son aide dans la mise en place du
tribomètre qui a servi à mes recherches, à Romain pour avoir repris le flambeau et complété
mes travaux par son approche mécanique et à Philippe pour m’avoir assistée lors de la partie
expérimentale de mon travail.
Merci à Christian Lutz Gerhardt qui m’a familiarisée avec les équipements de l’EMPA
et que j’aurais aimé mieux connaître.

2
 Je tiens bien sûr à ne pas oublier mes collègues de bureau ; Céline, Coralie, Mohamed,
Romain, Cyril, Jean-Maurice et Mohamed Ali, grâce à qui l’ambiance de travail a toujours été
sympathique et qui ont participé à l’aboutissement de mon travail par leur présence, leurs
conseils et leur bonne humeur.

Je pense encore aux acteurs du service technique ; Francis, Thien, Sébastien, Philippe,
Christian, Jean-Christophe, qui font en sorte que notre quotidien soit le plus aisé possible.
J’ai bien évidemment une pensée pour toutes les personnes qui font et ont fait de cette
école un endroit où il fait bon venir, travailler, vivre, qu’ils soient thésards, chercheurs,
enseignants, IATOSS, élèves ou extérieurs. Je remercie tous ceux qui se sont intéressés de
près ou de loin à qui je suis et ce que je fais.

Enfin, j’aimerais terminer par une pensée pour mon mari, ma famille, celle de mon
côté et celle du côté d’Olivier, et mes amis qui m’ont soutenue, écoutée, secouée, encouragée,
conseillée, de manière indéfectible tout au long de ma thèse. Leur aide m’a été extrêmement
précieuse.

3
 CONTENTS

Remerciements _____________________________________________________________ 2

Textile glossary _____________________________________________________________ 5

Introduction _______________________________________________________________ 13

Chapter I: state of the art _____________________________________________________ 16

Chapter II: materials and methods _____________________________________________ 71

Chapter III: experimental results ______________________________________________ 100

Chapter IV: modelling of sock-to-skin friction ___________________________________ 161

Conclusion _______________________________________________________________ 187

4
Textile glossary

5
Acetate
Acetate is a manufactured fibre of cellulose ethanoate (acetate) wherein less than 92%, but at
least 74% of the hydroxyl groups of the original cellulose are acetylated.

Acrylic
Acrylic is a manufactured fibre composed of synthetic linear macromolecules having in the
chain at least 85% (by mass) of recurring acrylonitrile groups.

Back
The back is the reverse of a fabric as opposed to the face. Face and back are identified in
function of the way the fabric is used or in function of the position of the fabric on the loom.

Bast fibre (fibre libérienne)

A bast fibre is a fibre obtained from the stems of various plants.

Burlap
Burlap is a plain cloth made from single yarns of approximately the same linear density in
warp and weft, usually made from bast fibres, particularly jute.

Cast-off position
It is the position in the knitting cycle where the needle is retracted to its lowest point, allowing
the new loop to be drawn through the old loop and then casting the latter over the head of the
needle hook.

Compressibility
Compressibility is a term relating to the hand of fabrics and referring to the resistance to
squeezing. It ranges from soft (high compressibility) to hard (low compressibility).
Measurements of thickness at various pressures provide an objective mean of testing
compressibility. This notion is closed to the inverse of hardness for a standard material.

Compressional resilience
Compressional resilience is the recovery of thickness or energy through a compression and
recovery cycle.

6
Course
A course is a row of loops across the width of a knitted fabric, i.e. perpendicular to the
machine direction.

Crimp
The crimp of a yarn is the waviness or distortion of the yarn that is due to interlacing in the
fabric. In woven fabrics, the crimp is measured by the relation between the length of the
fabric sample and the corresponding length of yarn when it is removed there from and
straightened under suitable tension. Crimp may be expressed numerically as (a) percentage
crimp, which is 100 divided by the fabric length and multiplied by the difference between the
yarn length and the fabric length, and (b) crimp ratio, which is the ratio of yarn length to
fabric length.

Elastane
Elastane is a manufactured fibre that is composed of synthetic linear macromolecules having
in the chain at least 85% (by mass) of segmented polyurethane groups and that rapidly reverts
substantially to its unstretched length after extension to three times that length. The most
famous trademark of elastane is Lycra®.

Face of fabric
The face of a fabric is the side which is intended to be the use surface or which is to be visible
in an end product.

Fibre
(1) General. Fibre is the fundamental element used in the fabrication of textile yarns and
fabrics.
(2) Specific. Fibre is a unit of matter characterized by having a length at least 100 times its
diameter or width, and with the exception of noncrystalline glass fibre, having a definitely
preferred orientation of its crystal unit cells with respect to a specific axis.
(3) Textile. Fibres can be spun into a yarn and made into a fabric by interlacing in a variety of
methods, including weaving, knitting and twisting. The essential requisites for fibres to be
spun into yarn include a length of at least 5mm, pliability, cohesiveness, and sufficient
strength.

7
Filament
A filament is a fibre of indefinite length.

Fleece
A fleece is a fabric with a deep, napped, or pile face.

Float
In warp knitting: a float is a length of yarn not received by a needle and connecting two loops
of non-consecutive courses.
In weft knitting: a float is a length or lengths of yarn not received by a needle and connecting
two loops of the same course that are not in adjacent wales.

Gauge
The gauge is the number of needles per unit length along a needle bed of a knitting machine.
In current practice, the common unit length of one inch (25.4 mm) is used for all types of
warp and weft knitting machines.

Hosiery
Hosiery stands for the knitted coverings for the feet and legs.

Jute
Jute is the fibre obtained from the bast layer of the plants Corchorus capsularis and
Corchorus olitorius.

Knitting
Knitting is the process of forming a fabric by the intermeshing of loops of yarn. Warp knitting
is a method of making a fabric by normal knitting means in which the loops made from each
warp thread are formed substantially along the length of the fabric. It is characterized by the
fact that each warp thread is fed more or less in line with the direction in which the fabric is
produced. In weft knitting, the loops made by each weft thread are formed substantially across
the width of the fabric. It is characterized by the fact that each weft thread is fed more or less
at right angles to the direction in which the fabric is produced.

8
Knock-over
Knock-over is the action of casting off the old loop over the head of the needle.

Latch needle
A latch needle is a needle having a small terminal hook closed by a pivoting latch. The action
is automatic as the fabric loop overturns the latch and allows the loop to be knocked over. The
newly formed loop is drawn by the hook, and loop-forming and knock-over proceed
simultaneously.

Moisture regain
Moisture regain is the ratio of the mass of moisture in a material to the oven-dry mass. It is
usually expressed as a percentage for a standard atmosphere of 60% relative humidity and
22°C.

Nap
A nap is a fibrous surface produced on a fabric or felt by raising in which part of the fibre is
lifted from the basic structure.

Needle bed
A needle bed is a flat slotted plate in which knitting needles operate under the influence of
jacks or cams.

Nylon
Originally Nylon was the trademark of Dupont for polyamide 6.6. It is now a common noun.

Pile
Pile is a surface effect on a fabric formed by tufts or loops of yarn that stand up from the body
of the fabric.

Plain weft-knitted fabric

A plain weft-knitted fabric is a fabric in which all of the component knitted loops are of the
same sort and meshed in the same manner.

9
Polyamide
Polyamide is a manufactured fibre composed of synthetic linear macromolecules having in
the chain recurring amide groups, at least 85% of which are attached to aliphatic or
cyclo-aliphatic groups.

Polyester
Polyester fibre is a manufactured fibre composed of synthetic linear macromolecules having
in the chain at least 85% (by mass) of an ester of a diol and benzene-1 ,4-dicarboxylic acid
(terephthalic acid).

Polypropylene
Polypropylene is a manufactured fibre composed of synthetic linear macromolecules having
an aliphatic saturated hydrocarbon chain in which alternate carbon atoms carry a methyl
group, generally in an isotactic disposition and without further substitution.

Polytetrafluoroethylene
Polytetrafluoroethylene is a fibre made from a synthetic linear polymer in which the chief
repeating unit is –CF2- CF2-.

Polyurethane
Polyurethane is a manufactured fibre composed of synthetic linear macromolecules having in
the chain recurring aliphatic groups joined via urethane groups which together comprise at
least 85% (by mass) of the chain.

Raising
It consists in producing a layer of protruding fibres on the surface of fabrics by brushing,
teazling or rubbing. The fabric, in open width, is passed over rotating rollers covered with
teazles, fine wires, etc., whereby the surface fibres are pulled out or broken to give the
required effect.

Relative humidity
Relative humidity is the ratio of the actual pressure of the water vapour in the atmosphere to
the saturation pressure of water vapour at the same temperature and same total pressure. The
ratio is usually expressed as a percentage.

10
Sinker
A sinker is a blade that works in conjunction with knitting needles and assists with loop
formation and fabric holding.

Spandex
Spandex is an elastane fibre.

Spinning
Spinning is the process or the processes used in the production of yarns or filaments.

Stitch
A stitch is an intermeshed loop.

Viscose
Viscose is a manufactured fibre of cellulose obtained by dissolving sodium cellulose xanthate
in a dilute solution of sodium hydroxide.

Wale
A wale is a column of loops along the length of a fabric, i.e. in the production direction of the
knitting machine.

Worsted
Descriptive of yarns spun wholly from combed wool in which the fibres are reasonably
parallel, and fabrics or garments made from such yarns. In most countries, fabrics with a
small proportion of non-wool decorative threads can be described as worsted.

Yarn
A yarn is a product of substantial length and relatively small cross-section consisting of fibres
and/or filament(s) with or without twist.
Note: Assemblies of fibres or filaments are usually given other names during the stages that
lead to the production of yarn, e.g., tow, slubbing, sliver, or roving. Except in the case of
continuous-filaments or tape yarns, any tensile strength possessed by assemblies at these
stages is generally the minimum that can hold them together during processing.

11
Yarn count
The count is the mass per unit length. It can be expressed in various units.
Tex: it is a recognized international unit which corresponds to the mass in grams of one
kilometre of product.
Metric count (Nm): it is defined as the number of kilometres of yarn to obtain one kilogram.
Cotton count (Ne (c)): it is the number of hanks of 840 yards per pound.
1000
N m = N e (c) × 1.693 =
Tex

The definitions given here are drawn from the following books [1-3]:
1. Milton Harris, Handbook of textile fibers. first edition. 1954, Washington: Harris
research laboratories, inc. 356.

2. Textile Institute Textiles Terms and Definitions Committee, Textile Terms and
Definitions. Ninth Edition. 1991, Manchester: M C TUBBS BSc Tech AMCT CText
FTI FRSA P N DANIELS BA MIInfSc. 367.

3. George E. Linton, The modern textile and apparel dictionary. Fourth revised enlarged
edition 1973, Plainfield, New Jersey 07060, USA: textile book service, a division of
Bonn Industries, Incorporated 716.

12
Introduction

13
 Various skin troubles can be caused by sports activities such as allergic reactions, and
chafing when swimming, frictional alopecia or occlusive acne when bicycling, blisters, corns,
calluses and talon noir when running. We investigated the contact of the sock and the skin of
the foot during running. Friction blister is the most common cutaneous disorder for joggers.
Blisters are caused by clothes rubbing on the skin. They appear where skin is glabrous and
epidermis is thick and result from frictional forces that mechanically separate epidermis cells.
The formation of blisters depends on the magnitude of the frictional forces and the number of
times that an object cycles across the skin. During jogging, the foot is deformed in each stride.
A running shoe has to provide support and shock absorption. Due to its higher stiffness, the
shoe will not perfectly follow the skin deformations i.e. relative movements take place and
rubbing occurs. Various scientific publications about the structure of running shoes and the
fibre composition and properties of running socks are available. However, very few studies
investigated the influence of socks structures regarding friction induced on the skin. The
influence of the sock structure on friction must be a subject of interest for sock manufacturers
but such research seems to be kept confidential.

In this study, we investigated the friction which different running socks structures
exert on the skin during running by means of friction experiments using a mechanical skin
model. We also modelled the friction of textile surfaces covered with terries on the foot skin.
The objectives were to examine and simulate the effect of sock structure on friction for the
recreational runner. We aimed to determine whether one specific textile construction provides
better performances in terms of friction exerted on the skin and intended to develop a simple
and quick model that can predict the friction a given sock structure will induce on the skin of
a runner.

The thesis is divided into four chapters.

Chapter 1 gives an overview of the current state of the knowledge in skin injuries,
friction and compression properties of textile fabrics and skin and socks technology regarding
friction and compression in contact with the skin in the shoe.
Chapter 2 describes the testing conditions, the socks which were investigated and the
friction and compression analysers used.
Chapter 3 reports the experimental results of this research; the friction results are first
presented and then confronted to compression results in order to conclude on the
performances of the different textile structures.

14
 Chapter 4 describes and analyses previous friction and compression models before
introducing in details the analytical model of sock-to-skin friction that is proposed.
The conclusion finally highlights the most important findings of this thesis and
suggests outlooks for future works.

15
Chapter I: state of the art

16
 Résumé

Ce premier chapitre présente dans un premier temps les affections cutanées qui
peuvent être générées lors de pratiques sportives. Le cas de la course à pied et plus
spécifiquement les phlyctènes sont abordées en détail. Dans un deuxième temps, les
principales lois de la tribologie sont exposées ainsi que les propriétés de frottement et de
compression des surfaces textiles et de la peau. Pour terminer, l’influence de certaines
caractéristiques comme par exemple la composition ou bien l’épaisseur d’une chaussette de
course sur le frottement généré au niveau de la peau est discutée.

La peau est une barrière de protection pour les organes et les tissus qui forment le
corps humain. Lors de pratiques sportives, elle peut être exposée de manière prolongée ou
répétitive au frottement, à la pression, à l’humidité, au soleil, au froid ou à la chaleur.
Différentes affections cutanées peuvent s’ensuivre, nous n’en citerons ici que quelques unes.
Les cors et callosités sont des zones où la peau s’est épaissie de manière à mieux résister aux
frottements répétitifs. Selon le sport pratiqué et l’équipement utilisé, les callosités peuvent
apparaître à divers endroits. Les cors sont le plus souvent localisés au niveau des
proéminences osseuses des pieds. Les cors durs apparaissent là où les doigts de pieds sont en
contact avec la chaussette alors que les cors mous apparaissent entre les doigts de pieds. Une
malformation ou bien le fait de porter des chaussures et des chaussettes mal adaptées
augmente le risque d’apparition de cors. Parmi les problèmes cutanés fréquemment rencontrés
chez les marathoniens, les phlyctènes sont les plus fréquentes. Nous avons donc concentré nos
recherches sur cette affection particulière.
Il y a principalement quatre contraintes qui s’exercent sur le pied lorsque l’on court :
le choc dû au contact initial entre la chaussure et le sol lors de la foulée, la pression de
l’ensemble chaussure + chaussette, le frottement et le cisaillement dus au mouvement du pied
dans la chaussure. La probabilité d’apparition d’une ampoule dépend de l’intensité des forces
qui s’exercent sur la peau et du nombre de foulées. Plus les forces sont importantes plus
l’apparition de phlyctènes est rapide. Les étapes de formation d’une ampoule sont les
suivantes :
1 Les couches superficielles de l’épiderme subissent tout d’abord une exfoliation
et une zone rougie et douloureuse apparait.

17
 1 Ensuite, il y a séparation de deux couches de l’épiderme et l’espace laissé vide
entre les deux couches se remplit de liquide sous l’effet de la pression
hydrostatique.
Les phlyctènes apparaissent communément aux endroits où la peau est glabre et
l’épiderme épais tandis que les zones pileuses de la peau où l’épiderme est fin sont sujettes à
l’exfoliation mécanique et aux irritations. En ce qui concerne la course, les ampoules sont
plus nombreuses à l’avant du pied. La peau est pourtant plus fine à l’avant qu’à l’arrière du
pied mais l’avant du pied est soumis à plus de pression et de cisaillement que la région du
talon. Plusieurs facteurs jouent sur l’apparition de phlyctènes : selon le terrain choisi, les
forces de cisaillement qui s’exercent sur le pied sont plus ou moins importantes, la
transpiration, la chaleur ou des exercices inhabituels contribuent à accélérer l’apparition
d’ampoules, tandis que le fait de bien lacer ses chaussures permet de réduire leur nombre.

Nous étudions le contact tribologique entre la chaussette de sport et la peau de manière

à mesurer, simuler et diminuer le frottement s’exerçant sur la peau. Le contact tribologique est
déterminé par les deux corps en contact (rugosité, composition, propriétés mécaniques),
l’interface ou troisième corps, l’environnement (température, humidité relative) et le système
mécanique (force normale appliquée, vitesse, géométrie du contact). Les forces de frottement
dépendent, à l’échelle macroscopique, du système tribologique, à l’échelle mésoscopique, des
mécanismes de déformation des deux corps en contact au niveau de leurs aspérités de surface
et, à l’échelle moléculaire, des mécanismes d’adhésion entre les deux corps. Deux matériaux
en contact subissent des déformations i.e. des modifications dimensionnelles. On distingue
trois types de déformation. Les déformations élastiques apparaissent dès l’application de la
contrainte et disparaissent immédiatement et complètement lorsque celle-ci cesse. Les
déformations plastiques sont elles irréversibles tandis que les déformations viscoélastiques
disparaissent totalement mais pas immédiatement. Une partie de la déformation disparaît
immédiatement (effet élastique) et l’autre partie disparaît au fil du temps (effet visqueux).
Les principales lois du frottement sont celles d’Amontons, de Coulomb et de Bowden
et Tabor. Les lois d’Amontons indiquent que la force de frottement est indépendante de l’aire
de contact, que cette force est proportionnelle à la force normale appliquée i.e. F = µ × W
avec F force de frottement tangentielle en Newton, W force normale en Newton et µ
coefficient de frottement qui dépend des matériaux en contact, et que la force de frottement
est indépendante de la vitesse de déplacement. Coulomb a introduit la notion de frottement
statique et dynamique. La force de frottement statique est la force requise pour initier le
18
mouvement entre les deux corps alors que la force de frottement dynamique est la force
requise pour entretenir ce mouvement, généralement moins importante que la première. Selon
Bowden et Tabor, les lois du frottement dépendent des mécanismes d’adhésion et de
déformation qui sont tous deux liés à la notion d’aire de contact réel. L’aire de contact réel est
généralement bien inférieure à l’aire de contact apparente et dépend des rugosités des surfaces
en contact, des propriétés des matériaux, de la force normale appliquée et de l’échelle
considérée.
Certaines de ces lois ne s’appliquent pas dans le cas des matériaux textiles, notamment
la deuxième loi d’Amontons (la force de frottement est proportionnelle à la force normale
appliquée). Pour les fibres textiles, la relation F = KW n où K et n sont des constantes
déterminées empiriquement, 0 ≤ n ≤ 1 , est admise. C’est en fait l’aire de contact réelle qui
dépend de W. Si l’on considère une surface recouverte d’aspérités sphériques, dans le cas
d’une déformation purement plastique, l’aire de contact est directement proportionnelle à W
2
3
alors que dans le cas d’une déformation purement élastique, elle est proportionnelle à W .
Pour une surface réelle, dans le cas d’une déformation purement élastique, l’aire de contact
2
réelle sera donc proportionnelle à W n où ≤ n ≤ 1 . Pour des forces normales très élevées, la
3
déformation plastique du matériau sera importante et la force de frottement pourra alors être
modélisée par la relation F = αW + KW n où 1 est une constante empirique.
En ce qui concerne la peau, la force de frottement sur cette dernière est proportionnelle
2

à W 3 . Les propriétés tribologiques de la peau dépendent de sa souplesse et de son

hydratation. Le coefficient de frottement de la peau en contact avec une surface textile
augmente avec l’humidité. L’hydratation de la peau entraine en effet une adhésion plus forte
entre la peau et la surface textile.

Les chaussettes de sport sont principalement tricotées soit en jersey simple, soit en
jersey bouclette qui a la même structure que le jersey simple mais est tricoté avec deux fils ;
un fil de fond pour former la structure et un fil de bouclette qui forme des boucles sur la face
envers du tricot lui donnant ainsi de l’épaisseur. La composition d’une chaussette influence de
manière importante ses propriétés, comme par exemple sa compressibilité, son coefficient de
frottement et son taux de reprise. Pour une vitesse de frottement de 0,6 m.s-1, les coefficients
de frottement de différentes matières textiles peuvent être classés du plus grand au plus petit
dans cet ordre : coton, lin, coton peigné, laine, et soie. Cet ordre est indépendant des
19
conditions atmosphériques. Le taux de reprise d’un textile permet de quantifier son caractère
hydrophile. La laine est la fibre la plus hygroscopique ; l’eau représente 12 à 15% de son
poids dans des conditions normales d’humidité et peut atteindre jusqu’à 30 à 50% de celui-ci
dans un environnement humide. La laine absorbe l’humidité de manière lente et sèche aussi
lentement. La soie est presque aussi hydrophile que la laine mais sèche beaucoup plus
rapidement. Les chaussettes en coton sont également très hydrophiles, elles retiennent
l’humidité au contact de la peau ce qui provoque une sensation désagréable et de nombreuses
affections cutanées. En revanche, les fibres de polyester comme les fibres de Coolmax® sont
très hydrophobes et évacuent l’humidité. Pour donner un ordre de grandeur, les fibres de
coton retiennent trois fois plus d’humidité que les fibres d’acrylique et quatorze fois plus que
celles de Coolmax®. Plus la surface spécifique de la fibre est importante plus la fibre évacue
l’eau facilement. Les fibres artificielles et synthétiques présentent des sections droites de
formes très variées. À titre d’exemple, entre une section ronde et une section scalloped oval
comme celle du Coolmax®, la surface spécifique augmente de 20%. Si l’on compare les
fibres synthétiques aux fibres naturelles, les fibres synthétiques provoquent moins
d’affections cutanées car leurs coefficients de frottement et leurs taux de reprise sont plus bas.
Une bonne chaussette de sport doit absorber l’humidité mais aussi la drainer vers
l’extérieur et permettre également l’évacuation de la chaleur générée par le frottement. Les
chaussettes en coton ne sont pas recommandées car leur conductivité thermique est faible ;
0,07 W/m/K, et qu’elles sont hydrophiles. Le polyester avec un taux de reprise de 0,4% et une
conductivité thermique de 0,14 W/m/K est un assez bon compromis, le polyamide également
bien qu’avec un taux de reprise de 3% car sa conductivité thermique atteint 0,25 W/m/K. De
plus, il est conseillé de choisir un produit qui permette d’avoir un coefficient de frottement
faible avec la peau de manière à ce que le pied soit libre de bouger et ne subisse pas de
contraintes de cisaillement importantes et en même temps un coefficient de frottement
important avec la chaussure de manière à limiter l’amplitude du mouvement dans la
chaussure.
Malgré de très nombreuses recherches sur les phlyctènes, il est difficile de prévoir
l’apparition et la sévérité d’une ampoule et aucune solution simple et efficace n’existe pour
prévenir sa formation. Le meilleur moyen consiste à réduire au maximum la pression et le
frottement exercés sur la peau en choisissant des chaussures et des chaussettes adaptées. Le
choix des chaussures dépend des caractéristiques des pieds de la personne et de l’activité
pratiquée. Les chaussettes doivent être adaptées aux chaussures et à la manière de les porter.
Afin de réduire la pression exercée sur la peau, l’aire de contact entre la chaussette et la

20
chaussure doit être maximale ; la distance entre la chaussure et la peau doit donc être si
possible égale à l’épaisseur de la chaussette non comprimée. Si cette distance est plus petite,
cela augmente la pression de contact et si elle est plus grande, cela augmente la distance de
frottement du pied dans la chaussure.

21
 CONTENTS

Chapter I: state of the art ____________________________________________________ 16

I. SKIN INJURIES AND FOOTWEAR ____________________________________ 24
1. Human skin structure 24
2. Sport-induced skin injuries due to mechanical influences 25
A. Damages from friction and pressure 26
B. Skin erythema due, among others, to running practise 27
a. Jogger’s toe ___________________________________________________ 27
b. Chafing and abrasion ____________________________________________ 27
c. Friction blister _________________________________________________ 27
3. Blister formation 27
4. Factors influencing the formation of skin blisters 29
A. Footwear influence 29
B. Influence of the lacing technique 30
5. Prevention of blisters 31
II. FRICTION AND COMPRESSION OF FABRICS AND SKIN ________________ 33
1. Tribology overview 33
A. Tribological contact 33
B. Laws of friction 35
a. Laws of Amontons _____________________________________________ 35
b. Laws of Coulomb ______________________________________________ 35
c. Bowden and Tabor approach______________________________________ 36
2. Friction behaviour of textiles 38
A. Effect of load on the friction of fibres 38
B. Variation of the area of contact with load 39
C. Case of high normal stresses 40
D. Effect of fibre cross-section shape 40
E. Effect of yarn sett and hairiness 41
F. Other factors affecting friction 41
3. Frictional properties of human skin 41
A. Influence of the applied load on skin friction 41
B. Influence of skin humidity 42
C. Influence of anatomical site and age 43
4. Fabric to skin friction measurements 44
A. Fabric friction measurement methods 44

22
 B. Fabric to skin friction devices 45
C. Investigations on friction blisters 46
D. Skin models 46
5. Plantar pressure measurements 47
A. Measurement devices 47
B. Plantar pressure in normal foot 48
6. Fabric compression measurements 49
III. SOCKS TECHNOLOGY ____________________________________________ 52
1. Textile overview 52
A. Fibres 52
B. Yarns 52
C. Knitted Fabrics 53
2. Common knitted structures for sport socks 54
A. Simple jersey 55
B. Terry jersey 59
3. Influence of fibre composition 61
A. On coefficient of friction 61
B. Moisture regain 62
4. Influence of sock thickness 64
5. Are there ideal socks? 65
Literature cited __________________________________________________________ 67

23
 In this chapter, skin injuries that may occur during sports activities will be described.
Running induced skin disorders, especially friction blisters will be detailed. The influence of
external factors on blisters incidence will also be dealt with. We will see that blister formation
is mainly due to friction and pressure forces. General principles of tribology, friction and
compression properties of textiles and skin will then be exposed. Finally, the effect of
different properties of a sock, such as its fibre composition and its thickness, on skin friction
and compression will be tackled.

I. SKIN INJURIES AND FOOTWEAR

12 3456789AB789CD4EC4DF8
3456789AB789CD4EC4DF8
Human skin has a surface of about 2 m², a weight of 5 kg and a thickness from 1 to
4 mm for an adult. It has two functions: it protects our organs and tissues against mechanical,
biological, thermal and solar stresses and it enables perspiration to be evacuated while it
prevents water from penetrating in our body. It is composed of several layers, from the most
superficial to the deepest: the corneal layer, the epidermis, the dermis and the hypodermis [1].
The multilayered structure of skin is illustrated in figure 1.
The corneal layer is a superficial layer which covers the epidermis and acts as a barrier
against the environment aggressions. The epidermis is composed up to 90% by keratinocytes;
cells which insure cutaneous regeneration and healing. Its thickness can vary from 0.07 mm
for the eyelid to 1.4 mm for the sole of the foot [2, 3]. It depends in part on the thickness of
the stratum corneum which is largest at the hands and feet [4]. The stratum corneum is the
upper layer of the epidermis. It is around 0.015 mm thick and is renewed continuously i.e.
new cells are continuously integrated in the stratum corneum, but on the other hand,
superficial cells are permanently abraded and lost. The epidermis is separated from the dermis
by the basal membrane. This junctional zone has a complex structure. Under the epidermal
layer, the lamina lucida is located, followed by the lamina densa which is interlinked through
anchoring filaments and fibrils with the basal cells and the dermal fibrous tissues. This
structure is subjected to bullous skin diseases. The dermis is composed of collagen fibres§ and
elastin matrix. It is responsible for the skin tear resistance and elasticity. Its thickness ranges

§
This symbol indicates that the word definition is given in the textile glossary.

24
from 1 mm for the face to 4 mm for the back. The deformation of the dermis presents an
elastic and a viscous part. The hypodermis or subcutaneous tissue has a loosen structure
whereby the dermis can slide on the bone or the muscle and it absorbs a great part of the
shearing energy.

Figure 1: Human skin layers [5].

2 DC
DCB74EF89AB78B74DBF984F8C85FE67BE68B74F7EF9
B74EF89AB78B74DBF984F8C85FE67BE68B74F7EF98
B74DBF984F8C85FE67BE68B74F7EF98
The skin is the primary human protective barrier. During sports activities, the skin is
exposed to multiple prolonged stresses such as friction, pressure from clothing and
equipment, high level of skin humidity from perspiration, sunlight, cold or heat. Various skin
injuries can result, for example infections and dermatoses. Skin injuries may be caused either
by sun, cold, chemicals or contact and friction [6]. Many publications investigated the
affections of skin relative to sports activity [7-13].

25
 12 34546789ABC59ABDEFDC949B788B79
During use and because of mechanical loads during sport activity, sport clothes and
equipment can generate friction, pressure, heat and occlusion on the athlete skin thus
provoking superficial folliculitis called acne mechanica. Many pieces of athletic equipment
have been reported to cause this acneiform eruption such as sweatbands, face guards, helmets,
football paddings or backpacks [7, 9, 12, 13].

Human skin also develops protective response to chronic and repetitive friction.
Calluses and corns are thus thickened, hyperkeratotic plaques which occur at points of
continuous friction like palms and soles. Depending on the sport and the equipment, calluses
may develop in various locations. Corns have a deep central core and can be hard or soft [13].
They occur as often as not over foot bony prominences, especially over the metatarsal heads,
along the inner aspect of the large toe, or over structurally or functionally defective areas of
the foot. Poor fitting footwear and skeletal deformities increase the risk of developing corns.
Hard corns occur on the external surface of the toes where drying occurs while soft corns are
interdigital and result from sweating.

Black heel (also called talon noir) is an asymptomatic lesion which often occurs on the
post lateral aspect of the heel of teenagers and young adults. It is characterized by discrete
brown or blue-black macules caused by the rupture of small peripheral blood vessels [9, 10].
It is frequently observed in sports which require frequent sudden stops i.e. basketball or
handball. Repeated stop-and-start motions, changes in direction and constant pounding on
hard surfaces lead to injury of the heel against the back of the shoe. The macules of talon noir
occur on the posterior, medial, and lateral sides of the heel, just above the thick plantar skin
where blood vessels are less protected by fatty tissue.

Most of the cutaneous reactions related to sports activities can be developed during
running practice, especially foot lesions.

26
 2 D97BF75497945C69CF7B89FC9BD69B4EFD879

a. Jogger’s toe
Approximately 0.1 to 14% of runners reported injuries to the toenails on marathon
day [10]. It is a subungual hematoma which commonly appears on the hallux, on the second
toenail, or on the lateral aspects of the third, fourth and fifth toes. It presents a black
discoloration, oedema and erythema and is due to repetitive shocks of the longest toe against
the toe box* of the shoe. It often occurs with downhill running.

b. Chafing and abrasion

The friction of moist skin can result in chafing which is a superficial inflammatory
dermatitis. The keratin is separated from the granular sub layer of the epidermis by friction in
a warm and humid environment. In the marathon literature, chafing was reported by 0.4 to
16% of runners [8, 10].
Long distance runners commonly suffer from jogger’s nipples which is a painful and
crusted abrasion of the areolae and nipples. Jogger’s nipples are caused by repetitive friction
against coarse shirts and occur in 2 to 16.3% of runners on marathon day.

c. Friction blister
Friction blister is the most common injury in marathon participants. Up to 39% of
marathon runners who participated to the surveys had post marathon blisters [8]. For infantry
soldiers carrying heavy equipment over long distances, blisters can account for 48% of the
total injuries [14]. We therefore focused our research on this particular skin phenomenon
which occurs at the interface between foot and sock. As blister is the subject of our research,
it will be fully detailed in the next paragraphs.

2 B9CFD8D56CB78
B9CFD8D56CB78
Knapik et al. reported that most activity-related blisters may be due to frictional
shearing forces [15]. However, frictional shearing forces do not seem to be sufficient for a
blister to occur. For Reynolds et al., it is the combination of pressure, shear and moderate
level of moisture which generates friction blisters [16]. Comaish suggested that mechanical
fatigue might be the most important mechanism in blister formation [17]. Four main stresses

*
The toe box is the part of the shoe where the toes are.

27
actually act on the foot: step shock, pressure, friction and shear. Step shock is the impact force
which arises from the initial contact of the foot with the ground. For a running athlete, this
force can be up to three times the body weight [18]. Friction results from the rubbing of the
footwear (shoe and sock) on the skin of the foot. It is often localized and provokes heat and
skin exfoliation. While friction corresponds to the external aspect of skin rubbing, shear can
affect both foot contact surface and deeper tissues. The initial contact between the foot and
the ground is made with the heel. When the heel touches the ground, friction between the foot
and the footwear is generated whereby a horizontal shear force acting backwards against the
foot is produced. The shearing forces are thought to be more damaging to the skin than
vertical forces [18]. Step shock, pressure, friction and shear are usually occurring at the same
time which results in injuries of the foot structures and the skin [19]. The magnitude of forces
and the number of times an object cycles across the skin determine the probability of blister
development: the higher the forces, the fewer the cycles necessary to produce a blister.
According to Knapik [20], there is a minimal force below which no blister will form,
regardless of the number of shear cycles.

The forces detailed above lead to an exfoliation of the stratum corneum and the
formation of a reddened region in and around the area of rubbing which is referred as a
“hot-spot” [20]. Then a cleavage of the epidermal layers occurs, usually between the stratum
granulosum and the stratum spinosum layers [3, 8, 10, 11, 13, 14, 20-23]. The three most
superficial skin layers, i.e. the stratum corneum, the stratum lucidum and the stratum
granulosum are then left intact and form the blister roof [21]. The separation of skin cell
layers may also take place subepidermally [11]. Once the layers are separated, hydrostatic
pressure causes the resulting free space to fill with blood or tissue transudation [8, 14, 20].

Friction blisters commonly occur on glabrous skin where epidermis is thick whereas
friction on hairy skin with thin epidermis usually leads to mechanical skin exfoliation and
irritation as illustrated figure 2.

Herring and Richie reported that friction blisters are peculiar to the skin of the hand
palms and foot soles where the thickened stratum corneum, stratum lucidum and stratum
granulosum develop an interfacial movement with the stratum spinosum [22]. According to
Mailler , Adams and Bergeron, blistering events usually occur on the tips of the toes, the balls
of the feet and the posterior heel [8, 11]. Forefoot region is more affected than heel region.

28
60.2 % of the blisters observed by Herring and Richie [22] on the feet occurred in the forefoot
region, 33.3 % in the midfoot region and 6.5% in the rearfoot area. The skin is actually
thinner on the forefoot than on the heel area. Moreover, Dai et al. measured larger magnitudes
of the plantar pressure and shear on the forefoot region than on the heel region thus explaining
this difference [24]. The size of blisters is very dependent on the region of the skin, the
contact and the environment, Herring and Richie [22] observed blisters from 8.7 to 81.2 mm².
For Bergeron [11], blisters of less then 5 mm in diameter are small blisters.

x10
counterpart movement
x1
liquid

healthy skin blister (glabrous skin, mechanical irritation

thick epidermis) (hairy skin, thin epidermis)

Figure 2: Friction blister and mechanical skin irritation phenomena.

2 6ECD98B74F7EB78CF8D56CB7889AB78B9CFD98
6ECD98B74F7EB78CF8D56CB7889AB78B9CFD98
The type of ground surface may increase the shearing forces on the foot thus
promoting blisters [18]. Skin moisture, heat or unusual exercises are others factors which
contribute to blisters formation [8, 16]. The number and severity of previous foot skin blisters
are also crucial. Patterson et al. studied the blistering events among 100 cadets in a summer
camp [6]. They found that women and cadets with a history of blisters in the 2 years before
camp had higher relative risk of blister formation. They recommended as well the
preconditioning of the foot to its footwear in order to reduce severe blisters. They thus
suggested a significant influence of the footwear on the formation of skin blisters.

12 CCF74B9DA7E79
Running shoes can be classified according to their level of stability. A runner should
buy running shoes corresponding to his weight and way of running. Running shoes that bring

29
a high level of stability are indicated for severe overpronators, flat-footed and bigger runners
while neutral shoes correspond to lighter and underpronating runners. Finally, running shoes
that give a moderate level of stability are designed to fit the majority of runners [25]. Richie
includes the footwear, notably the type of insole and the socks, in the environmental
parameters that can increase foot blisters formation [18]. He reported studies have been
carried out which investigated the role of footwear on foot lesions. These studies showed that
appropriate footwear reduce the occurrence of foot affections by redistributing and decreasing
plantar pressure and shear. The importance of footwear in blistering events will be further
discussed later in this manuscript. Footwear includes the shoe and the sock; consequently one
must be careful in the choice of both (size and technical properties) but also in the way of
wearing them. Shoe lacing system is thus of great importance.

2 A7E79CA9F794ED69F7ED79
In 2001, Sandrey et al. [26] compared the ability of different soccer shoes to control
rear foot motion and showed that a shoe with a special “pronated lacing technique” presented
a better effectiveness than others. In 2002, Polster [27] recommended X-lacing as the best
lacing technique. These studies inspired Hagen and Hennig [28] who investigated the effects
of different shoe lacing systems on rear foot motion, plantar pressure distribution and shock
attenuation during running.
The latter study investigated nine different lacing conditions varying the number of
laced eyelets. An additional test was carried out fixing the shoe lacing conditions and
changing the shoe lacing tightness from weak to normal and strong. Various parameters were
recorded during the running sequences: foot pronation during contact, tibial acceleration of
the right leg and plantar pressure under seven locations, i.e. medial and lateral heel, lateral
midfoot, first, third and fifth metatarsal heads and hallux.
Strong tightness and maximum number of eyelets laced conditions gave identical
results. The shoe lacing technique highly impacted foot-shoe coupling. Stronger lacing
induced a tighter coupling of foot to shoe leading to reduced peak vertical loading rates, peak
heel pressures and pronation velocity. On the other hand, the important sliding of foot within
the shoe in the weak tightness and low laced shoe conditions reduced peak vertical impact
forces and peak pressures under the forefoot. Hagen and Hennig could conclude that the
important thing was to well lace running shoes which is, to my point of view, the most
important thing because the better lacing conditions may surely differ according to which

30
shoe and which sock are worn and also among individuals. The footwear is a closed
environment, every part of it, i.e. the foot, the sock and the shoe interferes with the others so
the whole system should always be considered. Appropriate shoes and lacing system are
therefore not sufficient; one should also wear appropriate socks. This will be the matter of this
chapter third part. Besides adapted footwear, there are a couple of methods which may
prevent blisters to occur.

2 !DF"F7CB788B9CFD98
!DF"F7CB788B9CFD98
Healthy and well hydrated skin can bear more stress before breaking down and is then
less likely to develop friction blisters [29]. There are several products which can help reduce
friction and shear on the skin and thus prevent blister formation such as lubricants or skin
surface drying agents [6, 8]. Rubbing moist skin produces higher frictional forces than
rubbing very dry or very wet skin which suggests that reducing sweating, e.g. using
antiperspirants, may reduce friction and consequently reduce blisters [15].
Reynolds et al. [16] investigated the influence of an antiperspirant with emollient*
additives on frequency and severity of frictional blisters, hot spots, and irritant dermatitis. The
antiperspirant may cause irritant dermatitis; the emollient was thus added to reduce the
incidence of this specific skin disorder. They investigated 23 healthy subjects free of lower
extremity skin lesions and with no history of hypersensitivity to topical skin products. Their
feet were treated with either the antiperspirant (20% aluminium zirconium tetrachlorohydrex
glycine concentration plus water) with emollient additives, emollient additives alone (placebo
control), or nothing for 4 consecutive days. They were asked to walk on a treadmill for 200
minutes. During the walk, the amount of foot sweat, the number of injuries and the size of
blisters were assessed. The subjects were also asked to rate the severity of the blisters and hot
spots on a pain scale of 0 to 10. No differences were seen among treatment conditions for
sweat accumulation, blister incidence, hot spot formation, or blister severity. However, mean
blister size and perceived severity increased with exercise duration. As antiperspirants are
proved to decrease sweat rate, the authors thought that the emollients may have affected the
antiperspirant’s chemical properties and may have also acted as moisturizing agents,
increasing friction and macerating the outer epidermis.
Knapik and his colleagues [15] later examined whether an antiperspirant can reduce
foot blisters during hiking. They carried out a double-blind study separating into two groups
*
emollient: an emollient is a substance which relaxes and softens the skin (émollient in French).

31
cadets attending the United States Military Academy. One group used an antiperspirant (20%
aluminium chloride hexahydrate in anhydrous ethyl alcohol), the second group used a placebo
preparation (anhydrous ethyl alcohol). Their feet were examined for blisters before and after a
21 km hike. It was observed that the antiperspirant may be efficient to reduce foot blisters
during prolonged cross-country hiking if applied at least three times on 3 separate days before
a hike. Compared to the placebo group, cadets using the antiperspirant at least 3 nights before
the hike had indeed a 56% lower incidence of blisters. The lower incidence of blisters is
correlated to the number of participants who reported that their feet did not sweat during the
experiment. The results are detailed in tables 1 and 2.

PREPARATION APPLICATION FOR 0 TO 2 APPLICATION FOR 3 TO 5

USED DAYS DAYS
Antiperspirant 39% 21%
Placebo 41% 48%

Table 1: Overall incidence of blisters [15].

PREPARATION APPLICATION FOR 0 TO 2 APPLICATION FOR 3 TO 5

USED DAYS DAYS
Antiperspirant 15% 45%
Placebo 9% 17%

Table 2: Percentage of participants whose feet did not sweat [15].

Cadets who used the antiperspirant nevertheless reported significantly more irritation
than those using the placebo. The overall incidence of self-reported irritation was 57% in the
antiperspirant group and 6% in the placebo group. Knapik et al. suggested that using
antiperspirants less concentrated in aluminium chlorohydrate, applying antiperspirants every
other or every third night or combining the active ingredients of antiperspirants with a
cortisone-based preparation may be methods of reducing skin irritation while preserving the
favourable antiperspirant property.

32
 II. FRICTION AND COMPRESSION OF FABRICS
AND SKIN

12 #DB$8"FD"BF%8
#DB$8"FD"BF%8
The tribology is the academic discipline which studies the interactions between
contacting surfaces [30]. It essentially deals with solid-solid contacts, solid-fluid contacts,
friction, wear and lubrication. All domains of activities are concerned by the tribology as
contacts contribute to the mechanisms of solid to solid bonds in live systems or in devices.

The first practical aspects of friction have their roots in prehistory [31]. More than
400,000 years ago, our hominid ancestors were making use of friction when they chipped
stone tools. By 200,000 B.C., Neanderthals had achieved a clear mastery of friction,
generating fire by the rubbing of wood on wood and by the striking of flint stones. The second
application showing understanding of friction phenomena, namely the use of lubricants to
minimize the work required to transport heavy objects, dates back more than 4000 years.
However, the tribology has been recognized for less than fifty years.

Friction forces are opposed to the movement of a solid in contact, in our case, with
another solid. The rolling friction is usually lower than the sliding friction; it is the wheel
principle. The sliding friction slows down or stops the relative movement of rotation or
translation of two bodies in contact.

12 BD CC6DE49ECF4EF9
The different actors of a tribological contact are:
1 The two contacting bodies which are characterised by their surface roughness, their
composition, their mechanical properties etc…
1 The interface or third body is constituted of contaminants, lubricants, etc…It can be
solid, fluid, introduced on purpose in the contact (lubricant for example) or generated
by the contact mechanism (such as abraded particles).
1 The environment (temperature, relative humidity, etc…).

33
 1 The mechanical system which includes the applied normal load, the velocity, the
contact geometry, etc…

The contact geometry is the general shape of the bodies’ surfaces around the contact
area. There are three main contact geometries:
1 Point-shaped contacts (typically ball on a flat surface).
1 Linear contact (cylinder on a flat surface).
1 Surfaced contact (parallelepiped on a flat surface).

The friction forces depend on:

1 The tribological system on a macroscopic scale.
1 The deformation mechanisms of the two contacting bodies at the surfaces asperities
(mesoscopic scale).
1 The adhesion mechanisms between the two bodies at the atomic or molecular scale.

Lubrication aims to reduce friction i.e. to increase sliding or reduce wear and heat
generation or both at the same time. It consists in generating a free-motion thin (from 0.1 and
100 µm thick) film between the two contacting surfaces. This thin film can be a gas, a liquid
or a solid film and should have the lowest possible shear resistance; it will separate the two
surfaces and make the movement easier. The choice of the lubricant depends on its role in the
system. Wear is characterised by material dislocation. It can be adhesive, abrasive, corrosive
or with fatigue and does not necessarily follow the friction evolution. Adhesion is
characterised by the creation of interfacial bonds between the contacting surfaces. Abrasion is
due to the action of hard particles which tear off material from the contacting surfaces. When
two materials are in contact, deformations i.e. modifications of shape and/or dimensions may
occur.

There are three different types of deformations:

1 Purely elastic deformations which occur immediately under stress and disappear
immediately and completely when the stress stops. These are reversible
deformations.
1 Plastic deformations which are irreversible.

34
 1 Viscoelastic deformations which disappear in part or totally when the stress stops
acting. Two stages are distinguished: a part of the deformation clears immediately
(elastic effect) and the other part wears off with time (viscous effect).

2 !489CA9ABDEFDC9

a. Laws of Amontons
The first recorded systematic study in the field of tribology is due to Leonardo da
Vinci (1452-1519) who, in the XVth century, measured the force of friction between objects
on both horizontal and inclined surfaces. Cords attached to the object to be moved were
allowed to pass over fixed rollers or pulleys to weights which gave a measure of the resisting
force. Leonardo da Vinci also studied the influence of the apparent contact area upon friction
resistance. He wrote: “the friction made by the same weight will be of equal resistance at the
beginning of its movement although the contact may be of different breadth and length” [32]
and “friction produces double the amount of effort if the weight is doubled” but the first laws
of friction were attributed to Amontons (cf. figure 3):
1 First law: the friction force is independent of the contact area.
1 Second law: the friction force is proportional to the applied normal load (weight).
The friction can therefore be expressed as F = µ × W where F is the tangential
friction force in Newton, W the normal load in Newton and µ the friction coefficient
(without dimension). µ depends on the materials in contact.
1 Third law: the friction force is independent of the sliding velocity.

4F
F F

Figure 3: Amontons’ laws of friction (1699).

b. Laws of Coulomb

35
 In 1782, Coulomb rounded out Amontons second law introducing the concept of static
and dynamic friction (cf. figure 4). The static friction force is the force required to initiate
motion while the dynamic or kinetic friction force is the force required to sustain motion.
Static friction is usually higher than dynamic friction.

Figure 4: Static and dynamic friction forces.

Coulomb observed that the kinetic friction is almost independent of the sliding
velocity and this is sometimes referred to as the third law of friction. He also considered the
possibility that friction arises from adhesion between the surfaces in contact but abandoned
this theory when he observed that the friction was independent of the area of the bodies. He
then suggested that friction arises from the asperities present on the surfaces, the frictional
energy expended in sliding representing the work necessary to lift one surface over the
asperities of the other [33].
In many cases, the laws of Amontons and Coulomb are not verified. Nevertheless, the
coefficients of friction of several materials can be compared provided that they were
measured in the same conditions.

c. Bowden and Tabor approach

The current theory of basic friction mechanisms was proposed in 1950 by Bowden and
Tabor. According to this theory, two mechanisms permit to physically explain friction laws;
adhesion and deformation. These two phenomena are related to the notion of real contact area.
When two surfaces are in contact, high local stresses and possible increase in surface

36
temperatures lead to bonds generation. Macroscopic bodies always have rough surfaces, at
least on a microscopic level. When two solid materials are placed in contact, some regions on
their surfaces will be so close together that the surface atoms of one material “touch” the
surface atoms of the other material, while in other regions, the surface atoms are separated by
large distances (around 1 µm). The regions of contact are referred to as junctions
(cf. figure 5). The real contact area is the sum of the junctions’ areas. The real contact area is
commonly a lot lower than the apparent contact area. Real contact area depends on:
1 Surfaces roughness: the rougher the surfaces, the lower the real contact area.
1 Material properties: the higher the hardness, the lower the real contact area.
1 Normal load: when the normal load increases, the contact areas are deformed and new
junctions are created.
1 The scale which is considered.

F
junction

Figure 5: Contact between two solids; junctions.

The junctions are formed by adhesive contact. The adhesion between the contacting
bodies results from atomic, chemical or physical attractions at the junctions. The adhesion
force is the normal force Fa which is necessary to separate the two bodies by shearing under a
Fa
given normal load W. The adhesion coefficient is thus µ a = . The adhesion force depends
W
on the two materials in contact and on the real contact area. It may be increased with a
prolonged contact as the number and the surface area of the junctions may get higher with the
time of contact thus increasing the real contact area (by creep). Moreover, the adhesion force
is all the more important that the two contacting bodies are close together and usually gets
higher linearly with normal force.

37
 During friction, a displacement of material to the side and to the front of the contact or
even a removal of material can occur. The deformation force depends on the volume of
material that has been displaced. If the contacting bodies have different hardness, the hardest
body creates an imprint on the softest. Two actions are then combined; junctions are sheared
and materials are removed from the softest body.
To resume, if the interaction between adhesion and deformation is supposed
negligible, the friction force between two bodies in contact can be written as:
F = Fa + Fd with Fa: adhesion force and Fd: deformation force. This approach is adapted to
viscous materials, but in this case, the deformation force will also depend on the speed of
deformation.

2 DBECB78F6"B4D88CF&CBF98
DBECB78F6"B4D88CF&CBF98
Studies conducted during the latter half of the twentieth century, on natural and
synthetic fibres, have shown that the coefficient of friction is not a material property, but is a
function of the normal force and the geometric area of contact [34]. An increase in normal
force causes a decrease in the coefficient of friction. The size of the fibre, the smoothness of
the surface and the mode of contact (point, line or area) also affect the value of the coefficient
of friction. In the classical materials like metals, frictional force F is directly proportional to
the normal force W. However, a non-linear relation occurs with most polymeric materials.
Aliouche [35] investigated the friction of woven textiles and observed that the
coefficient of friction µ began to decrease sharply when the normal load increased from 0 to a
positive value then the coefficient of friction still got lower but slowly and became roughly
constant for high normal loads. He explained the stabilization of the value of the coefficient of
friction by the integration of the hairiness in the volume of the woven fabric. At high normal
loads, the distribution of loads on the contact surface is thought to be more homogeneous and
the number of points of contact to be maximal.

12 "AA7EF9CA9C49C9F79ABDEFDC9CA9AD B789

It seems satisfactory to use a simple relationship like: F = KW n where K and n are

empirical constants whose values are found by fitting the data on the model and conducting a
least square analysis [34, 36], n is less than unity. All workers in the field of fibre friction who
adopt this relationship have assumed firstly that friction arises from adhesion at the points of
38
real contact and secondly that the junctions at the interface have a constant shear strength s.
Therefore, they have all explained the variation of F with W in terms of the variation of the
true area of contact A with W [33]. If s is a constant, it follows from the previous equation
that A = k1W n where n has the same value as in the previous equation. Lincoln (1952)
examined the friction between a sphere and a flat nylon surface at low loads and found that F
2
3
varied as W . He then measured the geometrical area of contact between a nylon sphere and
a glass flat by optical interference methods and found that the area of the circle of contact
2
3
varied as W . Howell and Mazur (1953) investigated the friction between crossed fibres at
right angles and found a relationship of the form F = kW n where n was 0.8 for drawn nylon,
0.9 for undrawn nylon and 0.96 for cellulose acetate. In 1951, Finch studied the imprint
formed by curved cylinders of nylon on a glass flat and found the area of the imprint to be
proportional to W 0.89 . This appears to be in excellent agreement with the frictional
dependence on W. However, Howell and Mazur used an optical interference method similar
to that used by Lincoln and obtained a different result. They wound the fibre round a
cylindrical rod and pressed the curved fibre against an optical flat. The optical area was
observed and often had very irregular boundaries, varied from fibre to fibre, and was
time-dependent. Under standard loading conditions of 15 seconds, they found that for
practically all the fibres examined the optical area varied as W 0.73 . In addition, they observed
that a fibre bent over a rod of larger diameter gave a larger optical area for a given load. This
is in agreement with the type of behaviour to be expected for elastic deformation. Howell
pointed out that these optical measurements of the area of contact have marked limitations.
First, the exact location of the boundary of the ellipse or circle of contact is uncertain.
Secondly, the optical area of contact is the same as the true area of contact only if the surfaces
are smooth. With surfaces covered with asperities the real or true area of contact will always
be smaller than the optical area. The recognition of this difficulty has led to the development
of a number of models showing how the true area of contact may be expected to vary with
load.

2 #4BD4FDC9CA9F794B749CA9ECF4EF9DF9C49
Archard (1951) considered a surface covered with asperities of spherical shape [33].
The two extreme types of deformation are purely plastic and purely elastic. For purely plastic

39
deformation the area of contact is directly proportional to the load. For purely elastic
2
3
deformation the area of contact of each asperity is proportional to W . Two possible ways in
which elastic deformation can take place were foreseen:
1 Contact occurs over a fixed number of asperities and the effect of increasing the load
is simply to increase the elastic deformation of each asperity. In this case the area of
2

real contact is proportional to W 3 (smooth contact).

1 Contact occurs over a large number of asperities, the average area of each deformed
asperity being constant. Increasing the load in this case increases the number of
regions of contact and the area of real contact is directly proportional to W (rough
contact).

With real surfaces an intermediate behaviour may be expected. Consequently, for pure
elastic deformation, the true area of contact will be proportional to W n where n lies between
2/3 and 1. If this is the only parameter affecting the friction, the friction will also follow a
similar relationship.

$2 $4879CA9D69CB5498FB788789
At high normal stresses, a textile will have significant plastic deformation. A plausible
model for characterizing the behaviour in this region may be F = αW + KW n where 1 is an
additional empirical constant [33]. Using the classical definition of the coefficient of friction,
µ = α + KW n−1 . This model, initially proposed by Gralen for applying to fibre friction in
general, was used by Briscoe for characterizing friction in ballistic textiles. Under the ballistic
levels of stresses, the material is expected to lose its elastic recovery properties and be mostly
deformed plastically. The second term of the model will then be small and friction will largely
be given by the first term.

32 "AA7EF9CA9AD B79EBC88%87EFDC98479
El-Mogahzy and Gupta [34] investigated the effects of structural and morphological
factors on the friction of fibrous materials. They measured the friction of monofilaments of
polypropylene of three different cross-sectional shapes namely circular, triangular and

40
trilobal. In both the line and the point contact tests, the circular fibre gave a significantly
higher value of µ than did the triangular or the trilobal materials, the average difference being
about 31%. The values of K and n for the circular fibres were also higher by 28% and 3.6%
respectively than the corresponding average values for the non-circular materials. It was
visualized that the trilobal and triangular fibres were transversely stiffer and thus less
deformable than the circular.

"2 "AA7EF9CA94B987FF9494DBD7889
Ajayi [37] investigated the influence of yarn sett and surface hairiness on the friction
of woven fabrics. He found that the frictional resistance to motion of these fabrics increases
with the yarn sett and for the case of pile-type fabrics with the height of fibre tufts.

2 &F7B9A4EFCB894AA7EFD69ABDEFDC9
Some other factors that affect friction are the mode of contact (line, point or area)
during tests, the morphology of the surfaces (degree of roughness or smoothness), the testing
environment (temperature, and relative humidity), and the time of contact (time before sliding
and speed of sliding) [34]. The mode of contact used will affect the number of points of
contact (m) and through it the values of K, A, F and µ. One would generally expect that the
larger the area the larger would be the value of µ. Differences in surface morphology can
affect friction as a rougher surface, with fewer contacts per unit area, would mean a smaller
value of m and therefore should result in lower friction. Finally, since fibres are viscoelastic
materials, their properties can change with the time of loading. It is expected that the contact
area will increase as the sliding speed is decreased thus leading to an increase in the
coefficient of friction [36].

2 DBECB768
DBECB768DFDCBF9
68DFDCBF98
DFDCBF988456789AB78
8456789AB78

12 A7E79CA9F794D79C49C98D9ABDEFDC9
Amontons’ law stipulating the non-variance of the coefficient of friction with load
does not apply to the skin [38, 39]. Sivamani et al. [38] found that µ is proportional to W −0.32
where W is the applied normal load. Koudine et al. [39] found that the coefficient of friction µ

41
was proportional to W −0.28 . These results are very similar to the theoretical relationship
deduced by Wolfram [40] from the Hertz theory:
2/3
6K3
µ ∝ S4 1 W −1 / 3 where E is the Young’s modulus of skin, K is a term including
5E2
average dimension of adhesive contacts, their number, and frequency per unit area, and S is
the shear strength of the adhesive contacts.
2
3
These three studies showed that the friction force on the skin is proportional to W .

2 A7E79CA98D95DDF9
Friction properties of human skin depend on the skin texture, suppleness and
hydration. The stratum corneum can absorb up to 8 times its weight in water [2]. The stratum
corneum hydration degree is a very important parameter. When the stratum corneum is dry, it
is brittle and less resistant but when it is highly hydrated, it becomes delicate. The
brittle-to-ductile transition for skin has been reported to occur at 70% RH [41]. The stiffness
and coefficient of friction of the stratum corneum increase with humidity whereas an extreme
dryness and wetness will tend to decrease skin friction [6, 21]. Numerous studies investigated
the evolution of the coefficient of friction of skin in contact with textiles in function of skin
humidity [32, 38, 41-46]. They all reported an increase of the coefficient of friction with
higher skin humidity.
In a recent study, Gerhardt et al. [42] systematically varied the hydration state of skin
of the volar forearm in 22 subjects and found a highly positive linear correlation between skin
moisture and friction coefficients against textiles. They found that the evolution of skin
friction with moisture is gender influenced; the friction of female skin showed significantly
higher moisture sensitivity. Coefficients of friction increased typically by 43% in women and
by 26% in men when skin hydration varied from very dry to normally moist skin.

A study on the effects of wearing diapers on skin was carried out by Zimmerer et al.
[43]. They observed that skin wetness was proportional to diaper wetness and that with
increased skin wetness, there were increased coefficients of friction and increased abrasion
damage.

42
 Gwosdow et al. [44] pulled fabrics across skin in different humid conditions from
neutral to hot-dry and hot-humid. For them, greater humidity involves greater degree of
hydration of the stratum corneum which may lead to a better adhesion between skin and
fabrics. Greater adhesion could result in a higher fabric to skin contact area thus explaining
the increase of friction coefficient and friction force. Herring [29] proposed the same
explanation stating that skin moisture increases friction, torque and shearing forces by
increasing the adhesion of the skin to the sock.

Kenins [41] showed that fibre type and moisture influenced fabric-to-skin friction.
Moreover, it could be concluded that moisture on the skin was more important than fibre type
or fabric construction in determining the intensity and nature of fabric-to-skin friction and
also that glabrous skin friction is less affected by moisture than hairy skin.

Nacht et al. [45] investigated the changes in skin friction coefficient induced by skin
hydration and emollient application. They observed that when a drop of water is put on the
skin, the coefficient of friction immediately increases sharply and then decreases rapidly to its
initial value. They explained this phenomenon by the absorption of water in the stratum
corneum thus increasing the contact area between the probe and the skin and leading to an
increase in friction coefficient. As water is lost through evaporation, the stratum corneum
returns to its original state and the friction coefficient decreases to its baseline value.

$2 A7E79CA944FC5DE498DF79494679
In 1990, Elsner, Wilhelm and Maibach [47] studied the frictional properties of human
skin in two localised areas: the forearm and the vulva. 44 healthy female volunteers
participated to this study whereby the dynamic friction coefficient between skin and a teflon
probe was measured. Elsner et al were looking for correlations between the dynamic friction
coefficient and particular parameters which could influence skin properties: age, body weight,
height, transepidermal water loss, and skin capacitance. They found a higher coefficient of
friction for vulvar skin (0.66 ± 0.03) than for forearm skin (0.48 ± 0.01) which can be
explained by the important hydration of vulvar skin compared to forearm skin. Age seems to
influence transepidermal water loss and friction coefficient in forearm but not in vulvar skin.

43
 Cua, Wilhelm and Maibach [48] also measured the dynamic friction coefficient of skin
and the skin lipid content in 11 different anatomical regions: the forehead, the upper arm, the
volar and dorsal forearm, the post auricular, the palm, the abdomen, the upper and lower back,
the tight and the ankle of 29 volunteers of different age and gender. They found no particular
difference in dynamic friction coefficient when varying age and sex group, but dynamic
friction coefficient does vary significantly depending on the anatomical site.

Zhang and Mak [49] measured the coefficient of friction of skin in contact with
several materials over six anatomical regions, namely the dorsum of the hand, the palm of the
hand, the anterior side of the forearm, the posterior side or the forearm, the anterior leg and
the posterior leg. The highest coefficient of friction was measured at the palms of the hand.

Gerhardt et al. [50] studied skin-textile friction and skin elasticity in young and aged
people. They confirmed that skin elasticity is lower in elderly persons. For seniors, more
pronounced skin tissue displacements also occurred and greater shear forces took place during
frictional contact. These observations showed that deformation plays an important role in the
friction of aged skin.

2 6DBE8C89AB78DBECB785F694DF5F7C9
6DBE8C89AB78DBECB785F694DF5F7C98
DBECB785F694DF5F7C98

12 4 BDE9ABDEFDC95748B757F957FC89
The measurement methods of friction behaviour of fabrics are based on a signal
treatment in temporal domain by a statistical analysis of the friction force, on the difference
between static and dynamic friction or either on the analysis of the coefficient of friction
variation in function of the applied normal load [36, 51]. The classical measures can be
unidirectional or multidirectional. The unidirectional method the most commonly used
consists in fixing the fabric on a horizontal support. A probe connected to a force sensor is
then rubbed on the fabric. A statistical treatment of the temporal signal is realised whereby a
mean friction coefficient can be calculated. This is the principle of the friction and roughness
module of the Kawabata’s Evaluation System (KES; a range of measurements devices

44
developed by Kawabata and the Hand Evaluation and Standardization Committee and
commercialised by Kato Tech Co., Ltd) [52].

A multidirectional tribometer was developed by Bueno [51] because the KES friction
and roughness module was found not sufficiently sensible to characterise the effects of
finishing treatments. This device combines two characteristics; a study of the fabric in all the
directions of its surface and the use of the periodical character of a fabric surface which is
harnessed using a signal treatment in the frequency domain. Its use is thus limited [53]:
1 The textile fabric tested should have a periodical surface with a small period.
1 This tribometer is based on the “frequential signature” of a textile fabric; it consists in
computing the Fourier transform of the temporal signal of the fabric during rubbing.
The autospectrum obtained shows one or several frequency peak(s). The peak(s)
height characterizes the fabric state with great sensitivity. This tribometer thus
quantifies a modification of the surface but can hardly compare different fabrics.

Another multidirectional device was therefore developed, using a new probe. This
apparatus consists in a very thin steel plate (50 to 75 mm long, 10 to 30 mm wide and 30 to
50 µm thick) which is excited in vibration by the rotating textile fabric. The steel plate is
prestressed before a measure i.e. bent on the textile surface. The vibration of the plate is
measured with strain measurement gauges. The steel plate has natural frequencies which are
independent of the rubbed textile surface. The energy of each mode of vibration is analysed in
order to characterize the textile surface whatever structure it has [54, 55].

2 4 BDE9FC98D9ABDEFDC97'DE789
Fabric to skin friction was usually assessed by using panels of volunteers. It generally
consisted in slowly pulling fabric samples across the surface of a subject’s skin. The pressure
between fabric and skin was often applied by suspending a weight to the free end of the fabric
and the frictional force required to pull each fabric across the skin was recorded by a force
transducer [6, 44].
Friction between hand and object or between skin and textiles can also be analysed by
rubbing the skin against objects and surface samples attached to a multi-axial force plate or a
force transducer. Normal and friction forces are thus recorded whereby friction coefficients
can be calculated [56].

45
 An alternative approach consists in using a tribometer whereby probes made of
different materials can be slid over the skin of subjects [46].

$2 '78FD64FDC89C9ABDEFDC9 D8F7B89
Studies on friction blister formation and prevention are usually performed by
recording the prevalence and size of blisters among a group of subjects which either have
routinely heavy load of activities, such as athletes or military personals or are asked to
perform specific trials [8, 15, 16].
A device was designed in the late 1960s and early 1970s in order to produce friction
blisters on human skin [6]. The rubbing head of this apparatus could be moved over the
surface of any chosen skin site at a selected stroking rate under a chosen load. Various
materials, including textiles, could be attached to the rubbing head. Frictional coefficient and
temperature could be recorded. It was observed that the rubbing head geometry, weight and
attached material all affect the friction coefficient measurements. The effect of skin moisture
was also studied and it was shown that a dry skin reduced the friction, while a moist skin
increased the friction [6, 14, 32].

32 D95C789
Fabric-to-skin friction was often assessed by measurements on voluntary subjects’
skin but the appearance of skin models is providing new possibilities to scientists who study
human skin. To model fabric-to-skin friction and compare this model to experiments, the best
is to use a skin model rubbing on the fabric as it will present a very low variability in
frictional properties. Skin models have indeed the advantage to be reproducible compared to
human skin and enable scientists to use classical friction measurement devices to assess fabric
to skin friction.
Derler et al. [57] investigated the friction of untreated human finger skin against a
reference textile with 12 subjects using a force plate. Fabric samples were attached to a
dynamometer and rubbed by the participants with the index finger. The sliding movements of
the finger ranged from 5 to 10 cm and were parallel to the longitudinal axis of the force plate.
The stroke frequencies were 1 ± 0.3 Hz and the contact areas between finger and fabric were
between 1.5 and 3 cm². For roughness assessments, the subjects used normal loads of

46
1.5 ± 0.7 N which resulted in typical contact pressures between 3 and 10 kPa. The results for
the human skin were compared with friction measurements of different silicone and
polyurethane materials as mechanical skin equivalents. A polyurethane coated polyamide
fleece with a surface structure similar to that of skin; Lorica®, showed the best
correspondence with human skin under dry conditions. The artificial leather Lorica® Soft is
also used for testing the slip resistance of floor coverings in barefoot areas.

2 !67C6D8
!67C6D8DF994DF85F694DF5F7C98
DF994DF85F694DF5F7C98
Plantar pressure is the distribution of force over the foot sole. Early investigations of
foot pressure patterns were obtained from the imprint in soft materials [58, 59]. These
methods gave qualitative results as they only recorded the shape of the foot and the deepest
impressions in the surface. The first quantitative measurements of foot pressure were carried
out using an air filled chamber by Carlet and Marey respectively in 1872 and 1873. Arrays of
deformable materials which left an ink print or were recorded by optical methods were later
developed.
Nowadays, a wide variety of plantar pressure measurement systems are available
which rely on electromechanical sensors. Pressure sensors are force transducers whereby the
force exerted on a known surface can be measured. Pressure is then calculated dividing force
by area. Electromechanical transducers usually convert a mechanical stress into an electrical
signal which can be recorded and stored for later analysis.
The characteristics of the pressure measurement apparatus should be adapted to the
research subject. For example, a higher spatial resolution is needed to measure children’s feet
pressure than adults’ feet pressure because a child’s foot is smaller than an adult’s foot. To
measure pressure on small anatomical structures, one should not use a too large sensor as it
will underestimate the real pressure due to the lower pressures around the structure. There are
different sensor principles such as resistive or capacitive methods and different measurement
devices i.e. insoles or single transducer systems. All the sensors detailed below [58] measure
the normal component of the ground reaction but neglect shear forces.

12 9(748B757F97'DE789
In-shoe systems are used to record the pressure at the foot-shoe interface. Multiple
steps can be recorded and easily averaged and analysed provided that the orientation of the

47
foot remains the same with respect to the sensor. Some systems include a portable data logger
which allows field measurements.
Matrix devices are a succession of rows or columns of sensors whereby the whole
plantar surface of the foot can be covered by an active sensor area.
Sole systems are composed of sensitive elements placed at selected anatomical sites on
the sole. They offer faster sampling rates as less data has to be processed.

2 9)4F4B9B788B79D9CB549ACCF9
Many parameters may affect foot pressure. These include shoe construction, foot
shape, age, gender, body mass and gait velocity. Body mass does not seem to influence
plantar pressure as much as one would expect. A low correlation was found between these
two parameters and only under the lateral midfoot and the third metatarsal head [58]. This
correlation was more pronounced in female subjects than in men. The whole plantar pressure
increases linearly with higher gait velocity but the same behaviour is not observed in all the
anatomical areas. The pressure at the heel and under the medial forefoot follows the whole
plantar pressure evolution while under the midfoot and the lateral forefoot the pressure
decreases with increasing speed [58].

Many studies investigated plantar pressure distributions, but either these investigations
were carried out on abnormal feet such as diabetic feet or the subjects were walking or
jogging barefoot [23, 60-65]. One can assume that abnormal feet may feel higher pressure
than normal feet and that in-shoe pressure may be lower than barefoot pressure but it depends
on the area of the foot and the footwear. It is thus difficult to estimate the pressure exerted on
normal feet during jogging with adapted footwear. However, the analysis of the running gait
and the structure of the foot indicate which foot areas suffer from higher pressures. Gait
analysis of running is indeed well-known: first the heel contacts the ground, then the others
structures of the foot are used from the heel to the toes and finally propulsion is given with the
forefoot. The heel and the forefoot are thus subjected to higher pressure than other elements
of the foot. Moreover, the structure of the foot shows bone prominences which also
experience high pressures e.g. the metatarsal and hallux [59, 66].

48
 '2 6DBE8
6DBE8E5DF99B785F694DF5F7C98
5DF99B785F694DF5F7C98
A fabric is not an elastic solid; it presents a hysteresis during a
compression - decompression test. To test the dynamic compression of a fabric, one can either
use the compression module of KES-F or an extensometer.

Peirce [67] studied compression test curves which provide the thickness of a fabric in
function of the applied pressure. He interpreted them as exponentials of which shape depends
on the fabric. Three different phases can be distinguished on these curves (cf. figure 6). The
first phase corresponds to the fabric surface hairiness squashing. The hairier the fabric, the
slower the reduction of thickness with increasing pressure. The second stage refers to the
compression of the fabric structure; the way yarns are organized may be slightly changed.
Once the air trapped in the fabric has been released, the fibres are finally deformed. The curve
asymptote corresponds to the material maximum deformation.

Hairs
Thickness

Structure

Material

Pressure

Figure 6: Pierce curve.

Kawabata compression module KES-FB3 (cf. figure 7) is composed of a circular

presser foot of 2 cm². When the driving motor switch is turned on, the plunger starts to
descend and compress the sample. When the compressional force reaches the preset
upper-limit force, the motor automatically turns to a recovery process. The
compression - decompression cycle is completed at a constant rate of compressional

49
deformation (0.02 mm.s-1 in standard mode, 0.007 mm.s-1, and 0.2 mm.s-1 if changing gear
system). The displacement of the plunger is detected by a potentiometer. The compressional
force transducer is a steel ring with a linear differential transformer. There are two relevant
pressures regarding the analysis of the compression results; P0 which equals 50 Pa and the
maximum pressure Pm which equals 5 kPa in standard mode.

The following parameters are obtained (cf. figure 8):

1 T0: thickness under P0, in mm
1 Tm: thickness under Pm, in mm
T0 − Tm
1 EMC: compressibility EMC (%) = 100 ×
T0
x = Tm

1 WC: compression work WC = 7 P × dx

x = T0
where P is the increasing pressure in

[Link]/cm²
WC
1 LC: compression linearity LC = without unit
1
× (T0 − Tm ) × Pm
2
It is the compression work of the fabric related to the compression work of an elastic solid
with same initial and final thicknesses i.e. divided by the area of the triangle formed by
the dotted lines on figure 8.
WC '
1 RC: resilience or restored energy rate RC (%) = 100 × where WC’ is the recovery
WC
x = Tm

work WC ' = 7 P'×dx where P’ is the decreasing pressure

x = T0

50
Figure 7: Schematic illustration of Kawabata’s compression tester.

Figure 8: An example of pressure thickness curve.

51
 III. SOCKS TECHNOLOGY

12 #F&CBF8"FD"BF%(8

12 D B789
Fibres and filaments§ are the basic elements of textile. Their length is a lot higher than
their transversal dimensions, they are therefore very flexible. Fibres lengths can be up to a
few centimetres while filaments have lengths of hundreds of meters, even several kilometres.
Filaments are obtained spinning§ either artificial materials such as viscose§ or acetate§, or
synthetic materials i.e. polyamide§, polyester§, acrylic§, polyurethane§. Silk is a natural
filament which length is comprised between 400 m and 1 km. Textile fibres can have a
natural origin such as wool, silk, cotton, linen or come from cut or cracked filaments.
The length and the fineness of fibres are two major parameters. The fineness of a fibre
can be characterised by its diameter if it has a round cross section or by its count§ i.e. its mass
per unit length. The count is usually given in Tex (1 Tex = 10-3g.m-1). Fibres and filaments
correspond to the microscopic scale of textiles (from 1 µm to 0.1 mm).

2 *4B89
Spinning is the process whereby yarns§ can be obtained from fibres or filaments.
Fibres must be cleaned, individualised and parallelised before being spun. The combing
process is a stage whereby the shortest fibres can be eliminated thus giving a better regularity
and resistance to the yarn. Yarn cohesion is usually obtained by twisting it but some yarns are
obtained by friction spinning.
Yarn twist is characterised by its direction (S if fibres are oriented like follows: \, and
Z if they are oriented in the opposite direction: /, cf. figure 9) and its value in turns per meter.
Yarn fineness is given in Tex. Yarns correspond to the mesoscopic scale of textiles (from 0.1
to 1 mm).

*
The information given in this section is drawn from my courses on textile metrology (taught by Jean-François
Le Magnen), spinning (given by Marc Renner and Artan Sinoïméri), and knitting (taught by Marie-Ange Bueno)
during my textile engineering education at the École Nationale Supérieure des Industries Textiles de Mulhouse.

52
 Figure 9: Z and S yarn twist direction.

$2 +DFF794 BDE89
A fabric is a textile surface obtained from yarns, fibres or filaments. The most
common textile fabrics are knitted fabrics, woven fabrics and nonwovens.
Knitting§ is the process whereby knitted fabrics are obtained. It consists in interlacing
one or several yarns to form stitches§. The kind of knit is the loops interlacing basic pattern.
In addition to the kind of knit, a knitted fabric is characterised by the yarns used to knit it, and
its stitch length. There are two different knitting methods: weft and warp knitting
(cf. figures 10 and 11).
In a weft knitted fabric, a single yarn forms successive courses of stitches; it is the
principle of hand knitting. In a warp knitted fabric, several yarns are knitted at the same time
and the stitches formed with the same yarn are in different courses. The thickness of a knitted
fabric depends, among others, on the yarn twist. The higher the twist, the thicker the knitted
fabric [53].

53
 x

Figure 10: Example of weft knitted fabric; simple jersey. A course§ of stitches is coloured in
grey. This knitted structure is realised with a single yarn.

Figure 11: example of warp knitted fabric; 1 and 1 lapping structure.

Fabrics correspond to the macroscopic scale of textiles (from 1 mm to 1 m). They are
flexible, deformable and viscoelastic materials. Fabrics behaviour depends on the three scales
that have been described. Fibres, yarns and fabric structure contribute to fabric mechanical
properties.

2 )5578A7BCCF8
)5578A7BCCF89CD4EC4DF9
A7BCCF89CD4EC4DF98
9CD4EC4DF98D89DC89EA98
D89DC89EA988
9EA988
Knitted fabrics are composed of curved yarns to form interlaced loops [68]. With
regard to the method of using the yarn, there are two knitting principles, warp knitting and

54
weft knitting. Socks are weft knitted fabrics. In weft knitted fabrics, a single yarn is used to
knit the fabric. This yarn travels horizontally to make a succession of loops. To form a loop
in a weft knitted fabric, each needle is inserted through each one of the original loops and thus
draws a yarn through the original loop to form a loop on the next course. The newly formed
loop takes the place of the original loop on the needle and remains there until the needle pulls
the next yarn through it [69].
There are two widely used weft knitted structures for socks products; simple and terry
jersey. These two kinds of knit are detailed and illustrated below.

12 D579,7B879
In simple jersey, all the stitches are identical; they are straight loops. The two faces are
very different. The face§ shows tightened wales§ of V which fit together. On the back§, one
can see successive and parallel waves (cf. figure 12). It is the weft knitted fabric the most
widely used in clothing (tee-shirt, pull-over, socks, stockings, etc…) and also in technical
textiles applications (coating support, composites reinforcements …).
This kind of knit is produced with a single needle bed§ knitting machine. Figure 13
shows the parts of the knitting machine which are involved in the loop formation. The needle
(1) is mobile in the vertical groove of the cylinder (2) via the action of the cams (3). It is kept
in place by the spring which is all around the cylinder (4). The sinkers§ (5) move horizontally
and slide in the grooves of the sinker ring (6) and in the grooves of the upper part of the
cylinder (9). They slide in a cam track (7) whereby they can be moved forward and back. The
different parts of a sinker and of a latch needle§ are respectively detailed in figures 14 and 15.
The yarn guides (8) are fixed on the cam track. The needles and the sinker ring rotate whereas
the cams and the yarn guides are immobile. The jersey loop formation is illustrated in
figure 16.

55
 Figure 12: Face (left) and back (right) of a simple jersey fabric.

Figure 13: Parts of the knitting machine involved in the jersey loop formation. The knitted
fabric is inside the machine (on the right hand side). The needle hooks are turned towards the
outside of the knitting machine.

56
Figure 14: The different parts of a sinker; 1) the butt, 2) the nose, 3) the throat, 4) the body.
The body behaves as a casting-off device and the nose prevents the knitted fabric from moving
up.

Figure 15: The different parts of a latch needle [70].

57
 The needle has just moved up from its casting-off
position§. Sinkers are moving towards the knitting
machine centre. The throat and the nose of the sinkers
will therefore keep the fabric in place while the needles
will move up.

The needles are moving up, the needle latch is opened

by the previous loop. The fabric is maintained by the
sinkers in their final forward position via the connecting
floats.

The needle is in its highest position, the previous loop

falls on the needle blade. The yarn guide prevents the
latch from closing.

The needle is moving down and takes the yarn in its

hook. The sinkers are moving back.

The needle keeps moving down pulling the yarn. The

needle latch is closed by the previous loop. The sinkers
keep moving back.

The needle reaches its lowest position pulling the yarn

through the previous loop. The sinkers reached their
final back position.

Figure 16: Main phases of the jersey loop formation.

58
 2 7BB9,7B879
Terry jersey is usually covered with loops on one side. Two yarns are knitted together,
the ground yarn and the pile§ yarn (cf. figure 17). The ground yarn is held while the pile yarn
is released to form loops [71]. The pile yarn must therefore be much more crimped§ than the
ground yarn in order to form terries on the back side of the fabric. The face has the same
aspect as simple jersey face while terries are visible on the back [68]. The formation of the
terry jersey loop is illustrated in figure 18.

Figure 17: Terry jersey knitted structure.

59
 The stitch of the previous course has just been knocked
over§ and the needle is moving up. The sinker is in its
final forward position; the stitches are maintained by the
throat of the sinker and the terry by the nose of the

1 sinker.

The needle reaches its highest position. The yarn guide

prevents the latch of the needle from closing. The sinker
is moving back and the terry is released.

The sinker is in its final back position and the yarn guide
is near to the needle.

1
pile yarn

The sinker moved forward for the terry to be formed

when the pile yarn is taken by the needle. The ground
ground yarn
yarn is at the back of the sinker throat.

1
The needle is in its casting-off position. The ground and
the pile yarns are drawn through the previous stitch. The
height of the terry depends on the height h of the sinker
nose. The higher the sinker nose, the higher the terry.
2.5, 3.1 and 3.5 mm are classical heights for a sinker
1
nose.

Figure 18: Main phases of the terry jersey loop formation.

60
 2 *74F7EF88BDF8E59BCB78
*74F7EF88BDF8E59BCB78
Sock fibre composition may influence many sock properties such as tensile strength,
compressibility§, heat transfer, friction coefficient and moisture regain§. Only the last two will
be treated below as coefficient of friction and moisture regain are essential parameters with
regard to fabric to skin friction and can thus have an effect on the size and the frequency of
blisters.
For example, Herring and Richie [22] carried out a double-blind study about the effect
of sock fibre composition on the frequency and the size of blistering events in long-distance
runners. The socks tested were identical in every aspect except fibre composition; one was
100% acrylic-made and the other was composed of 100% natural cotton fibres. They observed
less blisters and their size was smaller with the acrylic-made socks.
Mailler and Adams [8] suggested to wear moisture wicking synthetic socks in order to
prevent the formation of blisters.
Knapik and his colleagues also investigated the effect of sock composition on the
incidence and severity of foot blisters. Three boot-sock systems were tested namely the
standard military boot sock composed of a wool-cotton-nylon§-spandex§ combination, this
standard military boot sock with an additional thin inner polyester sock and a very thick outer
wool-polypropylene§ sock with the same liner as before. In terms of reduction of the
incidence of blisters and severe blisters, the socks can be ranged in this order: thick outer sock
with thin liner, then the standard military sock with the liner and finally the standard military
sock alone [6]. In this study, not only the fibre composition but also the thickness of the socks
was varied, the results may then be a combination of both effects. The influence of sock
thickness on sock to skin interaction will be discussed in the next paragraph.

12 &9EC7AADED7F9CA9ABDEFDC99
Gwosdow et al. [44] tested six different unwashed fabrics on eight healthy male
volunteers in an environmental chamber. Worsted§ wool, combed cotton, cotton, silk, linen
and burlap§ were pulled across the inner surface of the volunteers’ forearm at 0.6 m.s-1. The
fabrics can be ranged from the highest friction force to the lowest: cotton, burlap, linen,
combed cotton, wool and silk. This order was independent of the ambient conditions.
Zhang and Mak [49] investigated in vivo frictional properties of human skin in contact
with the following materials: aluminium, nylon, silicone, and cotton sock. Nylon showed an

61
average coefficient of friction of 0.37 with skin, while the mean coefficients of friction
between skin and cotton sock and between skin and silicone were 0.51 and 0.61 respectively.

2 (CD8FB79B764D9
Fabrics absorb water in stages. Water vapour first penetrates the fibre mass. The
amount of absorption depends on the chemical composition of the fibres; it is low for most
synthetics and high for most natural fibres. With increasing humidity, many layers of water
molecules may be bound and water may condense on the fibre surface. Then the fibres may
swell as plasticization of the monocrystalline structures occurs and liquid water will penetrate
the fibre mass. The amount of water taken up during theses processes is called regain [41].
Moisture regain is a parameter whereby the hydrophilicity of a textile material can be
quantified. Wool is the most hygroscopic textile fibre. It carries moisture up to 12 or 15% of
its weight under ordinary conditions and can take up 30 to 50% of its weight in a moist
environment. It absorbs moisture and dries slowly. Silk is nearly as hygroscopic as wool but it
loses moisture much more rapidly [72]. Cotton socks are also highly hydrophilic, they hold
moisture against the foot which leads to less cushioning and more moisture related cutaneous
problems like blisters and athlete’s foot. On the opposite, polyester fibres such as Coolmax®
are hydrophobic; they draw moisture away from the foot [73]. Cotton fibre retains three times
the moisture of acrylic and fourteen times the moisture of Coolmax®.
Moisture regain depends on relative humidity and temperature. It increases with
relative humidity but the order of fibres from the highest to the lowest regain does not change
with humidity [71]. Swelling properties are linked to moisture regain, the more hydrophilic a
fibre, the more it swells in water. Swelling is usually expressed as a percentage, for example
cotton fibre surface increases by 44 to 49% when plunged in water (cf. table 3). When a fibre
swells, it reduces the free space in the shoe and thus decreases the sweat transport from the
inside to the outside of the sock.
Moisture regain is not the only characteristic that determines the degree of
hydrophobicity of a fibre; overall wicking performance should also be considered. Fibres
should also have a great specific surface in order to facilitate water transport from the inside
to the outside of the sock. Spinning is a process whereby artificial and synthetic fibres can be
obtained. Textile material goes through the holes of a die and is then stretched. Artificial and
synthetic fibres cross-sections can thus be varied changing the shape of the die holes
(cf. figure 19). For example, Coolmax® section is scalloped oval increasing its specific

62
surface by 20% compared to a round cross-sectional geometry of fibre. Vapour transport is
thus easier for a Coolmax® than for a cotton sock.

COTTON WOOL SILK POLYESTER ACRYLIC NYLON

(PA 6-6)
Moisture regain at
21°C and 65% 8.5 17 11 0.4 1.6 3
relative humidity in
percentage
Swelling in water in 44-49 32-38 30-41 None 0-5 0-2
percentage

Table 3: Moisture regain and swelling of textile fibres [71]. The abbreviation PA 6-6 stands
for polyamide 6-6.

a) b) c)

d) e)

Figure 19: Different cross-sections of Polyester fibres photographed with a scanning electron
microscope: a) hexachanel, b) trilobal, c) scalloped oval, d) round and e) cruciform [74].
When comparing synthetic fibre compositions to natural fibre contents, synthetic
fibres are more efficient to prevent friction induced skin injuries because of lower friction
coefficients and moisture regain.
63
 2 *74F7EF889EA8CBEA7F998
*74F7EF889EA8CBEA7F998
Socks manufacturers propose various friction-reducing socks. In most of the case,
thick paddings are placed at selected areas such as the toes, the forefoot and the heel. Double
layer socks are also widely used. With these sock systems, designers try to make the friction
occur between the two layers of socks instead of on the skin [29, 75]. Using thick paddings or
double layer socks, the aim is to increase the thickness of the sock in order to decrease the
pressure exerted on the foot skin.

Mailler and Adams [8] suggested wearing two pairs of socks from different materials.
Richie [18] also advocates the use of double layer sock material of which he detailed the
composition: a synthetic inner layer coupled with a natural fibre outer layer. As high plantar
pressure is an acknowledged risk factor in the development of plantar ulcers in the diabetic
neuropathic foot and of skin irritations and blisters in healthy foot, Garrow et al. [19] studied
the ability of a multilayered sock (preventive foot care PFC) to reduce plantar foot pressure in
a sample of high-risk patients with diabetes. PFC socks have a double-layer construction
consisting of a padded outer layer to cushion the feet combined to a low-friction inner layer to
reduce friction at the sock/foot interface. They compared PFC socks to ordinary supermarket
socks recording in-shoe plantar pressure measurements with F-Scan, an in-shoe system
marketed by Tekscan, Boston, MA. They measured the differences in maximum foot contact
area and plantar pressure for the whole foot, forefoot and peak plantar pressure areas. They
found an increase of 8% in maximum foot contact area when subjects wore the PFC socks.
Total foot pressure was also decreased by 9%. Similar results were observed at the forefoot
with a 14.2% increase in contact area and a 10.2% reduction in peak forefoot pressure.

Wearing thicker socks seems therefore to increase the contact area between foot and
sock thus decreasing foot pressure. However, it is important to remember that hosiery§
offering thicker construction will reduce the volume within the shoe. Compared to the
supermarket socks which were around 0.7 mm thick, the maximum thickness of the PFC
hosiery was 2.7 mm around the toes, metatarsal phalangeal area and the heel. Although 2 mm
may not represent a huge increase in bulk inside the shoe, additional care in the fitting of
shoes is nonetheless required because such hosiery may induce localised increase in foot
pressure.

64
 2 +DF8CFDF8
+DF8CFDF8BF689EA9,88
BF689EA9,88
Perspiration can exceed 0.5 litre per foot during sport activity [18], moisture
absorption is thus needed. However, in addition to the perspiration of the foot, moisture
produced elsewhere on the body may accumulate in the sock due to gravity and this amount
far exceeds the absorption capacity of any sock product. Moisture absorption capacity is not
sufficient; a good athletic sock must also draw water away from the foot. Ideally, one should
wear hydrophobic socks and shoes with a hydrophilic inner liner. The socks should not be
too hydrophobic because they may not allow water transfer. It is also important that the sock
allows the evacuation of the heat produced by friction. Cotton fibres are not recommended
since they have a low thermal conductivity of 0.07 W/m/K and are hydrophilic. Polyester is
hydrophobic (moisture regain of 0.4%) and has an average thermal conductivity of
0.14 W/m/K, it can thus be a good compromise. Polyamide is also a good solution even if its
moisture regain equals 3% because its thermal conductivity reaches 0.25 W/m/K [76].

Dai et al. [24] investigated the effect of wearing socks with different frictional
properties on plantar shear. They simulated the foot-sock-insole contact using a 3-D finite
element model. With this model, they studied the dynamic plantar pressure and shear stress
during the stance phases of gait. They simulated three cases: a barefoot with a high frictional
coefficient against the insole (0.54), a sock with a high frictional coefficient against the skin
(0.54) and a low frictional coefficient against the insole (0.04) and another sock with opposite
frictional properties. They found that wearing socks of low friction against the insole allow
more relative sliding between the foot and footwear and thus reduce the shear force compared
to the barefoot. If the foot cannot move in the shoe, high pressure and shear can result in
blisters or other lesions on the forefoot. To avoid cutaneous lesions, the foot must be free to
slide in the shoe. Therefore, low friction is required between foot and the insole and there
must be a certain space in the shoe for the foot to slide. However, rubbing in the shoe includes
friction between the foot and the inner surface of sock and friction between the outer surface
of sock and the shoe, and if the friction is too low on both interfaces, it may lead to an
excessive movement of the foot in the shoe and to discomfort and insecurity. Ideally, a sock
should have low friction with the skin to allow movement of the foot and high friction with
the insole.

65
 Sport-specific socks appeared in the 1980s [18]. Soon, a small range of specific socks
were available for key sports such as tennis and running. Nowadays, there are special socks
for every conceivable sport activity; it has become a marketing issue. Actually, there is very
little difference in hosiery products design within any particular company but it is extremely
profitable. Socks should rather be designed specific to the shoe and the foot as it seems to be
the case for professional athletes.

Despite extensive research on friction blisters, scientists can hardly predict the
apparition and severity of a blister and there is still no simple and sure solution to prevent its
formation. It may be due to the dramatic variation of skin conditions (surface roughness,
hydration, adhesion between skin layers, etc.) among individuals as well as among different
anatomical sites of the same person. The best way to avoid blisters is to reduce the
mechanical factors that provoke their formation i.e. pressure and friction. This can be
achieved choosing well adapted footwear. The choice of the shoes depends on the
characteristics of the feet and the sport practise. The socks should be fitted to the shoes and
the way the runner tightens his shoes. In order to reduce the pressure on the skin, the contact
surface between the sock and the shoe should be as large as possible; the distance between the
shoe and the skin should ideally be equal to the thickness of the sock when it is not
compressed. A lower distance will lead to an increase of the contact pressure while a higher
distance will increase the sliding distance. The design of a sock may reduce the friction and
pressure forces acting on the foot skin. The aspects of sock fibre composition and thickness
have been discussed. For Richie [18], the ability of a sock to dissipate damaging forces on the
foot surface depends more on its structure than on its fibre composition. Our research is
precisely focused on studying the effect of the sock structure on the friction between the
fabric and the skin. The friction force between terry jersey socks and the skin is also
modelled.

66
 Literature cited

1. Régnier, M. and M.-J. Staquet, La peau reconstruite. Pour La Science, 1999(266): p.

154-159.
2. Breugnot, C., Contribution à la caractérisation mécanique du toucher des surfaces
textiles à partir de critères neurosensoriels, in Sciences Pour l'Ingénieur. 2005,
Université de Haute Alsace: Mulhouse. p. 209.
3. Sanders, J.E., B.S. Goldstein, and D.F. Leotta, Skin response to mechanical stress:
Adaptation rather than break down-A review of the literature. Journal of
Rehabilitation Research and Development, 1995. 32(3): p. 214-226.
4. Elsner, P., What textile engineers should know about the human skin. Curr Probl
Dermatol, 2003. 31: p. 24-34.
5. [cited 2010 2010/10/12]; Available from: [Link]
6. Zhong, W., et al., Textiles and Human Skin, Microclimate, Cutaneous Reactions: An
Overview. Cutaneous and Ocular Toxicology, 2006. 25(1): p. 23 - 39.
7. Reichel, M. and D.A. Laub, From Acne to Black Heel: Common Skin Injuries in
Sports. the physician and sportsmedicine, 1992. 20(2): p. 111-118.
8. Mailler, E.A. and B.B. Adams, The wear and tear of 26.2: dermatological injuries
reported on marathon day. British Journal of Sports Medicine, 2004. 38: p. 498-501.
9. Kirkland, R. and B.B. Adams, Dermatological problems in the football player.
International Journal of Dermatology, 2006. 45: p. 927-932.
10. Mailler-Savage, E.A. and B.B. Adams, Skin manifestations of running. Journal of the
American Academy of Dermatology, 2006. 55(2): p. 290-301.
11. Bergeron, B.P., A Guide to Blister Management. The physician and sportsmedicine,
1995. 23(2): p. 37-46.
12. Basler, R.S.W., C.M. Hunzeker, and M.A. Garcia, Athletic Skin Injuries. Combating
Pressure and Friction. The physician and sports medicine, 2004. 32(5): p. 33-40.
13. Pharis, D.B., C. Teller, and J.J.E. Wolf, Cutaneous manifestations of sports
participation. Journal of the American Academy of Dermatology, 1997. 36(3): p. 448-
459.
14. Xing, M., et al., Skin friction blistering: computer model. Skin Research and
Technology, 2007. 13(3): p. 310-316.
15. Knapik, J.J., K. Reynolds, and J. Barson, Influence of an antiperspirant on foot blister
incidence during cross-country hiking. Journal of the American Academy of
Dermatology, 1998. 39(2): p. 202-206.
16. Reynolds, K., et al., Effects of an antiperspirant with emollients on foot-sweat
accumulation and blister formation while walking in the heat. Journal of the American
Academy of Dermatology, 1995. 33(4): p. 626-630.
17. Comaish, J.S., Epidermal fatigue as a cause of friction blisters The Lancet, 1973.
301(7794): p. 81-83.
18. Richie, D.H. Socks: Hosiery - Essential equipment for the Athlete. [cited
2009/03/11]; Available from: [Link]
19. Adam P., G., V.S. Carine H.M., and B. Andrew J.M., Efficacy of Multilayered Hosiery
in Reducing In-Shoe Plantar Foot Pressure in High-Risk Patients With Diabetes.
Diabetes Care, 2005. 28(8): p. 2001-2006.
67
20. Knapik, J.J., et al., Friction Blisters - Pathophysiology, Prevention and Treatment.
Sports Medicine, 1995. 20(3): p. 136-147.
21. Polliack, A.A. and S. Scheinberg, A New Technology for Reducing Shear and Friction
Forces on the Skin: Implications for Blister Care in the Wilderness Setting.
Wilderness & Environmental Medicine, 2006. 17(2): p. 109-119.
22. Herring, K.M. and D.H. Richie, Friction blisters and sock fiber composition A double-
blind study. Journal of the American Podiatric Medical Association, 1990. 80(2): p.
63-71.
23. Yavuz, M. and B.L. Davis, Plantar shear stress distribution in athletic individuals
with frictional foot blisters. Journal of American Podiatric Medical Association, 2010.
100(2): p. 116-120.
24. Dai, X.-Q., et al., Effect of sock on biomechanical responses of foot during walking.
Clinical Biomechanics, 2006. 21(3): p. 314-321.
25. [Link] Running Shoes by Level of Stability. 2010/04/14.
26. Sandrey, M.A., C.J. Zebas, and J.D. Bast, Rear-foot motion in soccer players with
excessive pronation under 4 experimental conditions. Journal of Sport Rehabilitation,
2001. 10(2): p. 143-154.
27. Polster, B., Mathematics: What is the best way to lace your shoes? Nature, 2002.
420(6915): p. 476-476.
28. Hagen, M. and E.M. Hennig, The influence of different shoe lacing conditions on
plantar pressure distribution, shock attenuation and rearfoot motion in running.
Clinical Biomechanics, 2008. 23(5): p. 673-674.
29. Herring, K.M., How Socks Make The Feet. [Link]
[Link], 2009/04/08.
30. Gras, R., Tribologie Principes et solutions industrielles. Mécanique et matériaux, ed.
l.u. nouvelle. 2008, Paris: DUNOD. 322.
31. Persson, B.N.J., Sliding Friction Physical Principles and Applications, ed.
P.D.D.h.c.K.v. Klitzing and P.D.R. Wiesendanger. 1998: Springer-Verlag Berlin
Heidelberg. 462.
32. Naylor, P.F.D., The Skin Surface and Friction. British Journal of Dermatology, 1955.
67(7): p. 239-248.
33. Howell, H.G., K.W. Mieszkis, and D. Tabor, Friction in textiles. 1959, London:
Butterworths Scientific Publications. 263.
34. Gupta, B.S., Friction in textile materials. 2008, Cambridge, England: Woodhead
Publishing Limited. 462.
35. Aliouche, D., Contribution à l'étude des mécanismes de frottement des étoffes, in
Sciences de l'Ingénieur. 1991, Université de Haute Alsace: Mulhouse. p. 152.
36. Virto, L. and A. Naik, Frictional Behavior of Textile Fabrics. Part I: Sliding
Phenomena of Fabrics on Metallic and Polymeric Solid Surfaces. Textile Research
Journal, 1997. 67(11): p. 793-802.
37. Ajayi, J.O., EFFECTS OF FABRIC STRUCTURE ON FRICTIONAL PROPERTIES.
Textile Research Journal, 1992. 62(2): p. 87.
38. Sivamani, R.K., et al., Friction coefficient of skin in real-time. Skin Research and
Technology, 2003. 9(3): p. 235-239.
39. Koudine, A.A., et al., Frictional properties of skin: proposal of a new approach.
International Journal of Cosmetic Science, 2000. 22: p. 11-20.
40. Wolfram, L.J., Friction of skin. Journal of the Society of Cosmetic Chemists, 1983.
34(8): p. 465-476.
41. Kenins, P., Influence of Fiber Type and Moisture on Measured Fabric-to-Skin
Friction. Textile Research Journal, 1994. 64(12): p. 722-728.

68
42. Gerhardt, L.-C., et al., Influence of epidermal hydration on the friction of human skin
against textiles. Journal of the Royal Society Interface, 2008. 5: p. 1317–1328.
43. Zimmerer, R.E., K.D. Lawson, and C.J. Calvert, The Effects of Wearing Diapers on
Skin. Pediatric Dermatology, 1986. 3(2): p. 95-101.
44. Gwosdow, A.R., et al., Skin Friction and Fabric Sensations in Neutral and Warm
Environments. Textile Research Journal, 1986. 56: p. 574-580
45. Nacht, S., et al., skin friction coefficient: changes induced by skin hydration and
emollient application and correlation with perceived skin feel. J. Soc. Cosmet. Chem.,
1981(32): p. 55-65.
46. Sivamani, R.K., et al., Coefficient of friction: tribological studies in man - an
overview. Skin Research and Technology, 2003. 9: p. 227-234.
47. Elsner, P., D. Wilhelm, and H. Maibach, Frictional properties of human forearm and
vulvar skin: influence of age and correlation with transepidermal water loss and
capacitance. Dermatologica, 1990. 181(2): p. 88-91.
48. Cua, A.B., K.P. Wilhelm, and H.I. Maibach, Skin Surface Lipid and Skin Friction:
Relation to Age, Sex and Anatomical Region. Skin Pharmacology and Physiology,
1995. 8(5): p. 246-251.
49. Zhang, M. and A.F.T. Mak, In vivo friction properties of human skin. Prosthetics and
Orthotics International, 1999(23): p. 135-141.
50. Gerhardt, L.-C., et al., Skin-textile friction and skin elasticity in young and aged
persons. Skin Research and Technology, 2009(15): p. 288-298.
51. Bueno, M.-A., et al., Tribological investigation of textile fabrics. Wear, 1996. 195(1-
2): p. 192-200.
52. Kawabata, S., The Standardisation and Analysis of Hand Evaluation. 1980, Osaka:
The Textile Machinery Society of Japan.
53. Bueno, M.-A., Caractérisation de l'état de surface des textiles en vue d'une meilleure
compréhension de leur toucher, in Sciences Pour l'Ingénieur. 2002, Université de
Mulhouse: Mulhouse. p. 107.
54. Fontaine, S., Contribution à la compréhension des mécanismes tribologiques lors
d'une usure abrasive ou chimique de structures fibreuses. Application industrielle aux
tissus en PET-laine. , in Sciences Pour l'Ingénieur. 2001, Université de Haute Alsace:
Mulhouse. p. 228.
55. Bueno, M.-A., S. Fontaine, and M. Renner, Dispositif pour mesurer l'état de surface
d'un matériau et procédé de mise en oeuvre, LPMT, Editor. 2000: France.
56. Derler, S., et al., Friction of human skin against smooth and rough glass as a function
of the contact pressure. Tribology International, 2009. 42(11-12): p. 1565-1574.
57. Derler, S., U. Schrade, and L.C. Gerhardt, Tribology of human skin and mechanical
skin equivalents in contact with textiles. Wear, 2007. 263(7-12): p. 1112-1116.
58. Rosenbaum, D. and H.-P. becker, Review article. Plantar pressure distribution
measurements. Technical background and clinical applications. . Foot and Ankle
Surgery, 1997. 3: p. 1-14.
59. Lord, M., D.P. Reynolds, and J.R. Hughes, Foot pressure measurement: A review of
clinical findings. Journal of Biomedical Engineering, 1986. 8(4): p. 283-294.
60. Hayafune, N., Y. Hayafune, and H.A.C. Jacob, Pressure and force distribution
characteristics under the normal foot during the push-off phase in gait. The Foot,
1999. 9(2): p. 88-92.
61. Cock, A.D., et al., A functional foot type classification with cluster analysis based on
plantar pressure distribution during jogging. Gait & Posture, 2006. 23: p. 339-347.
62. Burnfield, J.M., et al., The influence of walking speed and footwear on plantar
pressures in older adults. Clinical Biomechanics, 2004. 19(1): p. 78-84.

69
63. Donaghue, V.M., et al., Longitudinal in-shoe foot pressure relief achieved by specially
designed footwear in high risk diabetic patients. Diabetes Research and Clinical
Practice, 1996. 31(1-3): p. 109-114.
64. Nagel, A., et al., Long distance running increases plantar pressures beneath the
metatarsal heads: A barefoot walking investigation of 200 marathon runners. Gait &
Posture, 2008. 27(1): p. 152-155.
65. Kernozek, T.W. and E.E. LaMott, Comparisons of plantar pressures between the
elderly and young adults. Gait & Posture, 1995. 3(3): p. 143-148.
66. Mills, N.J., Running shoe case study, in Polymer Foams Handbook. 2007,
Butterworth-Heinemann: Oxford. p. 307-327.
67. Peirce, F.T., The "Handle" of Cloth as a Measurable Quantity. Journal of the Textile
Institute, 1930: p. T377-T416.
68. Troyes, C.d.P.d.l.B.d. and C.d.R.d.l. Bonneterie, Mémento de technologie bonneterie.
1968, Troyes: grande imprimerie de Troyes. 25
69. Abrams, A. and L. Mishcon, Pattern Wheel Designing for Circular Jersey Knitting
Machines. first edition ed. 1949, New York: Supreme knitting machine co., inc. 165.
70. Committee, T.I.T.T.a.d., Textile Terms and Definitions. Ninth Edition ed. 1991,
Manchester: M C TUBBS BSc Tech AMCT CText FTI FRSA P N DANIELS BA
MIInfSc. 367.
71. Harris, M., Handbook of textile fibers. first edition ed. 1954, Washington: Harris
research laboratories, inc. 356.
72. Woolman, M.S. and E.B. McGowan, Textiles: a handbook for the student and the
consumer. third ed. 1946, New York: The macmillan company. 388.
73. Pribut, S. and D. Richie, Separating the Buzz from the Biomechanics: A guide to
Athletic Shoe trends & Innovations. Podiatry Management, 2004: p. 85-97.
74. Pac, M.-J., Les surfaces textiles aux échelles micro-méso et macroscopiques :
propriétés thermiques et tribologiques, in Sciences Pour l'Ingénieur. 2001, Université
de Haute-Alsace: Mulhouse.
75. Bernhardt, F., Double layer sock with low friction layer to layer interface. 2000:
United States.
76. Bertaux, E., et al., Textile, Physiological, and Sensorial Parameters in Sock Comfort.
Textile Research Journal, 2010.

70
 Chapter II:
materials and methods

71
 Résumé

L’objectif de ce travail de recherche était d’étudier l’influence de la structure textile

sur le frottement entre la peau du pied et la chaussette lors de la course à pied. Le premier
chapitre a recensé les différentes affections cutanées qui peuvent survenir lorsque l’on
pratique le jogging. Ces affections peuvent apparaître à différents endroits et la pression
exercée sur la peau par l’ensemble chaussure + chaussette dépend de la zone considérée. Pour
pouvoir mesurer le frottement entre la peau et la chaussette dans les mêmes conditions que
lors de la course à pied, il faut déterminer l’aire de contact entre le pied et la chaussette,
l’amplitude de frottement, la fréquence de frottement et la force normale appliquée sur la
peau. Le chapitre II détaille successivement les zones du pied étudiées, les chaussettes et les
dispositifs de mesures de frottement et de compression utilisés.

Les informations disponibles dans la bibliographie sur les zones d’apparition des
phlyctènes ne sont pas suffisamment détaillées pour bien les identifier. J’ai par conséquent
réalisé un sondage afin de localiser les zones du pied où les ampoules apparaissent en
fonction de l’activité sportive pratiquée. Cinq sites ont ainsi été identifiés : le coussinet sous
le gros orteil sur la plante des pieds (cité dans 56% des cas), la voute plantaire (citée dans
50% des cas), le talon (cité dans 56% des cas), les côtés externes du gros et du petit orteil
(cités dans 25% des cas).

En ce qui concerne les chaussettes de course à pied, les modèles basiques sont en
coton renforcé de polyamide ou de polyester. Ils contiennent aussi parfois de l’élasthanne en
faible quantité (2 à 3% en masse) aux zones élastiques, ou bien de l’acrylique pour diminuer
l’humidité à la surface de la peau ou encore du PolyTetraFluoroEthylène afin de diminuer le
coefficient de frottement peau-chaussette. Les chaussettes d’hiver sont à base de laine tandis
que certaines marques proposent aussi des chaussettes à base de polyamide ou de polyester
qui permettent de garder les pieds au sec. Dans notre étude, nous utilisons des chaussettes en
coton car c’est le principal composant des chaussettes basiques. Quant à la structure tricotée,
les chaussettes de course à pied sont généralement soit en jersey simple soit en jersey
bouclette. Nous avons choisi 4 liages : le jersey simple qui a une structure fine tricotée avec
un seul fil, le jersey double qui présente la même structure mais tricotée avec deux fils, le
jersey ponts de fils qui, en plus de la structure de base en jersey simple, présente des flottés

72
formés avec un deuxième fil, et le jersey bouclette dont l’épaisseur importante est obtenue par
les boucles que forme le deuxième fil avec le fil de fond. De plus, six autres chaussettes ont
été étudiées afin d’évaluer l’influence du titre des fils, du liage et de la longueur de fil
absorbée. Les comptes i.e. le nombre de colonnes et de rangées de mailles par cm, ou bien le
nombre de ponts de fils ou de bouclettes par cm, ont été mesurés pour chaque type de
chaussette ainsi que le diamètre moyen des fils. Pour les jerseys bouclette, l’angle décrit par
les bouclettes avec la verticale au tricot et la hauteur des boucles ont également été
déterminés. Tous les tests effectués ont été réalisés à sec en atmosphère conditionnée standard
i.e. à 20°C ± 2°C et 65% ± 4% d’humidité relative. Les matériaux textiles ont été
conditionnés au préalable durant 48 heures.

Un tribomètre alternatif linéaire à poids mort a été mis en place au Laboratoire de

Physique et de Mécanique Textiles afin de reproduire le plus fidèlement possible le frottement
de l’ensemble chaussure + chaussette sur la peau lors de la course à pied. La chaussette est
collée sur une table oscillante dont le moteur est commandé par un programme informatique.
La table oscillante décrit un mouvement sinusoïdal. Le palpeur, fixé sur un bras, va venir
frotter la face envers de la chaussette i.e. celle qui est en contact avec la peau au porter. Le
dispositif de fixation de la charge normale est placé au dessus du palpeur. Un capteur à jauge
de déformation est placé sur le bras afin de mesurer la force de frottement. Le déplacement
vertical du palpeur est enregistré grâce à un capteur laser. Le palpeur a une longueur de 1,7
cm et une largeur de 0,9 cm pour une aire de contact de 1,53 cm² ce qui correspond à la
surface de l’index, mesurée grâce à un sondage sur 58 personnes. Le palpeur est recouvert
d’une peau artificielle en polyamide enduit de polyuréthane, le Lorica®, qui présente une
surface similaire à la peau et reproduit de façon satisfaisante les propriétés tribologiques de la
peau sèche. Deux autres palpeurs, de forme cylindrique en acier inoxydable de surface égale
au premier et qui ne diffèrent que par leur rugosité, ont également été utilisés afin de mieux
comprendre les phénomènes qui se produisent lors du frottement. La distance de frottement a
été fixée à 5mm. La force normale exercée sur la peau par l’ensemble chaussure + chaussette
ainsi que la fréquence des foulées ont été mesurées in situ grâce à un capteur piezorésistif
flexible, le Flexiforce®, pour une vitesse de course de 8 à 13 km/h. Nous avons concentré
notre étude expérimentale sur la zone de la voute plantaire. Pour cette zone, la fréquence des
foulées a été fixée à 1,5 Hz et la force normale appliquée à une plage de 0,44 à 2,14 N.

73
 Un deuxième tribomètre a aussi été utilisé, le Textile Friction Analyser (TFA), qui a
été développé par l’EMPA Swiss Federal Laboratories for Materials Science and Technology
St Gallen. Un bras d’élévation permet d’ajuster la force normale appliquée. Le mouvement
alternatif du support métallique est programmé informatiquement et géré par un moteur. Ce
tribomètre est assez différent du précédent ; le palpeur est rond de diamètre 2,8 cm, il a donc
une surface de 6,16 cm² alors que celui du LPMT est rectangulaire et a une surface de
1,53 cm². Les expériences de frottement ont donc été réalisées pour une pression donnée. De
plus, sur le dispositif du LPMT, le palpeur est libre de bouger verticalement selon l’influence
de la structure tricotée ce qui n’est pas le cas sur le TFA. Par conséquent, pour le tribomètre
du LPMT, la force normale appliquée est constante alors que, dans le cas du TFA, la force
normale varie autour d’une valeur d’équilibre plus élevée. Ni le tribomètre du LPMT ni la
TFA ne reproduit fidèlement le frottement de la chaussette sur le pied dans la chaussure lors
de la course à pied. Ce qui se produit dans la chaussure est probablement un cas intermédiaire,
c’est pourquoi il était intéressant de réaliser des mesures de frottement sur les deux
tribomètres et de les comparer.

La pression exercée sur la peau du pied par l’ensemble chaussure + chaussette dépend
de nombreux facteurs dont le comportement de la chaussette en compression. C’est pourquoi
des mesures de compression ont été effectuées sur un appareil KES FB3, dispositif de la
chaine de mesures Kawabata dédié à l’évaluation de la compression des matériaux textiles.
Un capteur Flexiforce® a été fixé sur le pied presseur, les chaussettes ont été placées avec la
face endroit vers le haut (face en contact avec la chaussure) et l’envers vers le bas (face en
contact avec le pied). Ainsi, la pression exercée par le pied presseur sur l’endroit est mesurée
par le Flexiforce® et correspond à la pression exercée par la chaussure sur la chaussette
tandis que la pression exercée sur l’envers de la chaussette est mesurée par le capteur de force
du KES FB3 et correspond à la pression exercée sur la peau. La vitesse de déplacement du
pied presseur a été fixée à 0,02 mm/s et la pression maximale à 15 kPa.

74
 CONTENTS

Chapter II: Materials and methods ____________________________________________ 71

I. TESTED FOOT AREAS ______________________________________________ 76
II. RUNNING SOCKS___________________________________________________ 77
1. Fibre composition 77
2. Textile structure 79
3. Setting, yarn diameter, terry height and orientation 83
4. Textile testing conditions 85
III. LINEAR RECIPROCATING DEAD WEIGHT TRIBOMETER _____________ 85
1. Design of the pin 87
A. Surface area 87
B. Artificial skin model 88
C. Additional pins 89
2. Characteristics of the reciprocating movement 89
A. Sliding movement amplitude 89
B. Sliding movement frequency and applied normal force 90
IV. TEXTILE FRICTION ANALYSER (TFA) ______________________________ 92
V. COMPRESSION EXPERIMENTS ______________________________________ 95
Literature cited __________________________________________________________ 99

75
 The aim of this research work was to study the influence of the textile structure on the
friction between the foot skin and the sock during running. The skin injuries that may be
induced by jogging were detailed in chapter I. They may occur on different anatomical sites
and the pressure on the skin is highly dependent on the anatomy. In order to measure the
friction between the foot and the sock in similar conditions than in a shoe during running, the
contact area between the foot and the sock, the sliding movement amplitude and frequency,
and the normal force applied on the skin need to be fixed. This chapter successively details
the foot areas we studied, the socks we used and the friction and compression measurement
devices and settings.

I. TESTED FOOT AREAS

According to Mailler, Adams and Bergeron, blisters usually occur on the tips of the
toes, the balls of the feet and the posterior heel [1, 2]. Herring and Richie [3] found that the
forefoot region is more affected than the heel region. 60.2 % of the blisters they observed
occurred in the forefoot region, 33.3 % in the midfoot region and 6.5% in the rearfoot area.
This information is not sufficiently detailed to identify the preferential areas for blister
formation on the foot. I thus carried out a survey asking people to localise on foot anatomy’s
pictures the zones where they usually have blisters and to specify how often these zones were
injured and during which sport activity. 16 participants answered the survey, 7 women and 9
men, from 25 to 55 years old. 5 preferential sites were identified, namely the plantar area
under the first metatarsal head, the plantar fascia area, the heel, the external side of the hallux
and the external side of the little toe. These five zones are illustrated in figures 1 to 3.

76
 Figure 20: Zone 1; under the first metatarsal head.

Figure 21: Zones 2 and 3; respectively the plantar fascia and the heel (left foot).

Figure 22: Zones 4 and 5; respectively the external sides of the big and the little toe.

Zones 1 and 3 were cited by 56% of the participants, half of the people who answered
the survey had blisters localised at zone 2 and zones 4 and 5 were mentioned by 25% of the
participants.

II. RUNNING SOCKS

12 BDF8E59BCB78
BDF8E59BCB78
Basic running socks are mostly made of cotton, usually combined with polyamide or
polyester. Polyester is less expensive but also less resistant than polyamide. When combined
with cotton, polyamide or polyester is used as plating* to increase the sock’s lifetime. Socks

*
plating: the plating (vanisage in French) is a reinforcement technique. For example, the toe and the heel of a
sock can be reinforced using an additional yarn.

77
may also contain elastane§ in very small quantity (2 to 3% in mass) in specific elastic areas, or
acrylic in order to reduce moisture at the sock - foot skin interface or either
polytetrafluoroethylene§ (PTFE) whereby the coefficient of friction between the skin and the
sock may be reduced. Winter running socks commonly contain a high percentage of wool
fibres for the runner to be warm. The compositions of different worldwide commercialised
running socks are given in Tables 1 to 3.

BRAND % OF % OF % OF % OF % OF % OF
NAME COTTON PA ACRYLIC PTFE PP ELASTANE
RUN 100 48 32 20
RUN 500 57 16 25 2
RUN 50 65 35
RUN 400 60 12 37 1
RUN 300 50 27 20 3
RUN 800 66 32 2
RUN 900
light 69 19 12
RUN 1000 57 16 25 2

Table 4: Fibre compositions of socks sold by Décathlon. The abbreviations PA, PTFE and PP
respectively stand for polyamide, polytetrafluoroethylene and polypropylene.

% OF % OF % OF % OF % OF
BRAND NAME COTTON PA PP ELASTANE PET
TACTEL X1 98 2
RUN X3 80 2 18
RUN X2 CMX 40 2 58
Oasics KAYANO 78 8 14
Oasics RUN
QUARTER 60 38 2
Nike Elite RUN 72 1 27

78
Table 5: Fibre compositions of socks sold by Go Sport. The abbreviation PET stands for
polyethylene terephtalate.
% % %
BRAND % OF % OF OF % OF OF % OF OF
NAME WOOL COTTON PA ACRYLIC PP ELASTANE PET
New
Balance 100
Rywan 17 3 80
Nike
ergonomique 26 3 71
Nike dry-fit 28 68
Oasics 32 68
Rohner
winter 45 16 27 11 1
running
Running
hurricane 17 72 8
X Socks
Speed One 38 38 18 6

Table 6: Fibre compositions of specific sport trademark running socks.

Sport’s brands such as New Balance, Oasics or Nike propose specific running socks
which are mostly made of polyamide or polyester for the foot to remain dry. These socks are
not commercial successes, maybe because they are expensive for recreational runners. For our
study, we selected cotton made socks as cotton is the principal component of basic socks.

2 #F&CBF89CD4EC4DF8
#F&CBF89CD4EC4DF8
There are two knitted structures widely used in running socks, namely terry and simple
jersey (cf. table 4). These two structures were described in the previous chapter
(paragraph III. 1).

79
 KNITTED RUN RUN RUN RUN RUN RUN RUN 900 RUN
STRUCTURE 100 500 50 400 300 800 LIGHT 1000
Simple jersey X X X
Terry jersey X X X X X

Table 7: Knitted structure of Décathlon’s running socks.

a) b)

c) d)

Figure 23: Labonal socks: a) simple jersey, b) double jersey, c) yarn bridges and d) terry
jersey.

80
 Four different kinds of knit were selected:
1 Labonal simple jersey
This thin structure is knitted with a single cotton yarn.

1 Labonal double jersey

The thickness of the structure is slightly increased knitting two cotton yarns
together.

1 Labonal yarn bridges

This structure is also derived from simple jersey. Here, floats§ are formed by a
second yarn.

1 Labonal terry jersey

The important thickness of this kind of knit is due to the loops formed by the pile
yarn with the back yarn.

Figure 4 shows these four kinds of knit photographed with a zoom ×40. These four
structures were knitted with a gauge§ 14.

Six others cotton-made socks were tested in order to study the influence of the yarn
count, the knitted structure and the stitch length:

1 Terry jersey 10 – 12, terry jersey 10 – 16 and terry jersey 10 – 20

These socks only differ by their yarn count (cf. table 5).

TERRY JERSEY TERRY JERSEY TERRY JERSEY

SOCK 10 - 12 10 - 16 10 - 20
Yarn count in Ne(c) 12 16 20
Yarn count in Tex 141 106 85

Table 8: Yarn count for terry jersey 10 - 12, terry jersey 10 - 16 and terry jersey 10 - 20.

81
 1 Terry jersey 10 – 16 and simple jersey 10 – 16
They only differ by their knitted structure; one is a terry jersey sock while the
other one is a simple jersey sock. The terries of the terry jersey sock are less
compact, less homogeneous and more crushed than Labonal terries.

1 Terry jersey 9 – 16, terry jersey 10 – 16 and terry jersey 11 – 16

They only differ by their size, respectively 9 ¼, 10 ¼ and 11 ¼ inches which is
linked to the stitch length (cf. table 6). The stitch length is the length of yarn
contained in one stitch and is expressed in cm of yarn per stitch.

TERRY JERSEY TERRY JERSEY TERRY JERSEY

SOCK 9 - 16 10 - 16 11 - 16
Stitch length in cm per stitch 1.19 ± 0.02 1.22 ± 0.01 1.28 ± 0.01

Table 9: Stitch length for terry jersey 9 – 16, terry jersey 10 – 16 and terry jersey 11 – 16.
The stitch lengths were measured using a Sodemat device (Troyes, France) in accordance
with standards NFG07101 and NFG07311.

leg

foot
heel
sole

toe

Figure 24: Different sections of a sock.

A sock can be decomposed into five sections (cf. figure 5): the leg, the foot, the toe,
the sole and the heel. Three sections, namely the toe, the sole and the heel are in contact with

82
skin areas where blisters may occur. Only these parts were studied. The way a sock is
assembled was not considered. It seems obvious that the more discontinuity and thickness the
sock sewing up presents, the higher friction it will induce on the foot skin.

2 FCCB7-8$6D78B65FCFD-8CFDD$8FBC
FCCB7-8$6D78B65FCFD-8CFDD$8FBC8
FBC8678DBF7C6CB78
678DBF7C6CB78
The setting i.e. the number of yarns or terries per cm in both directions of the
investigated socks in loom-finished state are given in table 7.
The diameters of the yarns were measured using a Projectina® 4002 device
(Projectina, Heerbrugg, Switzerland, serial number 8070). Objective 35004 (×40) and lens
170/10/0.25 were employed whereby a 1 mm read equals to 8 µm in reality. 10 and 6
measures were respectively carried out for Labonal and other socks. The results are shown in
table 7.
The height and the angle of terries with the vertical to the textile fabric were measured
from photographs. The mean values are reported in table 8 (at least 10 measurements) and
figure 6 shows an example of a photograph used to assess these parameters.

SOCK LENGTHWISE WIDTHWISE MEAN YARN

SETTING SETTING DIAMETER IN MM
Labonal simple jersey 16 courses 9 wales 0.36 ± 0.03
Labonal double jersey 14 courses 8 wales 0.43 ± 0.03
Labonal yarn bridges 7.5 bridges and 4.5 bridges and 0.38 ± 0.08
16 courses 10 wales
Labonal terry jersey 17 9.5 0.39 ± 0.05
Terry jersey 10 - 12 9 6 0.51 ± 0.05
Terry jersey 10 - 16 8.5 6 0.36 ± 0.03
Terry jersey 10 - 20 9 6 0.34 ± 0.03
Simple jersey 10 - 16 9 courses 6.5 wales 0.40 ± 0.06
Terry jersey 9 - 16 11 7 0.41 ± 0.07
Terry jersey 11 - 16 7.5 6.5 0.38 ± 0.04

83
Table 10: Lengthwise, widthwise settings of the socks and mean diameters of the yarns
without tension. For double jersey, it is the diameter of one single yarn which is given. For
yarn bridges, the values stand for the bridge yarn. For terry socks, the results are given for
the terry yarn.
1
1

KNITTED FABRIC TETAINIT (°) HEIGHT (MM)

Labonal terry jersey 5±5 1.38 ± 0.18
Terry jersey 10 - 12 -41 ± 6 1.06 ± 0.11
Terry jersey 10 - 16 -45 ± 7 1.04 ± 0.09
Terry jersey 10 - 20 -51 ± 8 1.07 ± 0.13
Terry jersey 9 - 16 -36 ± 11 1.09 ± 0.05
Terry jersey 11 - 16 -52 ± 8 1.11 ± 0.16

Table 11: Height and orientation of terries. Tetainit is the angle formed by the terries with the
vertical to the textile surface, it is positive when the terries are oriented towards the heel of
the sock and negative when the terries are oriented towards the toe of the sock. 1
1

Figure 25: Example of a photograph used to assess the height and the orientation of terries.

84
 2 #F&CBF8CF9CB78E7BCB798
#F&CBF8CF9CB78E7BCB798
As it has been detailed in the first chapter, textile materials can be highly hydrophilic.
In a moist environment, textile fibres absorb water from the air and swell. The properties of
textile materials are therefore highly influenced by ambient humidity and temperature. In
order to compare different textile materials, international standard conditions have been fixed
for textile testing experiments i.e. 20°C ± 2°C and 65% ± 4% Relative Humidity (european
standard NF EN ISO 139). Socks were preconditioned in this climate for at least 48 hours
before the experiments.
Comparative friction results of different textile surfaces obtained in dry conditions
remain valid in wet conditions [4]. Therefore, all the experiments described in chapters 2 and
3 were carried out in dry conditions.

III. LINEAR RECIPROCATING DEAD WEIGHT

TRIBOMETER

A specific device was designed to simulate the friction of the sock on the foot in the
shoe during running. It is composed of an oscillating motor-controlled table (Controls Linear
Table [Link] LT 100-2 ST controlled by 3 Chanel Stepper-Motor-Controller [Link], LINOS
Photonics GmbH & Co. KG, Göttingen, Germany) on which the sock is stuck using a
removable and high resistant glue. A computer program has been written in Visual Basic
language whereby the oscillating table can have a sinusoidal movement. The back side of the
sock i.e. the one which touches the skin in the shoe is in contact with a probe which is
mounted on an arm. Weights can be applied overhead of the probe. A strain-gauge sensor
(HBM PW4C3 sensor, maximum load 3 N, HBM France SAS, Mennecy, France) is clamped
on the arm whereby friction forces can be recorded. A laser sensor has been installed on the
tribometer in order to measure the vertical displacement of the pin during a friction
experiment.

Figure 7 is a photograph of the tribometer while figure 8 shows the curve shape of the
oscillating table displacement in function of time for the settings which will be detailed later.

85
 Figure 26: Linear alternating friction device.

7
927ABCD434EF56338

8
8 7 6 5 4 3 9 A
27

26

25

BCDE1FEDCC1
24
FDE1FEDCC12
BCC117361
23

12345678

Figure 27: Displacement of the oscillating table versus time.

86
 12 .F9B788CF8B7
.F9B788CF8B78
B78

12 BA4E794B749
When developing our tribometer, we tried to design a highly versatile apparatus which
can be used to evaluate at the same time the friction between the sock and the foot in the shoe
and the friction between the finger and a fabric. The index is the finger the most commonly
used to assess the tactile properties of a fabric. The skin of the index and the skin of the foot
are both glabrous and when considering the areas of the foot where blisters commonly occur,
the contact area involved in blister formation seems to be similar to the surface area of the
index. Moreover, the contact pressure used to evaluate tactile properties ranges from 0.15 to
4.5 N [5] and, as it will be detailed later in section III.2.B, the contact pressure of the foot
with the sock during running varies from 0.44 to 20 N. The other parameters namely the
sliding distance and the sliding velocity are adjustable. I therefore measured the surface area
of the index in order to manufacture a pin with the same surface.

To measure the surface area of the index, 58 volunteers, 40 men and 18 women, aged
from 20 to 65 years were asked to tincture their index with red ink using a stamp pad and to
press it against a paper sheet placed on a precision balance. They placed their index on the
paper sheet as if they were touching a textile fabric and pressed it until they reached a force of
0.78 N (i.e. 80g read on the precision balance). I then scanned the fingerprints, converted each
image into jpeg format and analysed them using ImageJ program. The calibration was realised
with a square of 2 cm side length. The fingerprints were considered ellipsoidal. The mean
contact area of the index with the paper sheet was 1.56 ± 0.38 cm². It was slightly higher for
men (1.60 ± 0.37 cm²) than for women (1.46 ± 0.37 cm²). The pin used in friction
experiments was thus 1.7 cm long and 0.9 cm wide i.e. it has a surface area of 1.53 cm². This
result is similar to the values given by Breugnot [5] and Derler [6] who respectively used a
surface area of 1.6 cm² for the mean contact area between the finger and fabrics and measured
contact areas from 1.5 to 3 cm² in touch experiments. The pin has fully rounded edges in
order to reduce boundary effects.

87
 2 1BFDADED498D95C79

Figure 28: Light-optical micrograph showing, on the left, the typical surface topography of
the skin model Lorica® and, on the right, that of in vivo human skin (back of the hand) [7].

The pin is covered with a polyurethane coated polyamide microfibre fleece; Lorica®
(polyurethane-coated polyamide, Lorica® Soft, Lorica Sud, Milano, Italy). Lorica® presents
a surface structure similar to that of skin (cf. figure 9) and was found to simulate the friction
behaviour of human skin in dry conditions [6]. The pin is covered with Lorica® using
double-face scotch-tape as shown on figure 10.

Figure 29: Tribometer pin recovered with Lorica®.

88
 $2 1DFDC49D89
In order to better understand the phenomena which take place during the friction of the
socks, we wanted to compare the friction results obtained with two pins which only differ by
their roughness. A product having the same material properties than Lorica® Soft combined
with a smooth surface is not available. Comparative friction tests were therefore carried out
with two cylinder-shaped pins made of stainless steel: a smooth one and one with 1mm
diameter balls made of stainless steel stuck on its surface which will be called the “rough” one
(figure 11).

Figure 30: the smooth and “rough” pins made of stainless steel. These pins have the same
surface area (in contact with the sock) than the Lorica® pin.

To investigate the influence of the counterpart material on the friction performances,

the socks were rubbed against smooth polyurethane because Lorica® is a polyurethane coated
polyamide. The mean arithmetic roughness of Lorica® and smooth polyurethane was
measured using a Taylor – Hobson Surtronic 3 roughness meter under 100 mN with a tip of
5 mm. It respectively equals 19.44 ± 1.69 µm and 0.51 ± 0.16 µm for Lorica® and smooth
polyurethane.

2 )6D6ECFDB9CBE988CF8
)6D6ECFDB9CBE988CF8DFEBDE6CB7
BE988CF8DFEBDE6CB78
DFEBDE6CB785"F5F7C8
5"F5F7C8

12 DD695C'757F945DF79
The distance over which the foot slides in the shoe during running is not easy to
measure. However, the size of blisters can give us an idea of this distance. According to
Knapik et al. [8], the size of blisters corresponds indeed to the area where the rubbing is
applied. Herring and Richie [3] observed blisters from 0.087 to 0.812 cm². Moreover, blisters

89
of less then 5 mm in diameter were qualified as small blisters by Bergeron [2]. We therefore
presumed that the sliding distance of the foot in the shoe should not far exceed 5mm and thus
fixed the sliding distance of the pin on the sock to this value.

2 DD695C'757F9AB77E9494D79CB549ACBE799
We have seen in the previous chapter that the pressure the shoe – sock system applies
on the foot skin depends on the area of the foot and on the running speed. The literature
indicates the areas of the foot which experience the highest forces but no data is available on
the forces exerted on the areas we wanted to study (section I). We therefore measured the
force at these specific areas. We thus needed a force measurement system which can be
inserted in the shoe of a volunteer without modifying the force exerted by the footwear on its
foot skin and without disturbing the runner so that its running gait remained unchanged.
Insoles systems have a nonnegligible thickness and therefore reduce the force on the foot.
Moreover, we needed to record the force on skin areas outside the foot sole. A flexible
piezoresistive force sensor FlexiForce® (Tekscan, South Boston, USA) matched our needs
(figure 12).

FlexiForce® is 208 µm thick, 14 mm wide and its length can vary from 51 to 203 mm.
Its active zone is round shaped and has a diameter of 9.53 mm i.e. it has a 0.71 cm² active
surface. The electrical resistance of the sensor is inversely proportional to the applied load.
The response time of this sensor is 5 µs; impact forces can thus be recorded. When running,
the force applied on a specific area of the foot may change very quickly as the foot is hardly
never motionless; a low response time was therefore necessary. FlexiForce® can work either
in cold or hot environments (from -9°C to 60°C) and can support forces up to 440N
depending on the model.

Figure 31: FlexiForce® sensor.

90
 Figure 32: Record of a run at 8 km/h with FlexiForce® placed at the plantar fascia.

Signals are recorded using a data acquisition system SDM 10 and CADM software
(Doerler Mesures, Vandoeuvre-lès-NANCY, France). Figure 13 shows a record obtained with
a FlexiForce® sensor. On this graph, the voltage, which corresponds to the force (sensor
calibration), is plotted in function of time. From the recorded force signal, peak forces, mean
force and stride frequency can be calculated.
5 volunteers were asked to run on a treadmill with their own footwear in a climate
controlled room (20°C and 65% RH). Three women and two men, aged from 25 to 47 years
old, with weights between 52 and 76 kg participated in this study. The sensor was fixed on
their foot with adhesive tape. Then, they chose their running speed and first got used to the
treadmill. Thereafter, one run was recorded for each foot area (zones 1 to 5). Running speeds
varied from 8 to 13 km/h. The sliding frequency of the foot in the shoe was assumed to be
equal to the stride frequency i.e. 1.2 to 1.5 Hz for a running velocity of 8 to 13 km/h. The
forces and sliding frequencies recorded are given in table 9.

91
 We focused our measurements on the plantar fascia area. The sliding frequency was
then fixed to 1.5 Hz and the vertical forces ranged from 0.5 to 2.2 N.

AREA OF FIRST PLANTAR HEEL BIG LITTLE

THE FOOT METATARSAL FASCIA TOE TOE
HEAD
Force range (N) 5 to 20 0.44 to 2.14 0.87 to 4.4 1 to 15 1.89 to 9.33
Sliding frequency
range 1.2 to 1.5 Hz

Table 12: Force and frequency ranges.

IV. TEXTILE FRICTION ANALYSER (TFA)

This device has been developed by the EMPA [9] to monitor normal and friction
forces at the same time. It includes an elevation arm (cf. figure 14) whereby friction test
functionalities can be implemented by measuring and controlling vertical load over an
adjustable force range up to 20 N. The reciprocating motion of the metallic support is created
by a stepper motor which is operated by a programmable controller. Friction forces are
measured using a highly sensitive quartz load cell (Type 9203, Kistler, Winterthur,
Switzerland) with a maximum resolution of 5 mN connected to a charge amplifier (Type
5011B, time constant: 10s). Vertical forces are recorded using a strain-gauge force transducer
(Type U9B, HBM, Darmstadt, Germany) with a maximum resolution of 10 mN. The position
of the manoeuvrable weight can be changed whereby the normal load can be adjusted. The
weight is driven by a DC motor (A-max 26, Maxon, Sachseln, Switzerland).

The TFA is different from the LPMT’s tribometer in terms of contact surface, applied
normal load and sliding velocity. Firstly, the LPMT’s tribometer has a chamfered rectangular

92
pin of 1.53 cm² surface while the TFA pin is round shaped with a diameter of 2.8 cm i.e. a
surface of 6.16 cm². Friction experiments were then realised for the same pressures.
Secondly, the LPMT’s tribometer is a dead weight tribometer i.e. the pin is free to
move up and down under the influence of the textile structure during a friction test. On the
TFA, the distance between the pin and the elevation arm is fixed when setting the device (the
manoeuvrable weight is moved until the strain-gauge sensor perceives the desired normal
load). In fact, the system is also composed of a dead weight but the elevation arm is linked to
its support by a pivot and a spring gives rigidity to the system (cf. figure 14). The pin thus
cannot raise itself up freely. Therefore, for the LPMT’s tribometer, the applied normal load is
constant and corresponds to the dead weight whereas for the TFA, the applied normal load
varies around a mean value which is higher than the dead weight of the LPMT’s tribometer
for the same setting value. The vertical load applied by the TFA versus time for a set point of
9.49 N is given figure 15.
Thirdly, the sliding velocity is triangular shaped for the LPMT’s tribometer while for
the TFA, the sliding velocity signal is 40% ramp shaped (20% ascending ramp to reach the set
point velocity then 60% constant velocity and 20% descending ramp).
Finally, friction force signals recorded using the LPMT’s tribometer are noisy
compared to the signals recorded using the TFA because on the LPMT’s tribometer the
friction force signal is not filtered while a low-pass filter is applied to the data recorded on the
TFA.

Neither the linear alternating dead weight tribometer nor the TFA exactly reproduces
the friction of the foot with the sock in the shoe. For the LPMT’s tribometer, the weight is
free to move vertically while in the case of the TFA, the system is fixed. In-shoe conditions
are probably an intermediate between these two cases. Friction experiments were therefore
carried out on both tribometers. Both the TFA and the LPMT’s tribometer can be improved in
order to better reproduce the friction between the sock and the foot of a runner during
jogging. The materials in contact, the pressure and the sliding frequency have been adapted
but the experimental measures do not entirely simulate the frictional conditions on anatomical
body shapes. The use of curved rubbing elements could, in a further study, simulate the
contact of textile fabrics against skin overlying bony prominences. The pin may also be
designed in order to reproduce the shape of the tested foot area. Moreover, a shoe insole may
be placed under the sock.

93
 .

Spring

Figure 33: TFA device [7].

94
 9

33
#ECF1BDE1EDF1!$"

3

43

4

53
8 83 7 73 6 63 5

C1!"

Figure 34: Normal load applied by the TFA in function of time for a set point of 9.49 N.

V. COMPRESSION EXPERIMENTS

The pressure exerted on the foot by the footwear depends on the shape and the
cushioning properties of the shoe, the friction and compression behaviour of the sock and on
the skin characteristics. Compression tests for socks were implemented on KES FB3 [10]. A
FlexiForce® sensor was placed on the compression plunger surface with double-face scotch-
tape and the socks were positioned with the face up and the back side down as shown on
figure 16. The pressure the plunger applied on the face of the sock was recorded by the
FlexiForce® sensor and corresponds to the pressure exerted by the shoe on the sock. The
pressure transmitted to the back of the sock is recorded by the force transducer of the KES
FB3 and corresponds to the pressure exerted by the footwear on the foot skin. The
FlexiForce® signal was recorded on a computer using the related software and the KES FB3
signal was saved using an oscilloscope.

95
 The following settings were chosen:
FlexiForce® software and oscilloscope:
1 100 points/s.
1 The measure is manually and automatically started for the FlexiForce® and
the oscilloscope respectively.
1 The measure lasts for 200 s for terry jersey socks and for 100 s for the other
knitted structures.

Figure 35: KES FB3 equipped with FlexiForce®.

KES FB3:
1 Plunger velocity: 0.02 mm.s-1.
This velocity is of course very low compared to the velocity with which the foot
touches the ground when running; it was limited by the sampling frequency of the
Flexiforce®. However, compression experimentations realised on velvet fabrics by
Camillieri [11] at velocities from 0.02 to 0.2 mm.s-1 show a negligible influence of the
velocity on the compression results.
1 Maximum pressure of 15 kPa.
96
 %DE&BD1C1'(1DEC

5
#ECF1BDE1BC1!/EC0BC.1$"

63

73

83

8
8 83 7 73 6 63 5

)CBC*CF1BDE1BC1!+,-.1$"

Figure 36: One calibration test; applied normal force (measured with Flexiforce®) versus
perceived normal force (measured with KES transducer). The black curve is the calibration
curve while the grey line is the expected result.

A calibration test was first carried out. During the calibration test, the KES plunger
applied a force on the surface of the Flexiforce® i.e. on 0.71 cm² while the force measured by
the KES force transducer was measured on 2 cm². As there is no sample on the compression
tester, the transmitted force should be equal to the applied force. On the other hand, for tests
with textile fabrics, as the plunger of the KES sinks in the textile sample, the applied force is
distributed over the entire surface of the plunger i.e. 2 cm², the same surface on which the
transmitted force is measured. Calibration test results are thus displayed in force while the
results of the compression tests with textile samples are given in pressure.

97
 Calibration test results are reported in figure 17 which displays the applied normal
force i.e. the force measured by the Flexiforce® in function of the perceived normal force i.e.
the force measured by the KES force transducer. It can be noticed that the behaviour of the
Flexiforce® is not linear, the grey area corresponds to a perceived normal force higher than
the applied normal force which is not possible physically. The compression results will be
analysed without considering this zone. The latter zone will be identified by a grey area on the
compression curves.

98
 Literature cited

1. Mailler, E.A. and B.B. Adams, The wear and tear of 26.2: dermatological injuries
reported on marathon day. British Journal of Sports Medicine, 2004. 38: p. 498-501.
2. Bergeron, B.P., A Guide to Blister Management. The physician and sportsmedicine,
1995. 23(2): p. 37-46.
3. Herring, K.M. and D.H. Richie, Friction blisters and sock fiber composition A double-
blind study. Journal of the American Podiatric Medical Association, 1990. 80(2): p.
63-71.
4. Gwosdow, A.R., et al., Skin Friction and Fabric Sensations in Neutral and Warm
Environments. Textile Research Journal, 1986. 56: p. 574-580
5. Breugnot, C., Contribution à la caractérisation mécanique du toucher des surfaces
textiles à partir de critères neurosensoriels, in Sciences Pour l'Ingénieur. 2005,
Université de Haute Alsace: Mulhouse. p. 209.
6. Derler, S., U. Schrade, and L.C. Gerhardt, Tribology of human skin and mechanical
skin equivalents in contact with textiles. Wear, 2007. 263(7-12): p. 1112-1116.
7. Gerhardt, L.-C., Tribology of human skin in contact with medical textiles for decubitus
prevention, in Sciences. 2008, ETH Zurich: Zurich. p. 206.
8. Knapik, J.J., et al., Friction Blisters - Pathophysiology, Prevention and Treatment.
Sports Medicine, 1995. 20(3): p. 136-147.
9. Gerhardt, L.-C., et al., Study of skin–fabric interactions of relevance to decubitus:
friction and contact-pressure measurements. Skin Research and Technology,
2008(14): p. 77-88.
10. Kawabata, S., The Standardisation and Analysis of Hand Evaluation. 1980, Osaka:
The Textile Machinery Society of Japan.
11. Camillieri, B., J.-M. Praëne, and M.-A. Bueno, Simulation du frottement des surfaces
fibreuses à partir de la compression transversale, in 22èmes Journées Internationales
et Francophones de Tribologie. 2010: Albi, France.

99
 Chapter III:
experimental results

100
 Résumé

Des expériences de frottement et de compression ont été réalisées afin d’étudier le

contact mécanique entre la chaussette et la peau lors de la course à pied. Une étude
phénoménologique identifiant le mouvement relatif du palpeur du tribomètre du LPMT et des
différents tricots a pu être effectuée en observant l’évolution simultanée de la force de
frottement et de la position verticale du palpeur. Pour le jersey bouclette, le palpeur s’enfonce
dans la structure en couchant les bouclettes dans le sens aller du mouvement. Plus le palpeur
s’enfonce plus la force de frottement augmente car l’aire de contact entre les deux corps et la
déformation augmentent. Une fois les bouclettes complètement couchées, le palpeur ne bouge
plus et la force de frottement reste constante. Au moment du changement de sens, les
bouclettes se relèvent avant d’être couchées dans le sens opposé. Pour des pressions
relativement faibles de 3 et 6 kPa, le palpeur se relève immédiatement lors du changement de
direction. Pour des pressions plus élevées de 10 et 15 kPa, il faut un nombre suffisant de
bouclettes sous le palpeur pour qu’il se relève, c’est pourquoi le palpeur remonte plus tard
quand la pression normale augmente. De plus, au moment du changement de sens, il
s’enfonce un peu plus profondément dans la structure de la chaussette car sa vitesse décroit.
On observe les mêmes phénomènes pour le jersey ponts de fils, le déplacement vertical du
palpeur est moindre car les ponts sont moins orientés que les bouclettes. En ce qui concerne
les structures jerseys simple et double, elles semblent symétriques, la force de frottement en
sens aller est très similaire à celle en sens retour mais le déplacement vertical du palpeur
montre une anisotropie de tribologie de surface. Si l’on mesure la longueur des têtes et des
pieds de mailles dans les deux structures, on constate que les pieds de mailles sont plus courts
que les têtes de mailles. En sens aller, le palpeur passe au-dessus des têtes de mailles avant de
frotter les pieds de mailles c’est pourquoi il est entrainé vers le haut, et inversement en sens
contraire.

Une étude sur l’influence de la Longueur de Fil Absorbée et du titre des fils sur les
propriétés tribologiques des chaussettes a également été menée. Aucune influence du titre des
fils n’a été constatée tandis que pour une pression normale faible de 3 kPa, la force de
frottement évolue en sens inverse de la LFA. Cela s’explique par la forme des boucles dans le

101
tricot, pour un titre de fil donné, plus la LFA augmente, plus les boucles sont plates. La
rugosité de la surface tricotée augmente donc lorsque le LFA diminue. Pour une pression
normale importante de 15 kPa, les boucles des tricots sont écrasées et la LFA n’a aucune
influence sur la force de frottement.

Nous nous sommes intéressés à l’influence de la structure tricotée sur le frottement

induit sur la peau d’un coureur amateur lors d’un jogging. Nous avons pu observer que la
forme et l’amplitude de la courbe de force de frottement dépend fortement de la structure
considérée. Les jerseys simple et double présentent des propriétés tribologiques similaires
avec un comportement symétrique tandis que le jersey bouclette a un comportement
asymétrique du fait de l’orientation des bouclettes. Pour une distance de frottement inférieure
à 2,5 mm, le jersey bouclette induit un frottement moindre que les jerseys simple et double.
Pour une distance supérieure à 2,5 mm, le jersey bouclette produit une force de frottement
égale ou supérieure à celle induite par les autres structures tricotées. Les bouclettes des
chaussettes Labonal sont orientées de la pointe de la chaussette vers le talon lorsqu’elles sont
portées en contact avec la peau. Or, lorsque le pied entre en contact avec le sol, il glisse vers
l’avant de la chaussure donc en sens opposé de l’orientation des bouclettes ce qui va générer
un frottement élevé sur la peau. Il serait plus approprié de tricoter les chaussettes en
commençant par la jambe de manière à avoir des bouclettes orientées dans le sens le plus
préférable. Deux qualités de chaussettes bouclettes ont été étudiées, les chaussettes Labonal
présentent des bouclettes plus homogènes, plus compactes et moins faciles à écraser. Les
courbes de frottement du jersey bouclette Labonal sont différentes de celles des jerseys simple
et double tandis que le comportement des autres jerseys bouclette se rapproche de celui des
jerseys simple. Nous pouvons donc conclure que la forme particulière des courbes de
frottement du jersey bouclette Labonal est due à la rotation des bouclettes lors d’un cycle de
frottement.

Nous avons aussi étudié l’influence de la rugosité du palpeur sur les signaux de
frottement des différentes structures tricotées. Nous avons observé que pour les jerseys simple
et bouclette la force de frottement augmente avec la rugosité du palpeur. Le frottement est dû
à de la déformation et à de l’adhésion. Ici, les contributions respectives de la déformation et
de l’adhésion ne sont pas mesurées indépendamment, on ne peut donc que comparer leur

102
importance relative dans le frottement. Les phénomènes d’adhésion augmentent avec l’aire de
contact i.e. quand la rugosité du contact diminue. Vu qu’ici la force de frottement croît
lorsque la rugosité du palpeur augmente, nous en déduisons que la contribution de la
déformation dans le frottement est essentielle. En comparant les courbes de frottement
obtenues pour les jerseys simple et bouclette en contact avec le palpeur lisse en acier et le
palpeur recouvert de Lorica®, on observe une force de frottement plus importante dans le cas
du palpeur recouvert de Lorica®, cependant la différence entre les deux résultats est
importante dans le cas du jersey simple et faible dans le cas du jersey bouclette. Dans le cas
du jersey bouclette, les propriétés du palpeur sont donc moins importantes que dans le cas du
jersey simple, la part de l’adhésion dans le frottement est moindre et la contribution de la
déformation est donc plus importante.

Nous avons comparé les différentes structures de chaussettes en termes de frottement

induit sur la peau. Cependant, la chaussette, de par sa structure et son épaisseur, peut atténuer
la pression ressentie sur la peau et ainsi réduire les risques de génération de phlyctènes. Des
mesures de compression ont donc été réalisées afin de mesurer, pour une pression appliquée
donnée (même sol, même chaussure et même coureur), la pression exercée sur la peau pour
les différentes chaussettes étudiées. On observe que la pression exercée sur la peau augmente
lorsque le titre des fils augmente. Pour une même LFA, lorsque le titre des fils augmente, les
fils sont moins souples et la chaussette moins compressible. Les résultats de compression
montrent que le jersey simple présente un meilleur « amortissement statique » que le jersey
bouclette. Cependant, la pression maximale ressentie est obtenue pour une profondeur de
pénétration plus importante dans le cas du jersey bouclette que dans le cas du jersey simple,
ce qui nous a fait penser que le jersey bouclette pourrait présenter un meilleur
« amortissement dynamique » que le jersey simple. Nous avons donc testé la capacité de
dissipation de l’énergie de ces deux tricots lors d’un choc en lâchant une bille sur ces deux
chaussettes d’une hauteur donnée et en mesurant la hauteur du premier rebond de la bille. Il
s’avère que le jersey bouclette dissipe effectivement plus d’énergie que le jersey simple.

Les structures jersey double et jersey ponts de fils ont été étudiées en tant
qu’intermédiaires entre le jersey simple et le jersey bouclette. Leurs résultats de frottement
révèlent effectivement un comportement intermédiaire mais beaucoup plus proche de celui du

103
jersey simple que de celui du jersey bouclette. Quant à leurs résultats de compression, ils
montrent que ces structures sont plus dures que les deux autres et donc moins adaptées pour
des chaussettes de sport. C’est pourquoi les conclusions de cette étude ne porteront que sur les
structures jerseys simple et bouclette. L’épaisseur d’une chaussette bouclette est un moyen
simple, efficace et bon marché d’améliorer le confort de personnes dont les pieds présentent
des proéminences osseuses et des ampoules préexistantes. Les expériences de frottement, de
compression et d’absorption des chocs ont montré que :
1 Pour une force normale donnée, une chaussette en jersey bouclette induit un
frottement sur la peau moins important qu’une chaussette en jersey simple à conditions que
les bouclettes soient compactes, homogènes et orientées dans la direction de la force de
frottement.
1 Lors d’une compression progressive, pour la même énergie de compression
reçue, le jersey simple transmet sur la peau une énergie de compression moins importante que
le jersey bouclette. Le jersey simple apporte donc un meilleur « amortissement statique » ou
quasi-statique.
1 Quand on lâche une bille d’une hauteur donnée sur des chaussettes en jersey
simple et bouclette, le premier rebond de la bille est moins haut dans le cas du jersey
bouclette. Le jersey bouclette présente par conséquent un meilleur « amortissement
dynamique ».
La confrontation des résultats de frottement, de compression et d’absorption des chocs
permet de conclure que :
1 Lors de la marche, quand le pied est soumis à la fois à du frottement et à de la
compression progressive, la structure tricotée la plus indiquée est le jersey simple.
1 Lors de la course à pied, la peau est d’abord soumise à un choc puis à de la
compression progressive et du frottement. Il reste à déterminer l’influence de ces trois
sollicitations sur la génération de phlyctènes. Le jersey simple présente de meilleurs résultats
en termes de frottement et d’ « amortissement quasi-statique » tandis que le jersey bouclette
est plus performant en termes d’ « amortissement dynamique ». Dans le cas du jersey
bouclette, il est à noter que sa structure peut être améliorée significativement par des
bouclettes compactes et homogènes orientées vers la pointe du pied de manière à réduire le
frottement exercé sur la peau lorsque le pied glisse dans la chaussure.
1 Si le coureur amateur n’a que des chaussettes en jersey bouclette, il lui faut
déterminer l’orientation des bouclettes. Si les boucles sont orientées du talon vers la pointe de
la chaussette lorsque la face bouclette est en contact avec la peau, il est alors recommandé de

104
les porter en contact avec la peau. Si les boucles sont orientées en sens inverse, ce qui est le
cas de la plupart des chaussettes vendues dans le commerce, il est préférable de porter les
bouclettes en contact avec la chaussure et l’autre face en contact avec la peau ce qui permettra
de réduire le frottement sur la peau.

105
 CONTENTS

Chapter III: experimental results ____________________________________________ 100

I. FRICTION EXPERIMENTS __________________________________________ 107
1. Phenomenological study 107
2. Pressure influence 116
A. On the friction force 116
B. On the coefficient of friction 123
3. Stitch length and yarn count influence 128
4. Sock structure influence 131
5. Influence of the pin material and roughness 139
A. Experiments using the Lorica® and the smooth polyurethane pins 139
B. Experiments using the smooth and the “rough” pins made of stainless steel
146
II. COMPRESSION EXPERIMENTS _____________________________________ 149
1. Sock structure impact 149
2. Yarn count influence 152
3. Energetic approach 154
III. CONCLUSION ___________________________________________________ 155

106
 Friction and compression experiments were carried out in order to investigate the
mechanical contact between the sock and the foot skin during running.
These experiments will be successively detailed, friction tests in a first part and
compression experimentations in a second part. The influence of the applied pressure, the
yarn count, the stitch length and structure, the pin material and roughness will be investigated
within the following pages.

I. FRICTION EXPERIMENTS

12 !F75F7BE689C4$8
!F75F7BE689C4$8
When considering the evolution of both friction force and vertical position of the pin
versus time, the phenomena which occur during a friction test between the Lorica® pin and
the different socks can be identified. The vertical position of the pin is a relative position i.e.
0 corresponds to the lowest position of the pin. The laser sensor has a sensibility of 5 µm in
dynamic experiments and 1 µm in static experiments.

It can be seen in figures 1a and 1b that, for Labonal terry jersey socks under 3 and
6 kPa, the pin first sinks in the terry structure as it flattens the loops. The more it sinks the
higher the friction force due to the increase in contact area and deformation. Once the terries
are laid down, the pin stays at the same position and the friction force keeps constant. When
the direction of the movement changes, the pin flattens the terries in the opposite direction, it
thus moves upwards until all the terries are laid down. For pressures of 10 and 15 kPa,
Figures 1c and 1d show the same phenomena except that when the direction of the movement
changes, the pin does not go up immediately but sinks little more in the textile structure
before going up flattening the terries. In the reverse direction, the number of terries under the
pin has to be sufficient for the pin to go up and lay down the terries. The pin therefore moves
upwards later when normal pressure increases and it before sinks deeper in the structure
because its velocity decreases at the direction change. These explanations are corroborated by
videos.

107
 Labonal terry jersey/Lorica Labonal terry jersey/Lorica

0.2 0.18 0.4 0.1

0.15 0.16 0.3 0.09

0.08

Vertical position of the pin

Vertical position of the pin

0.1 0.14 0.2
T 0.07
Friction force (N)

Friction force (N)

0.05 0.12 0.1
0.06
0 0.1 0

(mm)

(mm)
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0.05
-0.05 0.08 -0.1
0.04
-0.1 0.06 -0.2
0.03
-0.15 0.04 -0.3 0.02
-0.2 0.02 -0.4 0.01
-0.25 0 -0.5 0
Time (s) Time (s)

a) b)

Labonal terry jersey/Lorica Labonal terry jersey/Lorica

0.6 0.1 1 0.14

0.09 0.8

Vertical position of the pin (mm)

0.4 0.12
0.08 0.6
Vertical position of the pin

0.2 0.4 0.1

0.07
Friction force (N)
Friction force (N)

0.2
0.06 0.08
0 0
(mm)

0 0.2 0.4 0.6 0.8 0.05

-0.2 0 0.2 0.4 0.6 0.8
0.06
-0.2
0.04
-0.4
-0.4 0.03 -0.6 0.04
0.02 -0.8
-0.6 0.02
0.01 -1
-0.8 0 -1.2 0
Time (s) Time (s)

i. d)

Figure 37: Friction force and vertical position of the pin for Labonal terry jersey socks in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of
respectively a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The letter T identifies the half period
where the movement of the pin follows terries orientation.

The same phenomena were observed for Labonal yarn bridges (cf. figure 2). The
bridges are indeed oriented in the same direction than the terries with a smaller angle. As
bridges are less oriented than terries, the phenomena are less pronounced, i.e. the vertical pin
displacement is smaller in the case of the Labonal yarn bridges.

108
 Labonal yarn bridges/Lorica Labonal yarn bridges/Lorica

0.2 0.016 0.4 0.018

0.3 0.016
0.15 0.014

Vertical position of the pin (mm)

Vertical position of the pin (mm)
0.1 B 0.012 0.2
0.014

Friction force (N)

Friction force (N)

0.012
0.05 0.01 0.1
0.01
0 0.008 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0.008
-0.05 0.006 -0.1
0.006
-0.1 0.004 -0.2
0.004
-0.15 0.002 -0.3 0.002

-0.2 0 -0.4 0
Time (s) Time (s)

a) b)

Labonal yarn bridges/Lorica Labonal yarn bridges/Lorica

0.6 0.035 0.8 0.02

0.6 0.018

Vertical position of the pin (mm)

0.03
Vertical position of the pin (mm)

0.4
0.016
0.4
0.025 0.014
Friction force (N)
Friction force (N)

0.2
0.2 0.012
0.02
0 0 0.01
0 0.2 0.4 0.6 0.8 0.015 0 0.2 0.4 0.6 0.8 0.008
-0.2
-0.2
0.01 0.006
-0.4
0.004
-0.4 0.005 -0.6 0.002
-0.6 0 -0.8 0
Time (s) Time (s)

c) d)

Figure 38: Friction force and vertical position of the pin for Labonal yarn bridges socks in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of
respectively a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The letter B identifies the half
period where the movement of the pin follows bridges orientation.

As for Labonal double and simple jersey socks (cf. figures 3 and 4), it seems that these
two structures also present an anisotropy of surface tribology even if these structures seem to
be symmetrical relative to an axis in the course direction. The friction force for the forward
and the reverse direction show very small difference but the laser sensor that gives the vertical
position of the pin is sensible to this anisotropy. In order to understand this anisotropy, it is
necessary to consider the different parts of a stitch (cf. figure 5). When looking at figures 6a

109
and 6b, it is not obvious that the top arcs and the bottom arcs of Labonal double and simple
jersey do not have the same length but the bottom arcs are indeed smaller than the top arcs
(cf. table 1).

Labonal double jersey/Lorica Labonal double jersey/Lorica

0.2 0.008 0.4 0.008

0.15 0.007 0.3 0.007

Vertical position of the pin (mm)

Vertical position of the pin (mm)
0.1 0.006 0.2 0.006

Friction force (N)

Friction force (N)

0.05 0.005 0.1 0.005

0 0.004 0 0.004
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
-0.05 0.003 -0.1 0.003

-0.1 0.002 -0.2 0.002

-0.15 0.001 -0.3 0.001

-0.2 0 -0.4 0
Time (s) Time (s)

a) b)

Labonal double jersey/Lorica Labonal double jersey/Lorica

0.6 0.008 0.8 0.009

0.007 0.6 0.008

Vertical position of the pin (mm)

Vertical position of the pin (mm)

0.4
0.007
0.006 0.4
Friction force (N)
Friction force (N)

0.2 0.006
0.005 0.2
0.005
0 0.004 0
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8 0.004
0.003 -0.2
-0.2 0.003
0.002 -0.4
0.002
-0.4
0.001 -0.6 0.001

-0.6 0 -0.8 0
Time (s) Time (s)

c) d)

Figure 39: Friction force and vertical position of the pin for Labonal double jersey socks in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of
respectively a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa.

110
 Labonal simple jersey/Lorica Labonal simple jersey/Lorica

0.2 0.012 0.4 0.012

0.15 0.3

Vertical position of the pin (mm)

Vertical position of the pin (mm)
0.01 0.01
0.1 0.2

Friction force (N)

Friction force (N)

0.008 0.008
0.05 0.1

0 0.006 0 0.006
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
-0.05 -0.1
0.004 0.004
-0.1 -0.2
0.002 0.002
-0.15 -0.3

-0.2 0 -0.4 0
Time (s) Time (s)

a) b)

Labonal simple jersey/Lorica Labonal simple jersey/Lorica

0.6 0.008 0.8 0.012

0.007 0.6

Vertical position of the pin (mm)

Vertical position of the pin (mm)

0.4 0.01
0.006 0.4
Friction force (N)
Friction force (N)

0.2 0.008
0.005 0.2

0 0.004 0 0.006
0 0.2 0.4 0.6 0.8 0 0.2 0.4 0.6 0.8
0.003 -0.2
-0.2 0.004
0.002 -0.4
-0.4 0.002
0.001 -0.6

-0.6 0 -0.8 0
Time (s) Time (s)

c) d)

Figure 40: Friction force and vertical position of the pin for Labonal simple jersey socks in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of
respectively a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa.

In the forward direction, the pin goes up the top arcs of the stitches and sinks in the
bottom arcs of the stitches and inversely for the reverse direction. Considering the sensibility
of the laser sensor (5 µm in dynamic), figures 3 and 4 show that the pin goes up before the
direction change and goes down once the direction has changed which can be explained by
the larger length of top arcs compared to the bottom arcs of the stitches.

111
 top arc

leg

bottom arc

Figure 41: Different parts of a stitch.

Forward direction Reverse direction

1 mm 1 mm
a) b)

Figure 42: Back of a) Labonal double jersey and b) Labonal simple jersey.

SOCK MEAN LENGTH OF THE MEAN LENGTH OF THE TOP

BOTTOM ARCS (MM) ARCS (MM)
Labonal double jersey 1.08 1.22
Labonal simple jersey 0.97 1.10

Table 13: Mean length of the bottom and the top arcs of Labonal double and simple jersey
socks.

112
 In this chapter, I present experimental results concerning the friction behaviour for
Labonal socks and for the six other socks. It has to be noticed that the terries of these socks
are oriented towards the toe of the sock i.e. in opposite way compared to Labonal terry jersey
and yarn bridges.

For terry jersey 10 - 12, 10 - 16, 10 - 20, 9 - 16 and 11 - 16, the curves given in figures
7 and 8 show the same behaviour. These terry jersey socks have different behaviours for low
and for higher pressures. For a low pressure of 3 kPa, it can be noticed in figures 7a, 7c, 7e,
8c and 8e that they behave in the same manner than Labonal terry jersey and yarn bridges
socks. The vertical movement of the pin is just inversed compared to Labonal terry jersey and
yarn bridges socks i.e. the pin first goes up because it lifts and pushes the terries to the vertical
as the terries of these socks are oriented in opposite way compared to Labonal terry jersey and
yarn bridges socks (cf. table 8 chapter II). When the sliding direction switches, the pin sinks
in the terry structure flattening the terries.

The terries and yarn bridges of Labonal socks are more elastic and less easy to swat
than the other terries. They thus have the same behaviour for low and higher pressure while,
in the case of terry jersey 10 - 12, 10 - 16, 10 - 20, 9 - 16 and 11 - 16, for a pressure of 15 kPa,
the vertical displacement of the pin is inversed compared to the case of a pressure of 3 kPa,
When sliding in the opposite direction of terries orientation, the pin does not lift and push the
terries but goes down crushing the terry structure and when the sliding direction changes, the
pin goes up sliding along the terries which came back to their initial position (figures 7b, 7d,
7f, 8d and 8f).

Simple jersey 10 - 16 (figures 8a and 8b) presents the same surface anisotropy as
Labonal double and simple jersey. The top arcs of the stitches are 6% longer than the bottom
arcs. In the forward direction, the pin goes up the top arcs of the stitches and sinks in the
bottom arcs of the stitches and inversely for the reverse direction. The pin therefore goes up
before the direction change and goes down once the direction has changed.

113
 Terry jersey 10 - 12/Lorica Terry jersey 10 - 12/Lorica

0.25 0.06 0.8 0.04

0.2 0.6 0.035

Vertical position of the pin (mm)

Vertical position of the pin (mm)
0.05
0.15 0.4
0.03
0.1

Friction force (N)

Friction force (N)

0.04 0.2
0.05 T 0
0.025

0 0.03 0 0.2 0.4 0.6 0.8 0.02

-0.2
-0.05 0 0.2 0.4 0.6 0.8
0.015
0.02 -0.4
-0.1
0.01
-0.15 -0.6
0.01
-0.2 -0.8 0.005

-0.25 0 -1 0
Time (s) Time (s)

a) b)

Terry jersey 10 - 16/Lorica Terry jersey 10 - 16/Lorica

0.25 0.04 1 0.045

0.2 0.8 0.04

0.035

Vertical position of the pin (mm)

Vertical position of the pin (mm)

0.15 0.6 0.035

0.03
0.4
Friction force (N)
Friction force (N)

0.1 0.03
0.025 0.2
0.05 0.025
0.02 0
0
-0.2 0 0.2 0.4 0.6 0.8 0.02
0 0.2 0.4 0.6 0.8 0.015
-0.05 0.015
-0.4
0.01
-0.1 -0.6 0.01

-0.15 0.005 -0.8 0.005

-0.2 0 -1 0
Time (s) Time (s)

c) d)

Terry jersey 10 - 20/Lorica Terry jersey 10 - 20/Lorica

0.25 0.025 1 0.035

0.2 0.8
0.03
Vertical position of the pin (mm)

0.15 0.02 0.6 Vertical position of the pin (mm)

0.4 0.025
Friction force (N)
Friction force (N)

0.1
0.015 0.2
0.05 0.02
0
0 0.8 0.015
0.01 -0.2 0 0.2 0.4 0.6
0 0.2 0.4 0.6 0.8
-0.05
-0.4 0.01
-0.1 0.005 -0.6
0.005
-0.15 -0.8
-0.2 0 -1 0
Time (s) Time (s)

e) f)

Figure 43: Friction force and vertical position of the pin for terry jersey socks in contact with
the Lorica® pin for a friction cycle. The mean curves were calculated from 10 experiments
for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of respectively
3 kPa for cases a, c and e and 15 kPa for cases b, d, and f. The letter T identifies the half
period where the movement of the pin follows terries orientation.

114
 Simple jersey 10 - 16/Lorica Simple jersey 10 - 16/Lorica

0.2 0.025 1 0.018

0.15 0.8 0.016

Vertical position of the pin (mm)

Vertical position of the pin (mm)

0.02 0.6 0.014
0.1
0.4

Friction force (N)

Friction force (N)

0.012
0.05
0.015 0.2
0.01
0 0
0 0.2 0.4 0.6 0.8 0.01
-0.2 0 0.2 0.4 0.6 0.8 0.008
-0.05
0.006
-0.4
-0.1
0.005 -0.6 0.004
-0.15 0.002
-0.8
-0.2 0 -1 0
Time (s) Time (s)

a) b)

Terry jersey 9 - 16/Lorica Terry jersey 9 - 16/Lorica

0.25 0.05 1 0.08

0.2 0.045 0.8 0.07

Vertical position of the pin (mm)

Vertical position of the pin (mm)

0.15 0.04 0.6
0.06
T 0.035 0.4
Friction force (N)

Friction force (N)

0.1
0.2 0.05
0.03
0.05
0.025 0 0.04
0 0 0.2 0.4 0.6 0.8
0.02 -0.2 0.03
0 0.2 0.4 0.6 0.8
-0.05
0.015 -0.4
0.02
-0.1 0.01 -0.6
-0.15 -0.8 0.01
0.005
-0.2 0 -1 0
Time (s) Time (s)

c) d)

Terry jersey 11 - 16/Lorica Terry jersey 11 - 16/Lorica

0.25 0.03 1 0.04

0.2 0.8 0.035

Vertical position of the pin (mm)

0.025 Vertical position of the pin (mm)

0.15 0.6
0.03
0.4
Friction force (N)

Friction force (N)

0.1 0.02
0.2 0.025
0.05
0.015 0 0.02
0 0 0.2 0.4 0.6 0.8
-0.2 0.015
0 0.2 0.4 0.6 0.8
-0.05 0.01
-0.4
0.01
-0.1 -0.6
0.005
-0.15 -0.8 0.005

-0.2 0 -1 0
Time (s) Time (s)

e) f)

Figure 44 : Friction force and vertical position of the pin for terry and simple jersey socks in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and a pressure of
respectively 3 kPa for cases a, c and e and 15 kPa for cases b, d, and f. The letter T identifies
the half period where the movement of the pin follows terries orientation.

115
 2 !DF994DF8B74F7EF88
!DF994DF8B74F7EF88

12 &9F79ABDEFDC9ACBE79

Labonal terry jersey/Lorica Labonal double jersey/Lorica

0.8 0.8
3kPa 3kPa
6kPa 6kPa
0.6 10kPa 10kPa
15kPa 0.6 15kPa

0.4
0.4

0.2

Friction force (N)

Friction force (N)

0.2
0

−0.2
T 0

−0.2
−0.4

−0.4
−0.6

−0.8 −0.6
10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz
−1 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

Labonal simple jersey/Lorica Labonal yarn bridges/Lorica

0.8 0.8
3kPa 3kPa
6kPa 6kPa
10kPa 10kPa
0.6 15kPa 0.6 15kPa

0.4 0.4
Friction force (N)

Friction force (N)

0.2 0.2

0 0
B
−0.2 −0.2

−0.4 −0.4

−0.6 −0.6
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)

Figure 45: Friction force for a) Labonal terry jersey, b) Labonal double jersey, c) Labonal
simple jersey and d) Labonal yarn bridges socks in contact with the Lorica® pin for a friction
cycle. The mean curves were calculated from 10 experiments for a friction distance of 5 mm,
a friction frequency of 1.5 Hz and pressures from 3 to 15 kPa on LPMT’s tribometer. The
letters T and B identify the half period where the movement of the pin follows respectively
terries orientation and bridges orientation.

116
 1

Mean friction force (N)

0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 2 4 6 8 10 12 14 16

Apparent pressure (kPa)

Labonal simple jersey Labonal terry jersey

a)

0
0 0.5 1 1.5 2 2.5 3
-0.5 y = 0.9x - 2.6
-1 R2 = 0.99
nl(FF)

-1.5 y = 0.9x - 2.8

R2 = 0.99
-2

-2.5
nl(AP)

Labonal terry jersey Labonal simple jersey

b)

Figure 46: a) Mean friction force versus apparent pressure and b) fitting of the mean friction
force in function of the apparent pressure using the empirical equation FF = K × AP n for
Labonal terry and simple jersey socks. The abbreviation nl stands for the logarithm, FF is the
friction force, AP is the apparent pressure and K and n are empirical constants. The mean
friction force was calculated from 3 experimental friction curves of one friction cycle and
corresponds to the mean value of the plateaus of the curves.

The well-known fact that the friction force gets higher when the pressure increases can
be observed on figures 9 to 13. The mean friction force versus the apparent pressure is shown
for Labonal terry jersey and simple jersey socks in figure 10a. The fitting of the mean friction

117
force using the empirical equation FF = K × AP n gives n = 0.9 for both Labonal terry jersey
2
and simple jersey socks which is conform to the theory i.e. ≤ n ≤ 1.
3

Friction force (N)

3 Friction force (N)
3kPa 4
6kPa 3kPa
10kPa
2 15kPa
6kPa
3
10kPa
15kPa
1 2

0 1
T
−1 0

−2 −1

−2
−3

−3
−4
6 samples, 5mm, 1.5Hz, Labonal double jersey on Lorica
5 samples, Labonal terry jersey, 5mm, 1.5Hz, sock on Lorica −4
−5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Time (s)
Time (s)

a) b)
Friction force (N) Friction force (N)
4 4
3kPa 3kPa
6kPa 6kPa
10kPa 10kPa
3 15kPa 3 15kPa

2 2

1 1

0 0

−1 −1
B
−2 −2

−3 −3
3 samples, 5mm, 1.5Hz, Labonal simple jersey on Lorica
3 samples, 5mm, 1.5Hz, Labonal yarn bridges on Lorica
−4 −4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 47: Friction force for a) Labonal terry jersey, b) Labonal double jersey, c) Labonal
simple jersey and d) Labonal yarn bridges socks in contact with Lorica® for a friction cycle.
The mean curves were calculated from at least 3 experiments for a friction distance of 5 mm,
a friction frequency of 1.5 Hz and pressures from 3 to 15 kPa on the TFA. The letters T and B
identify the half period where the movement of the pin follows respectively terries orientation
and bridges orientation.

118
 Terry jersey 10−12/Lorica Terry jersey 10−16/Loric a
1 0.8
3kPa 3kPa
15kPa 15kPa
0.8 0.6

0.6
0.4

0.4
Friction force (N)

Friction force (N)

0.2

0.2 T
0
0

−0.2
−0.2

−0.4
−0.4

−0.6 −0.6
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

Terry jersey 10−20/Lorica

Simple jersey 10−16/Lorica
0.8 1
3kPa 3kPa
15kPa 15kPa
0.6 0.8

0.6
0.4
0.4
Friction force (N)

0.2
Friction force (N)

0.2
0
0

−0.2 −0.2

−0.4 −0.4

−0.6
−0.6
10 samples, 5mm, 1.5Hz
−0.8
10 samples, 5mm, 1.5Hz
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 −1
Time (s) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s)

c) d)

Terry jersey 9−16/Lorica Terry jersey 11−16/Lorica

0.8 0.8
3kPa 3kPa
15kPa
15kPa
0.6 0.6

0.4 0.4
Friction force (N)

0.2
Friction force (N)

0.2

0
0

−0.2
−0.2

−0.4
−0.4

−0.6
10 samples, 5mm, 1.5Hz −0.6

−0.8 10 samples, 5mm, 1.5Hz

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 −0.8
Time (s) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s)

e) f)

Figure 48: Friction force for a) terry jersey 10 – 12, b) terry jersey 10 – 16, c) terry jersey
10 – 20, d) simple jersey 10 – 16, e) terry jersey 9 – 16 and f) terry jersey 11 – 16 in contact
with the Lorica® pin for a friction cycle. The mean curves were calculated from 10

119
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and pressures of 3
and 15 kPa on LPMT’s tribometer. The letter T identifies the half period where the movement
of the pin follows terries orientation.

For a fixed sock and a given friction device, the shape of the friction curve is the same
for all the pressures. The lower the friction force, the flatter the curve. It can be noticed on
figure 11 that friction tests carried out using the TFA present two static peaks which
correspond to the change of the sliding direction. These static peaks do not occur for friction
experiments on the LPMT’s tribometer as the pin is free to move up and down while the TFA
pin is kept to the same vertical position by the elevation arm. TFA curves show less variations
as the friction force signal is filtered while the signal of LPMT’s tribometer is not filtered.

The values of the friction force related to the apparent contact surface are of the same
order of magnitude for both LPMT’s tribometer and TFA for Labonal socks (cf. tables 2
and 3). For terry jersey socks, when the sliding direction corresponds to terries orientation
(forward sliding direction), we can suppose that the deformation component of the friction
force due to the rotation of terries may induce more friction than the crushing of the structure
which explains the higher friction force values in the case of the LPMT’s tribometer (cf.
table 2). On the other hand , when the sliding direction is opposed to the terries orientation,
terries withstand the pin movement, and the phenomena differ upon the tribometer which is
considered. For the LPMT’s tribometer, the terries under and at the front of the pin are carried
along by the pin and rotate. As the pin is free to move vertically, it is lifted by the rotating
terries and does not exert extra pressure on the terries. On the other hand, for the TFA, the pin
cannot move vertically and is opposed to the natural movement of terries leading to an
additional pressure and a higher friction force (cf. table 3). For the others structures, the
values of the friction force related to the apparent contact surface are comparable for both
tribometers or slightly higher for TFA which may be due to the fact that the pressure can get
higher in the case of the TFA measurements. The effect of the hairiness orientation of
Labonal terry jersey and yarn bridges is visible in the friction force and the vertical
displacement of the pin measured on LPMT’s tribometer but is more pronounced in TFA
measurement data.

120
 LABONAL LABONAL LABONAL LABONAL
PRESSURE TERRY JERSEY DOUBLE SIMPLE YARN
(KPA) JERSEY JERSEY BRIDGES
LPMT TFA LPMT TFA LPMT TFA LPMT TFA
3 0.098 0.049 0.111 0.097 0.111 0.114 0.111 0.097
6 0.144 0.114 0.196 0.211 0.196 0.227 0.196 0.195
10 0.261 0.211 0.327 0.357 0.314 0.357 0.294 0.308
15 0.392 0.292 0.458 0.519 0.458 0.487 0.425 0.455

Table 14: Comparison of the LPMT’s tribometer and the TFA. Labonal socks friction force
in N for 1 cm² of apparent contact area for pressures from 3 to 15 kPa for the forward sliding
direction i.e. the first half of a friction cycle. No standard deviation is given here as these data
have been calculated from mean curves.

LABONAL LABONAL LABONAL LABONAL

PRESSURE TERRY JERSEY DOUBLE SIMPLE YARN
(KPA) JERSEY JERSEY BRIDGES
LPMT TFA LPMT TFA LPMT TFA LPMT TFA
3 0.111 0.146 0.111 0.114 0.098 0.114 0.111 0.146
6 0.229 0.260 0.196 0.211 0.196 0.179 0.209 0.260
10 0.392 0.422 0.327 0.325 0.327 0.292 0.307 0.373
15 0.523 0.617 0.458 0.487 0.438 0.438 0.438 0.519

Table 15: Comparison of the LPMT’s tribometer and the TFA. Labonal socks friction force in
N for 1 cm² of apparent contact area for pressures from 3 to 15 kPa for the back sliding
direction i.e. the second half of a friction cycle. No standard deviation is given here as these
data have been calculated from mean curves.

Friction experiments were also carried out changing the counterpart (the material the
socks are rubbing). The influence of the counterpart material and roughness will be discussed
later. At this stage we just check that the pressure influence is the same whatever the

121
counterpart material. It can thus be noticed on figure 13 that, for Labonal socks in contact
with smooth polyurethane, the friction force increases with pressure.

Friction force (N)

8 Friction force (N)
3kPa 6
6kPa
10kPa 3kPa
6 15kPa 6kPa
10kPa
4
15kPa
4

2
2

0 0

−2
−2

−4
−4
−6

3 samples, 5mm, 1.5Hz, Labonal double jersey on smooth polyurethane 3 samples, 5mm, 1.5Hz, Labonal simple jersey on smooth polyurethane
−6
−8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Time (s)
Time (s)

a) b)
Friction force (N)
6
3kPa
6kPa
10kPa
4 15kPa

−2
B
−4

−6

3 samples, 5mm, 1.5Hz, Labonal yarn bridges on smooth polyurethane

−8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)

c)

Figure 49: Friction force for a) Labonal double jersey, b) Labonal simple jersey and
c) Labonal yarn bridges socks in contact with smooth polyurethane for a friction cycle. The
mean curves were calculated from 3 experiments for a friction distance of 5 mm, a friction
frequency of 1.5 Hz and pressures from 3 to 15 kPa on the TFA. The letter B identifies the
half period where the movement of the pin follows bridges orientation.

122
 2 &9F79EC7AADED7F9CA9ABDEFDC9

Labonal terry jersey Labonal double jersey

0.4 0.4
3kPa 3kPa
6kPa 6kPa
0.3 10kPa 0.3 10kPa
15kPa 15kPa

0.2 0.2

Coefficient of Friction
Coefficient of Friction

0.1 0.1

0
T 0

−0.1 −0.1

−0.2 −0.2

−0.3 −0.3

−0.4 −0.4
10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz
−0.5 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

Labonal simple jersey Labonal yarn bridges

0.4 0.4
3kPa 3kPa
6kPa
6kPa 10kPa
0.3 0.3
10kPa 15kPa
15kPa
0.2 0.2
B
Coefficient of Friction

Coefficient of Friction

0.1 0.1

0
0

−0.1
−0.1

−0.2
−0.2

−0.3
−0.3
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 −0.4
Time (s) 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s)

c) d)

Figure 50: Coefficient of friction for a) Labonal terry jersey, b) Labonal double jersey,
c) Labonal simple jersey and d) Labonal yarn bridges socks in contact with the Lorica® pin
for a friction cycle. The mean curves were calculated from 10 experiments for a friction
distance of 5 mm, a friction frequency of 1.5 Hz and pressures from 3 to 15 kPa on the
LPMT’s tribometer. The letters T and B identify the half period where the movement of the
pin follows respectively terries orientation and bridges orientation.

123
 The mean coefficient of friction was obtained by dividing the mean friction force by
the normal force. It can be seen in figures 14 to 17 that the mean coefficient of friction either
remains stable or slightly decreases when pressure increases for the couple cotton / Lorica®.
The mean coefficient of friction versus the apparent pressure is detailed for Labonal terry
jersey and simple jersey socks respectively in figures 15a and 15b.

0,5
Mean coefficient of friction

0,45

0,4

0,35

0,3

0,25
0 2 4 6 8 10 12 14 16

Apparent pressure (kPa)

a)

0,35
Mean coefficient of friction

0,3

0,25
0 2 4 6 8 10 12 14 16

Apparent pressure (kPa)

b)
Figure 51: Mean coefficient of friction versus apparent pressure for a) Labonal terry and
b) simple jersey socks. The mean coefficient of friction was calculated from 3 experimental
curves of 1 friction cycle on the LPMT’s tribometer and corresponds to the mean value of the
plateaus of the curves.

124
 Coefficient Of Friction Coefficient Of Friction
0.3 0.4
3kPa
0.2 6kPa 3kPa
10kPa 0.3
6kPa
15kPa
0.1 10kPa
T 0.2
15kPa
0
0.1
−0.1 5 samples, 5mm, 1.5Hz, Labonal double jersey on Lorica
0
−0.2
−0.1
−0.3
−0.2
−0.4

−0.5 −0.3
5 samples, 5mm, 1.5Hz, Labonal terry jersey on Lorica
−0.6 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Coefficient Of Friction Coefficient Of Friction

0.4 0.5
3kPa
0.3 3kPa 0.4 6kPa
6kPa 10kPa
0.2 10kPa 0.3 15kPa

0.1
B 15kPa
0.2
0.1
0
3 samples, 5mm, 1.5Hz, Labonal yarn bridges on Lorica
0
−0.1
−0.1
−0.2
−0.2
−0.3 −0.3
−0.4 −0.4
3 samples, 5mm, 1.5Hz, Labonal simple jersey on Lorica
−0.5 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 52: Coefficient of friction for a) Labonal terry jersey, b) Labonal double jersey,
c) Labonal yarn bridges and d) Labonal simple jersey socks in contact with the Lorica® pin
for a friction cycle. The mean curves were calculated from at least 3 experiments for a
friction distance of 5 mm, a friction frequency of 1.5 Hz and pressures from 3 to 15 kPa on
the TFA. The letters T and B identify the half period where the movement of the pin follows
respectively terries orientation and bridges orientation.

125
 Terry jersey 10−12 / Lorica Terry jersey 10−16 / Lorica
0.5 0.5
3kPa 3kPa
15kPa 15kPa
0.4 0.4

0.3 0.3
Coefficient of Friction

0.2

Coefficient of Friction
0.2

0.1 0.1
T
0 0

−0.1 −0.1

−0.2 −0.2

−0.3 −0.3
10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz
−0.4 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)
Terry jersey 10−20 / Lorica Simple jersey 10−16 / Lorica
0.5 0.4
3kPa 3kPa
15kPa 15kPa
0.4 0.3

0.3 0.2

0.2
Coefficient of Friction
Coefficient of Friction

0.1

0.1 0

0 −0.1

−0.1 −0.2

−0.2 −0.3

−0.3 −0.4
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0.4 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)
Terry jersey 9−16 / Lorica Terry jersey 11−16 / Lorica
0.5 0.4
3kPa 3kPa
15kPa 15kPa
0.4
0.3

0.3
0.2
Coefficient of Friction

0.2
Coefficient of Friction

0.1
0.1
0
0

−0.1
−0.1

−0.2
−0.2

−0.3 −0.3
10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz
−0.4 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

e) f)

Figure 53: Coefficient of friction for a) terry jersey 10 – 12, b) terry jersey 10 – 16, c) terry
jersey 10 – 20, d) simple jersey 10 – 16, e) terry jersey 9 – 16 and f) terry jersey 11 – 16 in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10

126
experiments for a friction distance of 5 mm, a friction frequency of 1.5 Hz and pressures of 3
and 15 kPa on the LPMT’s tribometer. The letter T identifies the half period where the
movement of the pin follows terries orientation.

Coefficient Of Friction Coefficient Of Friction

0.8 0.8
3kPa
6kPa 3 samples, 5mm, 1.5Hz, Labonal simple jersey on smooth polyurethane
10kPa
0.6 15kPa 0.6 3kPa
6kPa
10kPa
0.4 15kPa
0.4

0.2
0.2

0
0

−0.2
−0.2
−0.4

−0.4
−0.6

3 samples, 5mm, 1.5Hz, Labonal double jersey on smooth polyurethane −0.6

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Time (s)
Time (s)

a) b)

Coefficient Of Friction
0.8
3kPa
6kPa
0.6
10kPa
15kPa
0.4

0.2

0
B
−0.2

−0.4

−0.6
3 samples, 5mm, 1.5Hz, Labonal yarn bridges on smooth polyurethane
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)

c)

Figure 54: Coefficient of friction for a) Labonal double jersey, b) Labonal simple jersey and
c) Labonal yarn bridges socks in contact with smooth polyurethane for a friction cycle. The
mean curves were calculated from 3 experiments for a friction distance of 5 mm, a friction
frequency of 1.5 Hz and pressures from 3 to 15 kPa on the TFA. The letter B identifies the
half period where the movement of the pin follows bridges orientation.

127
 The evolution of the coefficient of friction in function of the pressure depends on the
materials in contact. The couple cotton / smooth polyurethane does not react like the couple
cotton / Lorica® (cf. figures 16 and 18). The coefficient of friction indeed increases when
pressure gets higher as it is shown in figure 18. Friction includes a part of deformation and a
part of adhesion. When the normal load is increased, both the deformation part and the
adhesion part of friction get higher but the rise in deformation is less important compared to
the increase of the real contact area and the decrease of the distance between the two
contacting bodies. Figures 16 and 18 thus indicate that the deformation part of friction is more
important for the couple cotton / Lorica® than for the couple cotton / smooth polyurethane
and that the adhesion contribution to friction is higher for the couple cotton / smooth
polyurethane than for the couple cotton / Lorica®.

2 CBCE8F7C8
CBCE8F7C8678$6D78E47C8B74F7EF
678$6D78E47C8B74F7EF8
B74F7EF8
Three terry jersey socks that only differ by their stitch length namely terry jersey
9 - 16, terry jersey 10 - 16, and terry jersey 11 - 16 and three terry jersey socks that only differ
by their yarn count namely terry jersey 10 - 12, terry jersey 10 - 16, and terry jersey 10 - 20
were at our disposal. We could therefore investigate the influence of the stitch length and the
yarn count on the friction force and the coefficient of friction (cf. chapter 2 paragraph II.2).

Figures 19 and 20 respectively show the friction force and the coefficient of friction
for the socks that have just been cited. By these figures, the yarn count does not impact the
friction results. On the other hand, when the fabrics are loaded with a low pressure of 3 kPa,
the values of the friction force and the coefficient of friction increase when the stitch length
decreases. This is due to the shape of the loop in the knitted fabric. For the same yarn count,
the higher the stitch length, the flatter the loop. The fabric surface roughness is thus lower for
terry jersey 11 – 16 than for terry jersey 9 – 16 and the values of the friction force and the
coefficient of friction get higher with surface roughness. For a pressure of 15 kPa, the loops
are compressed and the stitch length has no influence on friction.

128
 3kPa − Sock on Lorica 15kPa − Sock on Lorica
0.25 0.8
Terry jersey 9−16 Terry jersey 9−16
Terry jersey 10−16 Terry jersey 10−16
0.2 Terry jersey 11−16 Terry jersey 11−16
0.6

0.15
0.4

0.1
Friction force (N)

0.2

Friction force (N)

0.05
0
0 T
−0.2
−0.05

−0.4
−0.1

10 samples, 5mm, 1.5Hz −0.6 10 samples, 5mm, 1.5Hz

−0.15

−0.2 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

3kPa − Sock on Lorica 15kPa − Sock on Lorica

0.25 1
Terry jersey 10−12 Terry jersey 10−12
Terry jersey 10−16 Terry jersey 10−16
0.2 Terry jersey 10−20 0.8 Terry jersey 10−20

0.15 0.6

0.1 0.4
Friction force (N)

Friction force (N)

0.05 0.2

0 0

−0.05 −0.2

−0.1 −0.4

−0.15 10 samples, 5mm, 1.5Hz −0.6 10 samples, 5mm, 1.5Hz

−0.2 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)

Figure 55: Friction force for a) terry jersey 9 – 16, terry jersey 10 – 16 and terry jersey
11 – 16 under 3 kPa, b) terry jersey 9 – 16, terry jersey 10 – 16 and terry jersey 11 – 16
under 15 kPa, c) terry jersey 10 – 12, terry jersey 10 – 16 and terry jersey 10 – 20 under 3
kPa and d) terry jersey 10 – 12, terry jersey 10 – 16 and terry jersey 10 – 20 under 15 kPa
in contact with the Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the LPMT’s
tribometer. The letter T identifies the half period where the movement of the pin follows
terries orientation.

129
 3kPa − Sock on Lorica 15kPa − Sock on Lorica
0.5 0.4
Terry jersey 9−16 Terry jersey 9−16
Terry jersey 10−16 Terry jersey 10−16
0.4 Terry jersey 11−16 Terry jersey 11−16
0.3

0.3
0.2

0.2
Coefficient of Friction

Coefficient of Friction
0.1
0.1

0 T 0

−0.1
−0.1

−0.2
−0.2

10 samples, 5mm, 1.5Hz −0.3 10 samples, 5mm, 1.5Hz

−0.3

−0.4 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

3kPa − Sock on Lorica 15kPa − Sock on Lorica

0.5 0.4
Terry jersey 10−12 Terry jersey 10−12
Terry jersey 10−16 Terry jersey 10−16
0.4 Terry jersey 10−20 Terry jersey 10−20
0.3

0.3
0.2

0.2
Coefficient of Friction
Coefficient of Friction

0.1
0.1
0
0

−0.1
−0.1

−0.2
−0. 2

10 samples, 5mm, 1.5Hz −0.3 10 samples, 5mm, 1.5Hz

−0.3

−0.4 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)

Figure 56: Coefficient of friction for a) terry jersey 9 – 16, terry jersey 10 – 16 and terry
jersey 11 – 16 under 3 kPa, b) terry jersey 9 – 16, terry jersey 10 – 16 and terry jersey
11 – 16 under 15 kPa, c) terry jersey 10 – 12, terry jersey 10 – 16 and terry jersey 10 – 20
under 3 kPa and d) terry jersey 10 – 12, terry jersey 10 – 16 and terry jersey 10 – 20 under
15 kPa in contact with the Lorica® pin for a friction cycle. The mean curves were calculated
from 10 experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the
LPMT’s tribometer. The letter T identifies the half period where the movement of the pin
follows terries orientation.

130
 2 EA89CD4EC4DF8
EA89CD4EC4DF8B74F7EF88
B74F7EF88
We aimed to determine the influence of the knitted structure of the sock on the friction
generated on the skin of the foot during running. We found that the shape and the value of the
friction force and the coefficient of friction curves depend on the sock structure (cf. figures 21
to 27).
Figures 21a to 21d and 22a to 22d show that Labonal double and simple jerseys have
similar friction forces and coefficients of friction when rubbing against Lorica®. Their
friction signals increase sharply before getting stable during the first half of the friction cycle
and when the sliding direction of the pin changes, the values of the friction force and the
coefficient of friction decrease sharply before stabilizing. Their curves are symmetrical.

On the other hand, for the Labonal terry jersey sock, the values of the friction force
and the coefficient of friction progressively increase without getting stable. For sliding
distances lower than 2.5 mm, Labonal terry jersey socks induce less friction than the Labonal
simple and double jersey structures while beyond 2.5 mm sliding distance, Labonal terry
jersey socks present similar or slightly higher friction than other Labonal structures. It can
also be noticed that the friction signals of Labonal terry jersey are asymmetrical. The friction
force and the coefficient of friction are indeed higher in the second half of the friction cycle
i.e. when the pin rubs the sock in the direction opposed to the terries orientation.

It has to be noted that for Labonal terry jersey socks, the terries are oriented towards
the heel and when the foot contacts the ground, it slides within the shoe towards the toe i.e. it
rubs the sock in the direction opposed to the terries orientation which is the less suitable case
regarding skin irritations as it induces higher friction on the skin. It is thus recommended to
knit the socks from the leg to the toe in order to reduce the friction when the foot contacts the
ground and is loaded with significant normal and tangential forces. As mentioned above,
Labonal and others socks are not from the same line of products. Labonal products are higher
quality socks. The terries of the socks that were not supplied by Labonal are easy to crush i.e.
they do not rotate during a friction cycle. Figures 21e, 21f, 22e and 22f show that one of these
terry jersey structure has similar friction force and coefficient of friction than a simple jersey
structure. It may thus be concluded that the particular shape of Labonal terry jersey friction
force is due to the rotation of terries during a friction cycle.

131
 3kPa−sock on Lorica 6kPa−sock on Lorica
0.2 0.4
Labonal terry jersey Labonal terry jersey
Labonal double jersey Labonal double jersey
0.15 Labonal simple jersey Labonal simple jersey
Labonal yarn bridges 0.3 Labonal yarn bridges

0.1
0.2

0.05
0.1
Friction force (N)

Friction force (N)

0 T B
0
−0.05

−0.1
−0.1

−0.2
−0.15
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0. 2 −0.3

−0.25 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)

10 kPa − Sock on Lorica 15 kPa − Sock on Lorica

0.6 0.8
Labonal terry jersey Labonal terry jersey
Labonal double jersey 0.6 Labonal double jersey
0.4 Labonal simple jersey Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges
0.4
0.2
0.2
Friction force (N)
Friction force (N)

0 0

−0.2 −0.2

−0.4
−0.4
−0.6

−0.6
−0.8
10 samples, 5mm, 1.5Hz
10 samples, 5mm, 1.5Hz
−0.8 −1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)

3kPa − Sock on Lorica

0.25 15kPa − Sock on Lorica
1
Terry jersey 10−16 Terry jersey 10−16
0.2 Simple jersey 10−16 Simple jersey 10−16
0.8
0.15
0.6
0.1
0.4
Friction force (N)

0.05
Friction force (N)

0.2
0
0
−0.05 T −0.2
−0.1
−0.4
−0.15
−0.6
10 samples, 5mm, 1.5Hz
−0.2
10 samples, 5mm, 1.5Hz −0.8
−0.25
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 −1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s)
Time (s)

e) f)

Figure 57: Friction force for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa, d) 15 kPa
and for terry jersey 10 – 16 and simple jersey 10 – 16 under e) 3 kPa and f) 15 kPa in
contact with the Lorica® pin for a friction cycle. The mean curves were calculated from

132
10 experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the
LPMT’s tribometer. The letters T and B identify the half period where the movement of the
pin follows respectively terries orientation and bridges orientation.

Yarn bridges structure behaves as an intermediate between terry jersey and simple or
double jersey; the bridges are indeed slightly oriented. The asymmetry of Labonal terry jersey
and yarn bridges friction curves must be due to the orientation of the terries and bridges; in
the terries or bridges direction, the structure rotate quite easily while in the opposite direction,
the structure resists to the movement.

Labonal socks friction experiments on the TFA are reported in figures 23 and 24. It
can be seen that for friction tests on the TFA, the shape of the friction force and the
coefficient of friction curves does not depend on the sock structure; all the curves have the
same shape. As the distance between the pin and the sock is constant, the terries and bridges
cannot rotate during a friction cycle and only the values of the friction force and the
coefficient of friction differ from one sock to another.

133
 3kPa − Sock on Lorica 6kPa − Sock on Lorica
0.4 0.4
Labonal terry jersey Labonal terry jersey
Labonal double jersey Labonal double jersey
0.3 Labonal simple jersey 0.3 Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges

0.2 0.2

0.1 0.1
T B
Coefficient of Friction

Coefficient of Friction
0 0

−0.1 −0.1

−0.2 −0. 2

−0.3 −0.3

10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz

−0.4 −0.4

−0.5 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

a) b)
10kPa − Sock on Lorica 15kPa − Sock on Lorica
0.4 0.4
Labonal terry jersey Labonal terry jersey
Labonal double jersey Labonal double jersey
0.3 Labonal simple jersey 0.3 Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges

0.2 0.2

0.1 0.1
Coefficient of Friction

Coefficient of Friction

0 0

−0.1 −0.1

−0.2 −0.2

−0.3 −0.3

10 samples, 5mm, 1.5Hz 10 samples, 5mm, 1.5Hz

−0.4 −0.4

−0.5 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

c) d)

3kPa − Sock on Lorica 15kPa − Sock on Lorica

0.5 0.4
Terry jersey 10−16 Terry jersey 10−16
Simple jersey 10−16 Simple jersey 10−16
0.4
0.3

0.3
0.2
0.2
Coefficient of Friction
Coefficient of Friction

0.1
0.1

0 0

−0.1 T −0.1

−0.2
−0.2
10 samples, 5mm, 1.5Hz
−0. 3
10 samples, 5mm, 1.5Hz
−0.3
−0.4

−0.5 −0.4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Time (s) Time (s)

e) f)

Figure 58: Coefficient of friction for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa,
d) 15 kPa and for terry jersey 10 – 16 and simple jersey 10 – 16 under e) 3 kPa and f) 15 kPa
in contact with Lorica® pin for a friction cycle. The mean curves were calculated from 10
experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the LPMT’s
134
tribometer. The letters T and B identify the half period where the movement of the pin follows
respectively terries orientation and bridges orientation.

Friction force (N) Friction force (N)

1 2
3 samples, 5mm, 1.5Hz, 3 kPa, socks on Lorica Labonal terry jersey
0.8 Labonal double jersey
Labonal terry jersey 1.5 Labonal simple jersey
0.6 Labonal double jersey Labonal yarn bridges
Labonal simple jersey 1
0.4 Labonal yarn bridges
0.5
0.2

0 0
T B
−0.2 −0.5
−0.4
−1
−0.6
−1.5
−0.8
3 samples, 5mm, 1.5Hz, 6 kPa, socks on Lorica
−1 −2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Friction force (N) Friction force (N)

3 4
3 samples, 5mm, 1.5Hz, 10 kPa, socks on Lorica Labonal terry jersey
Labonal terry jersey 3 Labonal double jersey
Labonal double jersey Labonal simple jersey
2
Labonal simple jersey Labonal yarn bridges
2
Labonal yarn bridges
1 1

0
0
−1

−1 −2

−3
−2
−4
3 samples, 5mm, 1.5Hz, 15 kPa, socks on Lorica
−3 −5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 59: Friction force for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa and
d) 15 kPa in contact with Lorica® for a friction cycle. The mean curves were calculated from
3 experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the TFA.
The letters T and B identify the half period where the movement of the pin follows respectively
terries orientation and bridges orientation.

135
 Coefficient Of Friction Coefficient Of Friction
0.6 0.5
Labonal terry jersey Labonal terry jersey
Labonal double jersey 0.4 Labonal double jersey
0.4 Labonal simple jersey Labonal simple jersey
Labonal yarn bridges 0.3 Labonal yarn bridges

0.2 0.2

0.1
0
T B 0
−0.2
−0.1

−0.4 −0.2

−0.3
−0.6
−0.4
3 samples, 5mm, 1.5Hz, 3kPa, sock on Lorica
3 samples, 5mm, 1.5Hz, 6kPa, sock on Lorica
−0.8 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s )

a) b)

Coefficient Of Friction Coefficient Of Friction

0.4 0.4
Labonal terry jersey Labonal terry jersey
0.3 Labonal double jersey 0.3 Labonal double jersey
Labonal simple jersey Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges
0.2 0.2

0.1 0.1

0 0

−0.1 −0.1

−0.2 −0.2

−0.3 −0.3

−0.4 −0.4
3 samples, 5mm, 1.5Hz, 10kPa, sock on Lorica 3 samples, 5mm, 1.5Hz, 15kPa, sock on Lorica
−0.5 −0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 60: Coefficient of friction for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa and
d) 15 kPa in contact with Lorica® for a friction cycle. The mean curves were calculated from
3 experiments for a friction distance of 5 mm and a friction frequency of 1.5 Hz on the TFA.
The letters T and B identify the half period where the movement of the pin follows respectively
terries orientation and bridges orientation.

When the sliding direction corresponds to the terries and bridges orientation, Labonal
terry jersey is the sock structure which induces less friction, followed by Labonal yarn bridges
and simple and double jerseys. For the opposite sliding direction, we find, from the lowest to
the highest friction: simple jersey, double jersey, yarn bridges and terry jersey structures. The
curves for Labonal simple and double jersey are almost identical.
As it has already been outlined for friction experiments on LPMT’s tribometer, the
friction curves of simple and double jersey structures are symmetrical while terry jersey and
yarn bridges socks curves are asymmetrical because of their oriented structure. The

136
asymmetry is more important in the case of terry jersey because terries are more numerous
and longer, therefore they are more oriented than bridges. Moreover, the asymmetry is also
more important when the friction test is carried out on the TFA; for terry jersey, the friction
force in the back direction is about 2 times the friction force in the forward direction for
experiments on the TFA (cf. figure 23) while the friction force in the back direction is about
1.5 times the friction force in the forward direction for LPMT’s tribometer tests (cf.
figure 21). This is due to the way the pin is loaded in both devices.

Friction force (N) Friction force (N)

1 2.5
Labonal double jersey Labonal double jersey
2 Labonal simple jersey
Labonal simple jersey Labonal yarn bridges
0.5 1.5
Labonal yarn bridges
B 1

0 0.5

−0.5 −0.5

−1

−1.5
−1
−2
3 samples, 5mm, 1.5Hz, 3kPa, socks on smooth polyurethane
3 samples, 5mm, 1.5Hz, 6kPa, socks on smooth polyurethane
−2.5
−1.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Friction force (N) Friction force (N)

5 8
3 samples, 5mm, 1.5Hz, 10kPa, socks on smooth polyurethane Labonal double jersey
4 Labonal double jersey 6
Labonal simple jersey
Labonal simple jersey
3 Labonal yarn bridges Labonal yarn bridges
4
2
2
1
0
0
−2
−1

−2 −4

−3 −6

−4 3 samples, 5mm, 1.5Hz, 15kPa, socks on smooth polyurethane

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 61: Friction force for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa and
d) 15 kPa in contact with smooth polyurethane for a friction cycle. The mean curves were
calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of

137
1.5 Hz on the TFA. The letter B identifies the half period where the movement of the pin
follows bridges orientation.

Coefficient Of Friction Coefficient Of Friction

0.6 0.6
Labonal double jersey Labonal double jersey
Labonal simple jersey
0.4 Labonal yarn bridges 0.4 Labonal simple jersey
Labonal yarn bridges
0.2 0.2

0
B 0

−0.2 −0.2

−0.4
−0.4

−0.6
−0.6
3 samples, 5mm, 1.5Hz, 3kPa, sock on smooth polyurethane
3 samples, 5mm, 1.5Hz, 6kPa, sock on smooth polyurethane
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Coefficient Of Friction Coefficient Of Friction

0.8 0.8
Labonal double jersey Labonal double jersey
Labonal simple jersey Labonal simple jersey
0.6 Labonal yarn bridges 0.6 Labonal yarn bridges

0.4 0.4

0.2 0.2

0 0

−0.2 −0.2

−0.4 −0.4

−0.6 −0.6
3 samples, 5mm, 1.5Hz, 10kPa, sock on smooth polyurethane 3 samples, 5mm, 1.5Hz, 15kPa, sock on smooth polyurethane
−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 62: Coefficient of friction for Labonal socks under a) 3 kPa, b) 6 kPa, c) 10 kPa and
d) 15 kPa in contact with smooth polyurethane for a friction cycle. The mean curves were
calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA. The letter B identifies the half period where the movement of the pin
follows bridges orientation.
As one can observe in figures 25 and 26, friction experiments with smooth
polyurethane counterpart gave similar results than friction experiments with Lorica®
counterpart. The friction force and coefficient of friction values are however 1.3 to 2 times

138
higher in the case of smooth polyurethane which is probably due to a greater real contact area
and adhesion force.

2 *74F7EF88CF8B7856CFDB68678D47F998
*74F7EF88CF8B7856CFDB68678D47F998

12 "-7BD57F89 8D69 F79 !CBDE4.9 49 F79 85CCF9 CB7F479

D89

Figures 27 to 32 give the friction force and the coefficient of friction of Labonal socks
for a friction cycle for two different counterparts; Lorica® and smooth polyurethane. These
curves confirm that friction is higher when the socks are rubbed against smooth polyurethane;
friction force and coefficient of friction are multiplied by about 1.2, 1.4, 1.7 and 1.9 for
pressures respectively equal to 3, 6, 10 and 15 kPa. The adhesion part of friction is more
important for the smooth polyurethane counterpart than for the Lorica® counterpart. For a
same increase in pressure, the real contact area therefore increases more in the case of cotton
rubbing against smooth polyurethane than in the case of cotton rubbing against Lorica®
which explains that the difference between the curves gets higher with pressure.

139
 Friction force (N) Friction force (N)
1 2.5
on Lorica
on Lorica 2 on smooth polyurethane
on smooth polyurethane
0.5 1.5

0 0.5

−0.5
−0.5
−1

−1.5
−1
−2
3 samples, 5mm, 1.5Hz, 3kPa, Labonal double jersey
3 samples, 5mm, 1.5Hz, 6kPa, Labonal double jersey
−2.5
−1.5 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)
Time (s)

a) b)

Friction force (N) Friction force (N)

5 8
3 samples, 5mm, 1.5Hz, 10kPa, Labonal double jersey
on Lorica
4 on Lorica
6
on smooth polyurethane on smooth polyurethane
3
4
2
2
1
0
0

−1 −2

−2 −4

−3
−6

−4 3 samples, 5mm, 1.5Hz, 15kPa, Labonal double jersey

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 63: Friction force of Labonal double jersey in contact with Lorica® and smooth
polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves were
calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA.

140
 Coefficient Of Friction Coefficient Of Friction
0.6 0.6
on Lorica
on smooth polyurethane on Lorica
0.4 0.4 on smooth polyurethane

0.2 0.2

0 0

−0.2 −0.2

−0.4
−0.4

−0.6
−0.6
3 samples, 5mm, 1.5Hz, 3kPa, Labonal double jersey
3 samples, 5mm, 1.5Hz, 6kPa, Labonal double jersey
−0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Coefficient Of Friction Coefficient Of Friction

0.8 0.8
on Lorica
on Lorica
0.6 0.6
on smooth polyurethane
on smooth polyurethane
0.4 0.4

0.2 0.2

0 0

−0.2 −0.2

−0.4 −0.4

−0.6 −0.6
3 samples, 5mm, 1.5Hz, 10kPa, Labonal double jersey
3 samples, 5mm, 1.5Hz, 15kPa, Labonal double jersey
−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 64: Coefficient of friction of Labonal double jersey in contact with Lorica® and
smooth polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves
were calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA.

141
 Friction force (N) Friction force (N)
1 2
3 samples, 5mm, 1.5Hz, 3kPa, Labonal simple jersey
0.8 1.5 on Lorica
on Lorica
0.6 on smooth polyurethane on smooth polyurethane
1
0.4
0.5
0.2
0
0

−0.2 −0.5

−0.4 −1

−0.6 −1.5

−0.8
−2
−1 3 samples, 5mm, 1.5Hz, 6kPa, Labonal simple jersey
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −2.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Friction force (N) Friction force (N)

4 6
3 samples, 5mm, 1.5Hz, 15kPa, Labonal simple jersey
on Lorica
on smooth polyurethane on Lorica
3
4 on smooth polyurethane
2
2
1

0 0

−1
−2
−2
−4
−3

3 samples, 5mm, 1.5Hz, 10kPa, Labonal simple jersey

−4 −6
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 65: Friction force of Labonal simple jersey in contact with Lorica® and smooth
polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves were
calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA.

142
 Coefficient Of Friction Coefficient Of Friction
0.5 0.6
on Lorica
0.4 on Lorica
on smooth polyurethane 0.4 on smooth polyurethane
0.3
3 samples, 5mm, 1.5Hz, 3kPa, Labonal simple jersey 0.2
0.2

0.1
0
0

−0.1 −0.2

−0.2
−0.4
−0.3

−0.4 −0.6
3 samples, 5mm, 1.5Hz, 6kPa, Labonal simple jersey
−0.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)
Time (s)

a) b)

Coefficient Of Friction Coefficient Of Friction

0.6
0.8
3 samples, 5mm, 1.5Hz, 15kPa, Labonal simple jersey
on Lorica
0.4 0.6 on Lorica
on smooth polyurethane on smooth polyurethane

0.2 0.4

0 0.2

−0.2 0

−0.4
−0.2

−0.4
−0.6
3 samples, 5mm, 1.5Hz, 10kPa, Labonal simple jersey
−0.6
−0.8 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)
Time (s)

c) d)

Figure 66: Coefficient of friction of Labonal simple jersey in contact with Lorica® and
smooth polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves
were calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA.

143
 Friction force (N) Friction force (N)
0.8 2
on Lorica
0.6 on Lorica on smooth polyurethane
1.5
on smooth polyurethane
0.4
1
0.2
B
0.5
0
0
−0.2
−0.5
−0.4
−1
−0.6
−1.5
−0.8
−2
−1
3 samples, 5mm, 1.5Hz, 6kPa, Labonal yarn bridges
3 samples, 5mm, 1.5Hz, 3kPa, Labonal yarn bridges −2.5
−1.2 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Friction force (N) Friction force (N)

4 6
3 samples, 5mm, 1.5Hz, 10kPa, Labonal yarn bridges
3 on Lorica
on Lorica 4
on smooth polyurethane
on smooth polyurethane
2
2
1
0
0
−2
−1

−2 −4

−3 −6

3 samples, 5mm, 1.5Hz, 15kPa, Labonal yarn bridges

−4
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 −8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 67: Friction force of Labonal yarn bridges in contact with Lorica® and smooth
polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves were
calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA. The letter B identifies the half period where the movement of the pin
follows bridges orientation.

144
 Coefficient Of Friction Coefficient Of Friction
0.6 0.6

on Lorica on Lorica
0.4 on smooth polyurethane 0.4 on smooth polyurethane

0.2 0.2

B
0 0

−0.2 −0.2

−0.4 −0.4

−0.6 −0.6
3 samples, 5mm, 1.5Hz, 3kPa, Labonal yarn bridges 3 samples, 5mm, 1.5Hz, 6kPa, Labonal yarn bridges

−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

a) b)

Coefficient Of Friction Coefficient Of Friction

0.6 0.8

on Lorica on Lorica
0.6
0.4 on smooth polyurethane on smooth polyurethane
0.4
0.2

0.2
0
0
−0.2
−0.2

−0.4
−0.4

−0.6 −0.6
3 samples, 5mm, 1.5Hz, 10kPa, Labonal yarn bridges
3 samples, 5mm, 1.5Hz, 15kPa, Labonal yarn bridges
−0.8 −0.8
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s) Time (s)

c) d)

Figure 68: Coefficient of friction of Labonal yarn bridges in contact with Lorica® and
smooth polyurethane under a) 3 kPa, b) 6 kPa, c) 10 kPa and d) 15 kPa. The mean curves
were calculated from 3 experiments for a friction distance of 5 mm and a friction frequency of
1.5 Hz on the TFA. The letter B identifies the half period where the movement of the pin
follows bridges orientation.

145
 2 "-7BD57F898D69F7985CCF949F79/BC609D895479CA9
8F4D78898F779

1,5

rough
1
smooth

0,5

T
Friction force (N)

0
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8

-0,5

-1

-1,5
Time (s)

Figure 69: Friction force of Labonal terry jersey in contact with both rough and smooth
stainless steel pins. The experiment was carried out on the LPMT’s tribometer for a sliding
distance of 30 mm, a sliding frequency of 0.63 Hz and a pressure of 15 kPa. The letter T
identifies the half period where the movement of the pin follows terries orientation.

As can be observed in figures 33 and 34, the friction force gets higher when the pin
roughness increases. Friction is due to deformation and adhesion. As the contributions of
adhesion and deformation can not be measured independently, it is not possible to compare
absolute values and one can only conclude on the relative importance of the two friction
components. Deformation mechanisms are important for the structured steel and Lorica®
surfaces. Adhesion usually gets higher when the contact surface increases, i.e., when the
surface roughness decreases. Here, the friction force rises with the surface roughness,
therefore it can be deduced that the contribution of the deformation component of friction is
essential. It is nevertheless interesting to see that basic and terry jersey structures are not

146
reacting in the same way when the pin roughness is changed. The difference between both
signals is less important in the case of terry jersey (40% rise) than for basic jersey (100%
growth); the contact area between the sock and the pin is indeed assumed to be smaller for the
terry jersey than for the simple jersey.
Moreover, experiments with the smooth stainless steel pin and the pin covered with
Lorica® give the same trend (figures 35 and 36). It can be observed that the friction force is
higher when the sock is in contact with Lorica® for both terry and simple jersey structures.
The difference is quite small for terry jersey socks while it is greater in the case of simple
jersey socks. Figure 35 indicates that the deformation of the textile structure is important (the
properties of the pin seem relatively unimportant) whereas figure 36 indicates that the
properties of the pin are important. It can therefore be assumed that the contribution of
deformation in friction is more important for the terry jersey structure than for the simple
jersey.

0,8 rough
smooth
0,6

0,4

0,2
Friction force (N)

0
0 0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8

-0,2

-0,4

-0,6

-0,8

-1
Time (s)

Figure 70: Friction force of Labonal simple jersey in contact with both rough and smooth
stainless steel pins. The experiment was carried out on the LPMT’s tribometer for a sliding
distance of 30 mm, a sliding frequency of 0.63 Hz and a pressure of 15 kPa.

147
 1,5
Lorica®
probe pin with Lorica
recovered
smooth
Smooth aluminium-made
stainless steel probe
pin
1

0,5
Friction force (N)

0
T
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7

-0,5

-1

-1,5
Time (s)

Figure 71: Friction force of Labonal terry jersey in contact with Lorica® and smooth
stainless steel pins. The experiment was carried out on the LPMT’s tribometer for a sliding
distance of 5 mm, a sliding frequency of 1.5 Hz and a pressure of 15 kPa. The letter T
identifies the half period where the movement of the pin follows terries orientation.

0,8

Lorica®
probe pin with Lorica
recovered
0,6 smooth aluminium-made probe
Smooth stainless steel pin

0,4

0,2
Friction force (N)

0
0 0,1 0,2 0,3 0,4 0,5 0,6 0,7

-0,2

-0,4

-0,6

-0,8
Time (s)

Figure 72: Friction force of Labonal simple jersey in contact with Lorica® and smooth
stainless steel pins. The experiment was carried out on the LPMT’s tribometer for a sliding
distance of 5 mm, a sliding frequency of 1.5 Hz and a pressure of 15 kPa.

148
 II. COMPRESSION EXPERIMENTS
The TFA and the LPMT’s tribometer allowed us to compare different sock knitted
structures in terms of friction. We have seen that the terry jersey and the yarn bridges
structures give higher friction than the simple and the double jersey when rubbed against
Lorica®. However, the friction exerted on the foot skin is not the only parameter to take in
account when choosing a running sock. The sock, by its structure and thickness, can reduce
the pressure on the skin which may lead to less severe skin irritations. Compression
experiments were therefore realised.
The terry jersey structure intuitively seemed better than the simple jersey structure for
running application because it is thicker and therefore is expected to present a higher static
“damping effect” i.e. a higher distribution of the mechanical stresses through the sock
structure. For a given applied pressure on the sock structure, the pressure perceived by the
skin is expected to be lower for the terry jersey than for the simple jersey. Compression
experiments were carried out in order to investigate this argument. The objective of these
experimentations was to measure, for a given applied pressure (same soil, same shoe and
same runner), the pressure perceived by the foot skin for the different sock structures. As
shown figure 17 chapter II, there is a zone that will not be considered for the compression
results. This zone is identified by a grey area.

12 EA89CD4EC4DF8B56EC8
EA89CD4EC4DF8B56EC8
The plunger stops sinking in the sock structure when the perceived pressure reaches
15 kPa. The maximal perceived pressure is thus the same for all the socks while the maximal
applied pressure necessary to get 15 kPa through the sock depends on the knitted structure. It
has to be specified that no saturation of the Flexiforce® was observed during the compression
experiments. Figures 37 and 38 respectively show the compression results of Labonal socks
and of terry jersey 10 – 16 and simple jersey 10 – 16. Figure 37a gives the maximal applied
pressure for Labonal socks, the socks can be ranged from the lowest to the highest maximal
pressure: yarn bridges, terry jersey, double jersey and simple jersey. Figure 38a confirms that
the maximal applied pressure is higher for simple jersey than for terry jersey structure.
Moreover, for a given applied pressure, figures 37c and 37d indicate that simple jersey
gives the lowest pressure on the skin, terry jersey and yarn bridges show similar perceived
pressure and the highest pressure is obtained with double jersey structure. Double jersey is

149
thicker than simple jersey but may not be soft like terry jersey so that less cushioning is
expected. This is supported by table 4 which gives the thickness of the socks under 49 Pa (To)
and 15 kPa (Tm) and the difference To – Tm which corresponds to the compressible thickness
of the socks. Figures 38c and 38d corroborate the fact that, for the same applied pressure,
simple jersey produces less pressure on the skin than terry jersey.

25 15
Labonal terry jersey Labonal terry jersey
Labonal double jersey Labonal double jersey
20 Labonal simple jersey Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges

Perceived pressure (kPa)

Applied pressure (kPa)

10
15

10
5

0 0
−0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 0.5 1 1.5
Depth penetration (mm) Depth penetration (mm)

a) b)

15 15
Perceived pressure (kPa)

Perceived pressure (kPa)

10 10

Going to Going fro

5 5
Labonal terry jersey Labonal terry jersey
Labonal double jersey Labonal double jersey
Labonal simple jersey Labonal simple jersey
Labonal yarn bridges Labonal yarn bridges

0 0
0 5 10 15 20 25 0 5 10 15 20
Applied pressure (kPa) Applied pressure (kPa)

c) d)

Figure 73: Compression results of Labonal socks; a) applied pressure versus depth
penetration, b) perceived pressure versus depth penetration, c) perceived pressure versus
applied pressure for the forward direction (compression) and d) perceived pressure versus
applied pressure for the back direction (decompression). These mean results are calculated
from 10 experiments.

150
 25 15
Terry jersey 10−16 Terry jersey 10−16
Simple jersey 10−16 Simple jersey 10−16
20

Perceived pressure (kPa)

Applied pressure (kPa)

10
15

10
5

0 0
−0.5 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
Depth penetration (mm) Depth penetration (mm)

a) b)

15 15
Perceived pressure (kPa)

Perceived pressure (kPa)

10 10

Going to
5 5
Terry jersey 10−16 Going fro
Simple jersey 10−16
Terry jersey 10−16
Simple jersey 10−16

0 0
0 5 10 15 20 25 0 5 10 15 20 25
Applied pressure (kPa) Applied pressure (kPa)

c) d)

Figure 74: Compression results of terry and simple jersey socks; a) applied pressure versus
depth penetration, b) perceived pressure versus depth penetration, c) perceived pressure
versus applied pressure for the forward direction (compression) and d) perceived pressure
versus applied pressure for the back direction (decompression). These mean results are
calculated from 10 experiments.

151
 KNITTED STRUCTURE MEAN TO (MM) MEAN TM (MM) TO - TM (MM)
Terry jersey 3.64 ± 0.06 2.43 ± 0.07 1.21 ± 0.13
Double jersey 1.95 ± 0.09 1.16 ± 0.01 0.79 ± 0.10
Simple jersey 1.89 ± 0.05 0.98 ± 0.02 0.91 ± 0.07
Yarn bridges 2.25 ± 0.05 1.27 ± 0.02 0.98 ± 0.07

Table 16: Labonal socks thickness under 49 Pa (To) and 15 kPa (Tm), mean results from 10
experiments.

2 /6D78
/6D78E47C8B74F7EF8
E47C8B74F7EF8
Figure 39 gives the compression results of terry jersey socks which only differ in their
yarn count (same stitch length).

It can be observed in figure 39a that the maximal applied pressure is similar for these
three socks. However, figure 39b points out that, for a given depth penetration, terry jersey
10 – 16 and 10 - 20 give similar perceived pressure while terry jersey 10- 12 induces higher
pressure. Moreover, it can be noticed in figures 39c and 39d that, for a given applied pressure,
the socks can be ranged from the lowest to the highest perceived pressure as follows: terry
jersey 10 – 20, terry jersey 10 – 16 and terry jersey 10 – 12.

The perceived pressure thus increases with the yarn count which can be related to the
sock softness. When the yarn count increases i.e. when the diameter of the yarn gets bigger,
the yarn gets less easy to bend and the sock structure gets less soft. Therefore, for a given
applied pressure and the same stitch length, the pressure the sock exerts on the skin increases
with the yarn count.

152
 25 15
Terry jersey 10−12 Terry jersey 10−12
Terry jersey 10−16 Terry jersey 10−16
20 Terry jersey 10−20 Terry jersey 10−20

Perceived pressure (kPa)

Applied pressure (kPa)

10
15

10
5

0 0
−0.5 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
Depth penetration (mm) Depth penetration (mm)

a) b)

15 15
Perceived pressure (kPa)
Perceived pressure (kPa)

10 10

Terry jersey 10−12

Terry jersey 10−12
Terry jersey 10−16
Terry jersey 10−16
Terry jersey 10−20
Terry jersey 10−20
5 5
Going to Going fro

0 0
0 5 10 15 20 25 0 5 10 15 20 25
Applied pressure (kPa) Applied pressure (kPa)

c) d)

Figure 75: Compression results of terry jersey 10 – 12, 10 – 16 and 10 – 20; a) applied
pressure versus depth penetration, b) perceived pressure versus depth penetration,
c) perceived pressure versus applied pressure for the forward direction (compression) and
d) perceived pressure versus applied pressure for the back direction (decompression). Let’s
remind that these three structures only differ by their yarn count which respectively equal to
12, 16 and 20 Ne(c) i.e. 141, 106 and 85 Tex. These mean results are calculated from
10 experiments.

153
 2 07FDFCBE86D6E8
07FDFCBE86D6E8
The ratio of the perceived energy divided by the applied energy was calculated for the
socks that have previously been studied in compression. As it has been outlined in chapter 2,
the Flexiforce® does not have a linear behaviour for low pressures, consequently the data
given here are qualitative ratios; the sock which transmitted the highest amount of energy has
a ratio equal to 1. These ratios are indicated in table 5.

SOCK RATIO PERCEIVED ENERGY / APPLIED ENERGY

Labonal simple jersey 0.69
Labonal double jersey 0.92
Labonal yarn bridges 1.00
Labonal terry jersey 0.73
Terry jersey 10 - 12 0.81
Terry jersey 10 - 16 0.84
Terry jersey 10 - 20 0.80
Simple jersey 10 - 16 0.58

Table 17: Socks energetic ratios. These ratios were calculated from the mean curves of the
applied pressure and the perceived pressure in function of the depth penetration.

This table indicates that Labonal double jersey and Labonal yarn bridges absorb less
energy than Labonal simple and terry jersey which have similar ratios. On the other hand, the
energetic ratio of terry jersey 10 – 16 is 50% higher than simple jersey 10 – 16 ratio. Terry
jersey 10 - 16 is thicker than simple jersey 10 – 16 but may not present enough softness to
absorb the energy transmitted to the sock (unlike Labonal terry jersey). The yarn count does
not seem to impact the energetic ratio.

154
 III. CONCLUSION
The pressure on the foot depends on the shoe (shape, size), the foot shape and the way
people tighten their shoes. If the space between the shoe and the foot is not sufficient, the
sock will exert a non negligible pressure on the foot even when it does not touch the ground
inducing a great friction force between the sock and the skin. On the other hand, shoes which
are not tightened enough will lead to excessive friction on the skin due to a higher sliding
distance. The footwear should thus be adapted to the runner foot. The sock structure should
notably be considered when choosing the size of the shoe. The size of the shoe has to be
higher for a thick sock structure like terry jersey than for a thin sock structure like simple
jersey.

Compression experiments indicated (figure 37 and table 5) that double jersey and yarn
bridges structures do not present a good static “damping effect”. These two structures will
then not be recommended.

The friction results obtained on the LPMT’s tribometer showed that, when friction
began, terry jersey coefficient of friction was lower than simple jersey coefficient of friction
and progressively got higher. Figure 40 shows the distance along which the coefficient of
friction of Labonal terry jersey is lower than the one of Labonal simple jersey versus the
applied pressure. It can be seen that this distance increases with the applied pressure. Labonal
terry jersey may therefore be a good compromise especially when high pressures are exerted
on the skin i.e. for stoutly built people. To be recommended, terry jersey socks need to present
compact, homogeneous and rather elastic terries which are oriented towards the toe of the
sock.

Moreover, in-shoe conditions are probably an intermediate between the conditions of

the LPMT’s tribometer and TFA’s conditions. Figure 41 gives the coefficient of friction
(calculated from figures 14 and 16 as the value of the coefficient of friction for the first half of
the curve) in function of the applied pressure for Labonal terry and simple jersey for both
tribometers. An intermediate between both tribometers would give a higher coefficient of
friction for Labonal simple jersey than for Labonal terry jersey.

155
 4

d (mm)
2

0
0 5 10 15 20
Normal pressure (kPa)

Figure 76: Distance along which the coefficient of friction of Labonal terry jersey is lower
than the one of Labonal simple jersey (d) versus the applied pressure. The data are taken
from friction experiments with the Lorica® pin on the LPMT’s tribometer for a friction
distance of 5 mm and a friction frequency of 1.5 Hz (cf. figure 22).

0.4

0.3
Coefficient of friction

0.2

0.1 Labonal terry jersey on LPMT's tribometer

Labonal simple jersey on LPMT's tribometer
Labonal terry jersey on TFA
Labonal simple jersey on TFA
0
0 5 10 15 20
Normal pressure (kPa)

Figure 77: Coefficient of friction of Labonal terry and simple jersey in contact with Lorica®
versus the applied pressure. The results are given for both the LPMT’s tribometer and the
TFA for a friction distance of 5 mm and a friction frequency of 1.5 Hz for the forward sliding
direction.

156
 The thickness of terry jersey socks may be appreciated by people with bony
prominences at specific areas of the foot or with pre-existing blisters. In this case, the terry
jersey structure will reduce the pain. In addition, terry jersey structure is easy and inexpensive
to realise whereas the technique of knitting socks with different thicknesses (higher thickness
at selected areas) is expensive. Terry jersey structure better corresponds to recreational
runners.

However, figures 37, 38 and table 5 indicate that the static “damping effect” of simple
jersey sock is greater than the static “damping effect” of terry jersey. It can be seen in figure
37d that for an applied pressure of 12.8 kPa, the perceived pressure equals to 10 kPa for
Labonal simple jersey and to 15 kPa for Labonal terry jersey. When comparing the friction
force of these two socks in contact with Lorica® (cf. figure 42), terry jersey under 15 kPa
induces 20 to 60 % more friction than simple jersey under 10 kPa. If we only consider the
static “damping effect” of the sock structure, it would therefore be recommended to wear
simple jersey rather than terry jersey socks.

0.8
Labonal terry jersey under 15 kPa
0.6
Labonal simple jersey under 10 kPa
0.4
T
0.2
Friction force (N)

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-0.2

-0.4

-0.6

-0.8

-1
Time (s)

Figure 78: Friction force of Labonal terry jersey under 15 kPa and Labonal simple jersey
under 10 kPa in contact with the Lorica® pin. The experiment was carried out on the LPMT’s

157
tribometer with a sliding distance of 5 mm and a sliding frequency of 1.5 Hz. The letter T
identifies the half period where the movement of the pin follows terries orientation.

Table 4 shows that the maximal perceived pressure is obtained for a greater depth
penetration in the case of the terry jersey compared to the simple jersey i.e. the maximal
perceived pressure is reached later in the case of the terry jersey. Terry jersey socks may then
provide a better dynamic damping than simple jersey socks. A vibratory analysis of terry and
simple jersey socks would determine which structure best resists to a shock and is proposed as
an outlook. We nevertheless investigated the ability of the different sock structures to
dissipate the energy of a shock dropping a ball of 9g from an initial height of 31.5 cm and
observing the height of the first rebound of the ball to calculate the coefficient of restitution e
which is given by the following expression: hn = e 2 n × h0 where hn is the height of the nth

rebound of the ball and h0 is the initial height. The terry jersey structure dissipated more
energy i.e. has a lower coefficient of restitution than the simple jersey structure as it is
reported in table 6.

SOCK COEFFICIENT OF RESTITUTION

Labonal terry jersey 0.39 ± 0.03
Labonal simple jersey 0.55 ± 0.02
Labonal double jersey 0.45 ± 0.01
Labonal yarn bridges 0.46 ± 0.02
Terry jersey 10 - 16 0.46 ± 0.02
Simple jersey 10 - 16 0.54 ± 0.02
Support 0.75 ± 0.01

Table 18: Coefficients of restitution. The coefficient of restitution of the support is given
because the socks are thin and smooth compared to the support and it therefore has an
influence on the quantitative results of shock absorption.

158
 1.0
0.8
0.6
Friction force (N)

0.4 Simple jersey - back side

0.2
0.0 Terry jersey - back side
-0.2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-0.4 Terry jersey - face side
-0.6
-0.8
-1.0
Time (s)

Figure 79: Friction force of Labonal terry jersey (face and back sides) and back side of
Labonal simple jersey under 15 kPa in contact with the Lorica® pin. The experiment was
carried out on the LPMT’s tribometer with a sliding distance of 5 mm and a sliding frequency
of 1.5 Hz. The curves given here are mean curves from 5 measurements.

When running, the foot skin is first subjected to a shock and then to progressive
compression and friction. It is not evident whether the shock or the friction energy has the
greater influence on skin irritations. We can thus conclude that the simple jersey structure is
better in terms of friction and quasi-static “damping effect” compared to the terry jersey
structure whereas the terry jersey structure has a higher dynamic “damping effect” than the
simple jersey structure. For hiking socks, we recommend wearing simple jersey socks and for
running socks either simple jersey or terry jersey socks are suitable. Terry jersey socks should
be worn with the terries in contact with the skin and oriented towards the toe for the friction
on the skin to be reduced. Figure 43 indeed shows that for a normal pressure of 15 kPa, the
friction force of the face of terry jersey (non terry side) is higher than the friction force of the
back of terry jersey (terry side) and of the back side of simple jersey.

We compared the different sock structures in terms of friction induced on the foot
during running. We have seen that Labonal terry jersey friction force has a particular shape
that may be due to the rotation of terries during a friction cycle. We have also discussed the
importance of the contribution of deformation and adhesion in friction for Labonal terry and
simple jersey socks. From this experimental study, it can be concluded that the deformation

159
component of friction is important in the case of terry jersey. Based on this experimental
result, we propose a discrete model of the friction of terry-covered textile surfaces described
on the next chapter.

160
Chapter IV: modelling
of sock-to-skin friction

161
 Résumé

Les résultats expérimentaux ont montré que le frottement des surfaces recouvertes de
bouclettes se déroule en 2 temps. Les bouclettes sont d’abord déformées i.e. orientées dans le
sens du mouvement du palpeur, puis elles sont frottées. Nous avons modélisé ce
comportement. Ce dernier chapitre présente tout d’abord les modèles de frottement et de
compression disponibles dans la bibliographie puis détaille le modèle proposé.

Les modèles de « brosses » pourraient être adaptés à la simulation du frottement entre

la peau et une chaussette bouclette car la structure d’une brosse et la déformation des poils de
la brosse sont très similaires à la structure du jersey bouclette et au mouvement des boucles
lorsqu’elles sont frottées. Comme dans les modèles de « brosses » étudiés, nous considérerons
chaque bouclette de manière indépendante.
Les modèles de frottement d’un pneu sur la route sont également intéressants car la
surface d’un pneumatique présente des sillons et pourrait être modélisée comme un polymère
dont la surface est recouverte de poils. Nous nous sommes inspirés du modèle de LuGre qui
permet la simulation de différents comportements en frottement, du purement élastique au
viscoélastique. Nous n’avons considéré que la partie élastique des équations de ce modèle.
L’équation de Van Wyk nous permet de déterminer la limite entre la compression des
bouclettes et la compression de la structure complète de la chaussette bouclette en réalisant un
ajustement grâce à une courbe expérimentale.

Nous avons modélisé la force de frottement d’une surface tricotée recouverte de

bouclettes sur la peau artificielle Lorica® Soft. Ce modèle à destination des industriels a pour
objectif de comparer de manière simple différentes chaussettes en termes de frottement induit
sur la peau afin de réduire le risque d’apparition d’irritations cutanées. Les modèles de
frottement et de compression existants sont tous fondés sur le fléchissement des poils, fils ou
fibres. Dans un premier temps, le mouvement des bouclettes a donc été simulé comme du
fléchissement et de la compression mais cette première approche a donné une modélisation
compliquée, très lente et peu satisfaisante. Nous avons donc simulé le comportement en
frottement comme une rotation des bouclettes.

162
 La partie expérimentale de ce travail a montré que la partie déformation du frottement
était plus déterminante que la partie adhésive qui sera négligée dans notre simulation. Le
frottement est divisé en deux phases, les bouclettes effectuent une rotation de leur angle initial
à leur angle final sans qu’il y ait de glissement entre les boucles et le palpeur, puis elles
restent en position finale et sont frottées avec un coefficient de frottement constant. Le
comportement des boucles est supposé élastique même s’il est plus vraisemblablement
viscoélastique. La partie non verticale des boucles n’est pas prise en compte. La force de
frottement est modélisée comme une fonction de l’orientation des bouclettes i.e. comme une
fonction du déplacement horizontal du palpeur. On considère deux groupes distincts de
bouclettes ; celles qui sont sous le palpeur et celles à l’avant du palpeur au début du
mouvement. La force de frottement totale résulte des forces de compression exercées par les
boucles sur le palpeur et est égale à la somme des forces de frottement individuelles des
bouclettes. L’instant auquel le palpeur touche la boucle, où la boucle arrive à sa position
finale ou encore l’instant où elle est lâchée par le palpeur dépendent de la place de la bouclette
dans la colonne de mailles. On calcule donc la force de frottement pour une colonne avant de
multiplier cette dernière par le nombre de rangées présentes sous le palpeur. Les bouclettes
sont considérées comme indépendantes les unes des autres.
Le moment d’un fil de bouclette peut être calculé à la fois à partir de la force de
frottement et de la force de compression exercées sur ce fil (cf. équation [E1]).
M (θ ) = Fc L sin θ = F f L cosθ [E1]

M est le moment en [Link], 2 l’angle formé par le fil de bouclette et la verticale au

tricot en rad, Fc la force de compression en N, Ff la force de frottement en N, L la hauteur de
la bouclette. Le déplacement vertical du fil de bouclette en mm est donné par l’équation [E2] :
d (θ ) = L(cos θi − cos θ ) [E2]
2i est l’angle initial du fil de bouclette en rad.
Des deux premières équations découle l’expression de la force de frottement [E3] :
Fc L sin θ
Ff = [E3]
L cos θi − d (θ )
Le déplacement horizontal du fil de bouclette est donné par l’équation [E4] :
z (θ ) = L(sin θ − sin θi ) [E4]
Pour exprimer la force de frottement en fonction du déplacement horizontal du fil de
bouclette, on suppose que :

163
 F f = λ (θ )z (θ ) [E5]

λ (θ ) est déterminé à partir des équations [E3], [E4], et [E5] :

Fc sin θ
λ (θ ) = [E6]
(L cosθi − d (θ ))(sin θ − sin θ i )
Nous avons considéré un coefficient λc constant pour calculer la force de frottement
d’un fil de bouclette :
Ff max Fc max sin θ max
λc = = [E7]
zmax (θ ) (L cosθi − d max (θ ))(sin θ max − sin θi )
θ max est l’angle de rotation maximal, Fc max et dmax sont obtenus à partir des courbes de
compression Kawabata en utilisant l’équation de Van Wyk. Pour le jersey bouclette Labonal,
λc = 2,3×10−3 N/mm/fil.
Les équations ci-dessus sont valables pour le sens aller du mouvement du palpeur.
La force de frottement modélisée est ajustée à une courbe expérimentale. L’ajustement
est réalisé de manière indépendante en sens aller et retour. Le paramètre d’ajustement est
l’angle de rotation maximal dans la direction considérée. Cet angle est calculé en appliquant
la méthode des moindres carrés à une courbe expérimentale. Il correspond à l’inclinaison
finale des bouclettes par rapport à la verticale au tricot. Les valeurs limites de cet angle sont
estimées à partir de photographies afin de corroborer les résultats de la modélisation.

La modélisation est très proche des résultats expérimentaux en ce qui concerne le

jersey bouclette Labonal. Afin d’obtenir un modèle simple facilement utilisable par des
industriels, nous avons fait de nombreuses hypothèses, les différences observées entre les
résultats de la simulation et les courbes expérimentales sont dues à ces hypothèses. Il se peut
notamment que la force normale appliquée ne suffise pas à coucher complètement les
bouclettes. Les bouclettes n’atteignent donc pas forcément leur angle maximal de rotation,
surtout en sens retour lorsqu’elles s’opposent au mouvement du palpeur, ce qui explique
l’asymétrie plus prononcée des résultats de la simulation.
L’introduction de la partie adhésive du frottement i.e. l’adhésion entre le palpeur et les
bouclettes mais aussi les forces d’attraction exercées par une bouclette sur ses voisines devrait
permettre d’améliorer le modèle. De plus, les bouclettes sont ici considérées comme
indépendantes alors qu’en réalité, chaque boucle touche et déforme sa voisine lorsqu’elle

164
bouge. La prise en compte de la force de compression exercée par chaque boucle sur la
suivante devrait également parfaire les résultats de la simulation.

165
 CONTENTS

Chapter IV: modelling of sock-to-skin friction __________________________________ 161

I. THEORETICAL BACKGROUND _____________________________________ 167
1. Friction models 167
A. Brush models 167
B. Road/tire dynamic friction models 170
2. Compression models 172
II. DISCRETE MODEL OF SOCK-TO-SKIN FRICTION _____________________ 176
1. Mechanical model’s assumptions 177
2. Mathematical expression of the friction force 179
3. Results and discussion 183
Literature cited _________________________________________________________ 186

166
 The previous chapter detailed the experimental friction results. These results indicate
that the friction of terry covered socks is a two stages process. First the terries are deformed
i.e. they are oriented in the pin sliding direction and secondly they are rubbed by the pin. We
modelled this behaviour and the simulation shows good correspondence with the experiment.
In this last chapter, previous studies on friction and compression models will first be outlined
then the model we propose will be fully detailed.

I. THEORETICAL BACKGROUND

12 DBECB785F988
DBECB785F988

12 B895C7899
Brush models may be adapted to simulate the friction between the skin and a terry
jersey sock since the structure of a brush and the deformation of the hairs of a brush are
respectively similar to the terry jersey structure and the movement of the terries when rubbed
against a counterpart material.

Kotomin and Avdeev [1] proposed a macromodel approach for experimental studies of
polymer brushes friction using ordinary tooth brushes with different stiffness. Experiments
were performed on pairs of brushes with various mutual penetrations and speeds of
displacement. Two types of toothbrushes were used for experiments: “medium” type brushes
i.e. mild brushes and “hard” type brushes i.e. rigid brushes. The speed of the crosshead
displacement varied in the range 1 – 500 mm/min. The penetration depth was 1, 3 and 7 mm.
The displacement range was 10 mm. A simultaneous recording of the shear force and the
normal displacement response with an accuracy of 0.2 µm were realised. A force-
displacement response during cyclic shift in two directions was studied. As could be
expected, the rate of force growth was higher for rigid brushes than for mild brushes. This rate
was therefore considered as characteristic of rigidity. Contrasting with the shear force, the
normal repulsive reaction for mild brushes is stronger than that for rigid brushes. Considering
the interaction of two single filaments of the opposite brushes, the authors identified two
different rubbing mechanisms. At the beginning of the movement, the filaments ends transfer

167
the stress to the whole filaments which are bent (cf. figure 1). Then, the disconnection of the
filaments of the two brushes may happen in two ways:
1 The sliding of bent filaments ends along each other up to separate the brushes
ends, called “slipping-pulling” mechanism
1 The deviation across common axis of the filaments axis leading to an
interpenetration of the opposite brushes without contact, called “rake-on-rake”
mechanism

Figure 80: Geometric model for filaments interaction in a) initial and b) separation
position [1].

The first mechanism was analysed to correspond to mild brushes with low penetration
and the second for rigid brushes with deep mutual penetration. A rather unexpected
phenomenon was observed: the normal displacement was negative at the beginning of the
cycle which corresponds to an “attraction” of the rubbing brushes while moving. This
attraction was considered as a consequence of the friction force reaction along the filament
axis. The normal component of this reaction actually acted as an attraction force. This force
was similar for both types of brushes as it depended mostly on the friction coefficient between
filaments. The existence of a critical velocity for the drop of the friction force was proved and
a possible link to critical displacement with interpenetration level was estimated.

168
 Spiegelberg and Andersson [2] used surface velocities obtained from a rigid body
model to simulate friction and wear in the contact between the rocker arm pad and valve
bridge in the cam mechanism of a diesel engine. They simulated friction using two different
friction models, a 3D brush model capable of handling transient conditions such as a varying
normal load and varying surface velocities and a Coulombian friction model. The 3D brush
model reproduced the friction in rolling and sliding contacts between rough surfaces. The
surfaces were modelled by bristles that bend depending on the relative tangential
displacement of the surfaces and the normal load at each point of contact. Each bristle was
assumed to deform independently. In a real contact, however, the tangential deformation in a
point would be influenced by the deformation in other points around it. Spiegelberg et al. [3]
previously simulated the motion of a cylinder between two planes, first with a step change in
velocity and then with an oscillating motion of the upper plane. They aimed to investigate the
friction in the contacts and the movement of the roller when the planes are in motion. The
friction in the rolling and sliding contacts was simulated with a brush model: one of the two
surfaces in contact was modelled with bristles on it (cf. figure 2).

Figure 81: The brush model. P: normal load, Qx: tangential load, R: equivalent contact
radius, qx: tangential traction, µp: limiting tangential traction [3].

The bristles were presumed to bend and give rise to traction when the opposite surface
moves relative to them. The imaginary bristles represented here the elastic deformation of the

169
surfaces in contact. A characteristic of the brush model is that each bristle deformed
independently of the others. The surfaces were assumed to be ideally smooth.

Brush models cited above all considered each bristle independently of the others. This
is an interesting hypothesis in our case as the interaction between a terry and the other terries
surrounding is complex to model.
Kotomin and Avdeev [1] investigated the interaction of two single filaments of two
opposite brushes while we are interested in the interaction of one single hair and Lorica®
smooth counterpart. Nevertheless, the analysis of the filaments deformation as bending may
also be pertinent in the case of terry jersey rubbing against skin during running.
Spiegelberg and Andersson [2] studied the friction and wear of two mechanical pieces
in contact. They considered that the hairs did not move vertically i.e. the distance between the
two counterparts was constant which is not the case for the skin of the foot in contact with a
sock. Spiegelberg et al. [3] first calculated the deformation of one single bristle then
multiplied this expression by the number of bristles in the contact to obtain the total friction
force. We followed a similar approach. However, they considered that sliding occurred in the
movement of a cylinder between two planes while we simulated friction without sliding.

2 1C42FDB7945DE9ABDEFDC95C789
Road to tire friction is interesting in connection with textiles as tire surface with its
tread designs may be modelled as a polymer with bristles on its surface.

Canudas de Wit et al. [4] suggested a dynamic model for friction, called the LuGre
model, which includes most of the friction behaviours which can be observed experimentally.
The model properties were investigated by analysis and simulation. Surfaces are very
irregular at the microscopic level and two surfaces make contact at a number of asperities.
This was modelled as two rigid bodies which make contact through elastic bristles (cf.
figure 3). When a tangential force is applied, the bristles will deflect like springs leading to
the friction force. If the force is sufficiently large, some of the bristles deflect so much that
they will slip. The phenomenon is highly random due to the irregular forms of the surfaces.
The LuGre model was based on the average behaviour of the bristles. The average deflection
of the bristles was modelled by z:

170
 dz v
= v− z where v is the relative velocity between the two surfaces. The
dt g (v )
function g is positive and depends on many factors such as material properties, lubrication,
and temperature. It needs not to be symmetrical. Direction dependent behaviour can therefore
be simulated. For typical bearing friction, g ( v ) will decrease monotonically from g(0) when
v increases which corresponds to the Stribeck effect (decrease of the friction force for low
velocities). The friction force generated from the bending of the bristles was described as:
dz
F = σ 0z + σ 1 where 30 is the stiffness and 31 a damping coefficient.
dt

Figure 82: The friction interface between two surfaces is thought of as a contact between
bristles [4]. For simplicity, the bristles on the lower part are shown as being rigid.

A term proportional to the relative velocity could be added to the friction force to
account for viscous friction so that:
dz
F = σ 0z + σ 1 + σ 2v
dt
Canudas de Wit and Tsiotras [5] later showed that the LuGre dynamic friction model
accurately simulated velocity and road/surface dependence of the tire friction force. They
therefore developed speed-dependent dynamic friction models based on simple punctual
contacts which described the tire/road interaction. The input parameters of these models had a
physical significance which allowed tuning the model parameters according to experimental
data. The models were also speed-dependent, which agreed with experimental observations.

171
 The LuGre model enables different friction behaviours to be simulated, from purely
elastic to viscoelastic behaviour. We drew our inspiration from the LuGre model, only the
elastic part of the equations above has been considered.

2 )5DF99B785F988
)5DF99B785F988
Van Wyk [6] studied the compression behaviour of a fibre wool mat and found a
mechanical expression allowing experimental curves to be adjusted. He exposed that the
compression of a mass of fibres involves several processes but that the problem is
considerably simplified when reduced to that of simple bending of the fibres. For a 3D fibre
wad*, it was assumed that the governing mechanism is that of fibres bending between contact
points. He further assumed that the crimps and scales of the wool fibres prevent them from
sliding at the contact points even though they may be titled with respect to the loading axis.
His key statement is that the vertical distance between contact points is proportional to the
contact point spacing. In his developments, he ignored possible twisting and extension of the
fibres. He further modelled the fibre elements as randomly oriented, the mass of fibres as
uniformly packed and neglected frictional forces. The fibres were supposed to all have the
same length and to be homogeneous, linearly elastic, with a round cross-section. It was
assumed that all the fibres were laid perpendicular to compression forces direction.
Compression and traction deformations at the contact points were ignored. He expressed
compression solicitation as bending between fibres which gets higher with the decrease of the
fibre mat volume during compression. Van Wyk’s equation can be written as follows:
m3 6 1 1 3
p = kY 3 4 3 − 3 1
ρ 5V V 0 2
p is the pressure applied on the fibre wad, k is a constant, m is the fibre wad mass, Y is
the Young’s modulus of the fibres, ρ is the bulk density of the fibres, V is the volume filled
with fibres under pressure, and V0 the volume filled with fibres under no pressure. Such a
simplified equation cannot be used near the asymptote of compression i.e. for assemblies
which have been compressed to a volume small enough for the incompressible volume of the
fibres to become significant or for assemblies for which the volume at zero pressure is finite,
so a more elaborate equation must be employed, such as:

*
fibre wad: tampon fibreux in French.

172
 3
6m3 D 1 1 A
p = kY 44 11 B − 38
5 ρ 2 C (V − V ') (V 0 − V ') 9
3

V’ is the maximum volume filled with a fibre wad under pressure i.e. the volume at
the asymptote of the compression curve. From this equation, compression energy W can be
calculated from 0 to p. Dupuis et al. [7] adapted the Van Wyk’s equation to the study of
textile fabrics. They considered De Jong et al. [8] assumption that the volume V0 is much
1
bigger than the volumes V and V’, so that the parameter is very small compared to
(V 0 − V ')3
1
the parameter . If V0 is much higher than V and V’, then:
(V − V ')3
3
1 6m3
p≈a with a = kY 44 11
(V − V ')3
5ρ2

aD A
p V
a 1 1
W = − 7 pdυ = − 7 dυ = B − 28
V 0 (υ − V ') 2 C (V − V ') (V 0 − V ') 9
3 2
0

Still supposing that V0 is much greater than V, it could be written:

= (V − V ')
a 1 p
W =
2 (V − V ')2
2
This gave:
2W 8W 3
V '= V − and a = 2
p p
Calculating V’ and a is therefore easy. For textile assemblies such as fabrics which do
not verify the condition that V0 is very high compared with V and V’, the following equations
were calculated:
D 1 1 A
p = aB − 38
C (V − V ') (V 0 − V ') 9
3

(V − V ') + 1.5a V − V 0 3
p
p
W = − 7 pdυ =
0 2 (V 0 − V ')

In order to get the minimum energy W, the parameters a and V’ have to be minimum.
W and V can be determined from experimental results. The couple amin and V’min whereby the
Van Wyk’s equation can be solved can thus be calculated. This equation enables us to

173
determine the limit between the compression of the terries of a terry jersey sock and the
compression of the whole knitted structure fitting a compression curve.

Toll [9] developed theories of fibre packing for use in manufacturing composite
materials. He defined the maximum unforced packing fraction as the volume fraction under
which fibres are compressed without opposition. The maximum packing fraction of force free
fibres was assessed from a statistical analysis of the distribution of fibre-fibre contact points.
Any compression beyond this limit requires the application of external pressure. Two regimes
have thus been defined in terms of fibre volume fraction: unforced packing, at sufficiently
low volume fractions for fibres to be separated with no forces acting between them; and
forced packing, at higher volume fractions where fibres formed a network and were bent
between contact points so that overall elastic forces were needed to keep the network from
expanding. Toll dealt with two aspects:
1 The limit between the two regimes, i.e. the maximum unforced packing
fraction
1 The mechanics of forced packing
The forced packing was believed to be governed by the bending of fibre segments
between contact points. A micromechanical theory was therefore developed, again based on
the contact point statistics, and equations relating the force response per unit area of a fibre
bed to the fibre volume fraction were derived for three basic types of assembly: a general 3D
wad, a planar mat of dispersed fibres, and a bundle of almost parallel fibres. Other types of
reinforcement structure, such as woven fabrics, and the effect of lubrication were also tackled.
Toll modelled the compressive response of fibres to network compression. His
objective was to provide a set of simple, yet mechanistic, models to describe the observed
phenomena and understand the underlying mechanisms. The fibre assembly was modelled as
a volume in which fibres of uniform and circular cross section, straight or curved, are
disposed at random positions, but without mutual intersection, and at orientations according to
a statistically homogeneous orientation distribution. There was a slight contradiction between
the randomness of position and the requirement of no intersection; the no intersection
condition supposed a short range correlation in the spatial distribution of fibres.

If the fibres are non-parallel and approximately straight, the governing deformation
mechanisms will be the same as for a fibre segment being bent between points of contact with
other fibres. Each fibre segment will touch each other as the assembly is compressed and thus

174
form new contacts points and new shorter fibre segments. This will happen at an increasing
rate toward higher volume fractions since the rate of encounter will be proportional to the
fibre volume fraction. Since a short segment is stiffer than a long one, and the number of
segments increases at an increasing rate, the overall compressive force per unit area will be a
nonlinear function of the fibre volume fraction. For the case of contact points formed by
fibre-fibre crossover, the chosen deformation unit was one fibre segment supported by two
other fibres and loaded by a third somewhere between the supports. The contact forces were
considered as punctual acting forces, unless the contacting fibres were very close to parallel,
in which case it was necessary to describe it as a line contact.
The contact zone therefore may no longer be confined to a negligibly small region
around the contact point but have a size comparable to the free segment length and will grow
along the fibres as they deflect. In the extreme case of closely packed, perfectly straight and
parallel fibres, the network will be almost as stiff as the fibres themselves, which, in this
context, are practically rigid. Consequently there must be some feature of the fibre
arrangement that lowers the packing efficiency and provides flexibility. Toll identified two
candidate mechanisms for this. One was that the fibres have a certain degree of misalignment
after all, and that this misalignment causes the fibres to cross over each other and bend
between the cross-over points in the same manner as in a non-aligned fibre network. The
other proposition was that the fibres are not quite straight, but rather, “wavy”, with one wave
corresponding to a beam segment loaded and supported at its extremities. Toll constructed a
simple model of weaves and mats where the structure is divided into two distinct regions: a
high volume fraction region a and a low volume fraction region b. He assumed that region a
behaved like the bundle itself, and was much stiffer than b. He presumed that no sliding takes
place across contact points and did not take the fibre length in account since fibre length is
important only if the number of contact points per fibre is small, so that the compliance of the
deformation units is influenced by the presence of free fibre ends. He considered that a fibre
network with reasonably long fibres; fibre length / fibre diameter ~ 1000 or more, would
already form a large enough number of contact points per fibre under its self-load to behave
as if the fibres were infinitely long.
Toll modelled a fibre assembly under compression. His results would be essential to
take in account the action of each fibre on the surrounding fibres.
Dayiary et al. [10] investigated the mechanism of pile deformation under a
compressive load in a cut-pile carpet and suggested a theoretical analysis of force and energy
of deformation. The elastic-stored bending energy was derived, the frictional effect due to the

175
sliding pile yarn with its neighbour and the frictional bending couple were also considered. It
was estimated that in a cut-pile carpet under vertical compressive load, the deformation of the
piles occurs in three steps. The first process was the beginning of loading which could be
different for new and used carpets because the hairs of a new carpet are quasi perfectly
vertical while the ones of a used carpet present a residual bending. In the second process, the
pile was deflected and a curvature occurred along its length. In this process, pile transferred
from a curved form to two sectional forms including a curved portion in bottom and a straight
portion on top. In the third and final position, there is a two-sectional form at the end of
loading or at jamming mode. Dayiary et al. believed that such a deformation could not be
created by bending phenomenon only and that friction between fibres into pile yarn may have
also been involved. The following assumptions were made:
1 A hair is an elastic vertical bar that is deflected under a compressive
concentrated load
1 A hair has a circular cross-section which is uniform along its length
1 Gravity is neglected compared with applied force
1 A hair has zero twist and consists of many fibres
1 Back of carpet is rigid
The total energy in a hair under loading is composed of two terms namely the bending and the
frictional energy. The total energy of pile deformation was shown to depend on geometrical
properties of yarn such as pile length and radius, on mechanical properties of yarn including
pile bending rigidity and coefficient of friction between pile yarns, and on applied
compressive load.

II. DISCRETE MODEL OF SOCK-TO-SKIN

FRICTION

Previous models that have just been detailed were all based on the bending of elements
that can be bristles, yarns or fibres. We modelled the friction force of knitted fabrics covered
with terries rubbing on the mechanical model of skin used in our friction experiments i.e.
Lorica® Soft. This model may be a useful tool to compare different socks and to assess the
potential of sock constructions to reduce friction induced skin injuries. The movement of
terries was first modelled as bending and compression as in previous studies but this model

176
was complicated, slow and did not give satisfying results. Therefore, the model we suggest is
not based on the bending of terries but on the rotation of terries from an initial to a final
position [11].

12 1FE67BE685F29869945CB79
1FE67BE685F29869945CB798
9945CB798
Theoretical friction includes deformation and adhesion. Terry-covered knitted fabrics
are very deformable materials. In the experimental chapter, it has been concluded that the
friction behaviour of these fabrics is dominated by deformation mechanisms. The model will
then consider the deformation of terries as the basic friction mechanism and adhesion will be
neglected. Dynamic friction is believed to primarily result from two mechanisms namely
terries rotation and friction. A similar two-phase mechanical solicitation has previously been
observed for velvet fabrics i.e. textile surfaces consisting of oriented hairiness of quasi-
constant length [12]. Moreover, the rotation is assumed to be reversible, i.e., the loops are
considered to have elastic or, more probably, viscoelastic behaviour. One loop is modelled as
two cut yarns. The input parameters of the model are basic textile and geometrical
characteristics of the sock. The modelling consists in formulating the friction force as a
function of the terry orientation which is determined by the horizontal displacement of the
contacting pin.

before friction during friction

F
Probe
θinitial
θfinal

Figure 83: two-phase friction of the terry yarns as simulated in the model.

177
 As it has been outlined, two different phases are distinguished in the terries movement.
Hairs are supposed to first rotate from an initial to a final position or angle without sliding
between them and the pin. Once in their final position, the hairs are assumed to keep this
position and be rubbed by Lorica® with a constant friction coefficient equal to the friction
force in final position divided by the normal load (cf. figure 4).

Two groups of terries are differentiated: terry yarns which are under the pin and terry
yarns which are in front of the pin when the pin begins to move. The total friction force for a
textile fabric is assumed to result from compression forces which are transmitted to the pin by
the rotating terries and to be equal to the sum of the friction forces for all individual terries.

The instants when the pin touches the terry, when the loop reaches its final position
and when it leaves the contact depend on the position of the terry in the wale. Therefore, the
friction force of a stitch wale is first calculated before being multiplied by the number of rows
present under the pin.

Each terry has been considered independent of the other terries. The final rotation
angle of the terries before the change of the movement direction depends on the position of
the terry. The last yarns taken by the pin have a smaller final angle than the other yarns. The
friction force of a hair is thus not necessarily equal to zero at the beginning of the friction
cycle. It is equal to zero for the yarns which are at the front of the pin while the terries under
the pin have a residual deformation due to the precedent friction cycle. Once the yarn enters
the contact, the friction force increases from its initial value to its maximal value in the
forward direction i.e. the value corresponding either to the maximal angle of the yarn in this
direction or to the maximal displacement of the pin. If the pin is not already at its maximal
displacement, the friction force of the hair will keep its maximal value until the terry yarn
leaves the contact or the pin reaches its maximal displacement. If the yarn leaves the contact,
its friction force becomes equal to zero. Similar phenomena occur in the reverse direction.

The model input parameters for Labonal terry jersey are listed in Table 1.

178
 PIN WIDTH 9 MM
Pin length 17 mm
Rubbing distance 5 mm
Period of the alternating movement 0.67 s
Number of terry yarns widthwise 1/mm
Number of terry yarns lengthwise 1.7/mm
Terries height 1.38 mm
Terries initial angle 7°

Table 19: Input parameters of the model for Labonal terry jersey sock.

2 16CF56CBE68F&DF99B788CF8DBECB78DEF8
16CF56CBE68F&DF99B788CF8DBECB78DEF8
The equations which have been used to calculate the friction force as a function of the
compression force and terry orientation are explained below. The notations used are listed in
table 2 and some parameters are described in figure 5.

MODEL MATHEMATICAL DEFINITION

NOTATION
d Terry yarn vertical displacement (mm)
Fc Compression force (N)
Ff Friction force (N)
θ Angle formed by the terry yarn and the vertical to the
knitted fabric surface (rad)
θi Terry yarn initial angle (rad)

L Terries height (mm)

M Torque ([Link])
z Terry yarn horizontal displacement (mm)

Table 20: Model mathematical notations and definitions.

179
 The torque of a single terry yarn can be computed from both friction and compression
forces as expressed in equation [E1].
M (θ ) = Fc L sin θ = F f L cosθ [E1]

d (θ ) = L(cos θi − cos θ ) [E2]

The friction force [E3] can be expressed from equations [E1] and [E2] as:
Fc L sin θ
Ff = [E3]
L cos θi − d (θ )

z (θ ) = L(sin θ − sin θi ) [E4]

We aim to plot the friction force in function of the terry yarn horizontal displacement.
We thus write Ff as:
F f = λ (θ )z (θ ) [E5]

Finally, [E3], [E4], and [E5] give the following expression of λ (θ ) :

Fc sin θ
λ (θ ) = [E6]
(L cosθi − d (θ ))(sin θ − sin θ i )
λ (θ ) is not constant as shown on figure 6. We nevertheless considered a constant
coefficient λc to calculate the friction force of a single terry yarn. λc is the slope of the dotted
line D and equals 2.3×10−3 N/mm/yarn for the Labonal terry jersey, it can be expressed as:
F f max Fc max sin θ max
λc = = [E7]
zmax (θ ) (L cosθi − d max (θ ))(sin θ max − sin θi )
θ max is the maximal rotation angle. Fc max and dmax are determined from Kawabata
compression curves using Van Wyk’s equation [6].
Figure 5 schematizes the geometry of a terry yarn when the pin moves in the forward
direction. The hair has a similar geometry when the pin goes in the reverse direction. The
equations above are given for the forward direction of the pin. When the pin moves in the
reverse direction, equations [E4], [E6] and [E7] are slightly modified:
z (θ ) = L(sin θ + sin θi ) [E8]

Fc sin θ
λ (θ ) = [E9]
(L cosθi − d (θ ))(sin θ + sin θ i )
F f max Fc max sin θ max
λc = = [E10]
z max (θ ) (L cosθi − d max (θ ))(sin θ max + sin θi )

180
 Fc

Ff d
θi z

L
θ

Figure 84: Geometrical characteristics and loading of a terry yarn.

-3 Friction force(N)
x 10
2.5
D
2

1.5

0.5

0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Horizontal displacement (mm)

Figure 85: Calculated friction force of a single hair versus horizontal displacement.

181
 SOCK θ a (°) θ r (°)
Labonal terry jersey 41 ± 6 -49 ± 9
Terry jersey 10 - 12 40 ± 12 -58 ± 6
Terry jersey 10 - 16 56 ± 9 -58 ± 12
Terry jersey 10 - 20 53 ± 12 -66 ± 7
Terry jersey 9 - 16 36 ± 16 -45 ± 12
Terry jersey 11 - 16 54 ± 6 -63 ± 8

Table 21: Mean maximal rotation angles in the forward ( θ a ) and the reverse direction ( θ r )
for terry jersey fabrics. These values were calculated from at least 9 measurements with
ImageJ software.

θa

(a)

θr

(b)

Figure 86: Photographs (zoom × 40) used to estimate (a) terries maximal angle for the
forward direction and (b) terries maximal angle for the reverse direction. These photographs
were taken under 15 kPa during a static experiment with the Lorica® pin.

182
 The modelled friction force is fitted to the experimental data. The fitting is realised in
both forward and reverse directions independently. The fitted parameter is the maximal
rotation angle in the studied direction. The maximal rotation angles in the forward and the
reverse directions are calculated applying the least squares method to an experimental friction
curve so that these angles do not exceed physical limits. These angles correspond to the final
inclination between terries and the vertical axis, their limit values can be assessed from
photographs as illustrated on figure 7. These estimations enable model results to be
corroborated. The maximal angles in both the forward and the reverse direction, namely θ a

and θ r , for the six terry jersey fabrics are given in table 3. Rotation angles for other terry
jersey fabrics than Labonal are indicated as comparison in table 3. It can be seen that the
standard deviations for these angles are very high; this is due to the heterogeneity of the
fabrics. When rubbed, not all the terries show the same behaviour, some rotate when the pin
moves, some just leave the contact. Simulations will therefore not be given for these fabrics.

2 3F94C98678B9E499B78
3F94C98678B9E499B78
A good correspondence was found between modelled and measured friction forces for
Labonal terry jersey as can be seen on figure 8.
Many hypotheses have been made to get a simple model. The differences between
experimental data and simulated curves are attributed to the model’s assumptions. A terry is
considered as two yarns i.e. the non-vertical part of the loop is not taken into account. Textile
fabrics are very deformable materials, especially terry jersey fabrics. The analysis of friction
experiments indicates that for terry jersey fabrics, the major part of friction is due to sock
deformation. The friction is thus restricted to terries rotation and rubbing i.e. adhesion is
neglected whereas theoretical friction includes both deformation and adhesion. Moreover,
terries are modelled as perfectly elastic while their real behaviour is rather viscoelastic.
Finally, the applied normal force may not be sufficient for the terry yarns to reach their
maximal rotation angles, especially in the reverse direction as they counter the movement
since they are oriented in the opposite direction. This may explain the more pronounced
asymmetry of simulated results compared to experimental data.

183
 Friction force (N)
1
Experimental
Simulation

0.5

−0.5

−1

−1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Time (s)

Figure 87 : Comparison between experiment and model for an apparent pressure of 15 kPa.

The displacement of the pin is modelled as a linear (triangular) displacement whereas

the movement of the oscillating table of the LPMT’s friction tester is sinusoidal. The model
was modified in order to simulate the real pin displacement but the result is very similar and
the computing interval was considerably increased. Some sliding was also introduced
simulating the hairs which did not reach their maximal rotation angle at the end of the forward
and the reverse movement to leave the contact once in their initial position. The
correspondence to experimental data was better for the modelling without sliding than for the
modelling including sliding. The maximal rotation angles of the terries were estimated using
photographs taken during a static experiment while real rotation angles correspond to a
dynamic situation. The angles that have been measured may therefore be overestimated as the
angle will increase with the time of loading. The overestimation is thought to be more
significant for the maximal rotation angle in the reverse direction since the resistance of
terries to sliding is higher in this direction. The maximal rotation angle in the reverse direction
was thus modelled as the angle for which the reaction of terries trims the pin weight. This

184
model gives similar results as the previous one for the steady states in the forward and the
reverse sliding directions.

The model may be improved by the introduction of the adhesion part of friction; the
adhesion between the pin and the yarns but also the attraction forces exerted by a yarn on the
neighbouring yarns around it. Moreover, the terry yarns are simulated here as independent of
each other while in reality, when the terry rotates, it will touch and deform the neighbouring
terry and so on. The model can thus be improved by considering the compression force
exerted by each hair on the next one.

The model may also be adapted to low and non hairy fabrics such as yarn bridges and
simple or double jersey. The parameter L can be measured for the yarn bridges fabric and may
be estimated as equal to the yarn diameter for non hairy fabrics. The initial angle is assumed
to be around 90° for non-hairy fabrics. For yarn bridges, it seems difficult to neglect the non-
vertical part of the loop as the most part of it is horizontal.

185
 Literature cited

1. Kotomin, S.V. and N.N. Avdeev, Macromodel approach to experimental study of

polymer brushes rubbing. Wear, 1998. 222(1): p. 21-28.
2. Spiegelberg, C. and S. Andersson, Simulation of friction and wear in the contact
between the valve bridge and rocker arm pad in a cam mechanism. Wear, 2006.
261(1): p. 58-67.
3. Spiegelberg, C., S. Björklund, and S. Andersson, Simulation of transient friction of a
cylinder between two planes. Wear, 2003. 254(11): p. 1170-1179.
4. Canudas-de-Wit, C., et al., A New Model for Control of Systems with Friction. IEEE
Transactions on Automatic control, 1995. 40(3): p. 419-425.
5. Canudas-de-Wit, C. and P. Tsiotras, Dynamic Tire Friction Models for Vehicle
Traction Control, in Conference on Decision and Control. 1999: Phoenix (Arizona,
USA).
6. Van-Wyk, C.M., Note on the compressibility of wool. Journal of Textile Institute,
1946: p. T285-T292.
7. Dupuis, D., G. Popov, and P. Viallier, Compression of Greystate Fabrics as a
Function of Yarn Structure. Textile Research Journal, 1995. 65(6): p. 309-316.
8. De-Jong, S., J.W. Snaith, and N.A. Michie, A Mechanical Model for the Lateral
Compression of Woven Fabrics. Textile Research Journal, 1986(56): p. 759-767.
9. Toll, S., Packing Mechanics of Fiber Reinforcements. Polymer engineering and
science, 1998. 38(8): p. 1337-1350.
10. Dayiary, M., S.S. Najar, and M. Shamsi, A new theoretical approach to cut-pile carpet
compression based on elastic-stored bending energy. The Journal of The Textile
Institute, 2009. 100(8): p. 688-694.
11. Baussan, E., et al., Experiments and modelling of skin-knitted fabric friction. Wear,
2010. 268(9-10): p. 1103-1110.
12. Praëne, J.M., Modélisation phénoménologique du comportement tribologique des
surfaces textiles, in Mécanique. 2007, Université de Haute-Alsace: Mulhouse. p. 174.

186
Conclusion

187
 This study deals with the mechanical contact between the foot skin of the recreational
runner and his socks. It aimed to determine which knitted structure should be recommended
for runners. There have indeed been several studies on the influence of the sock fibre
composition while the effect of the sock structure has not been investigated yet. The socks
that people commonly choose and the areas where blisters may appear have been pointed out.
The case of the plantar fascia zone is studied here. Cotton-made socks were investigated as
cotton is the material the most used in basic running socks. Four knitted structures were
chosen, namely simple jersey, double jersey, yarn bridges jersey and terry jersey. The socks
were studied independently of the shoes. The pressure applied on the foot and the stride
frequency have been determined using in-shoe measurements. In order to simulate the friction
between the socks and a skin equivalent, the sliding frequency was then fixed to 1.5 Hz and
the vertical forces ranged from 0.5 to 2.2 N. Friction experiments were carried out on two
different tribometers; a linear alternating tribometer especially designed for this study and the
EMPA Textile Friction Analyser. Neither the linear reciprocating dead weight tribometer nor
the TFA exactly reproduces the friction of the foot with the sock in the shoe. In-shoe
conditions are probably an intermediate between these two cases. The transverse properties of
the different knitted structures were also examined through compression experiments and a
first approach of shock absorption tests.

It is difficult to give universal conclusions about which sock is the best to wear during
running as the pressure and friction the foot skin undergoes highly depend on the foot shape,
the shoe shape, the mechanical properties of the shoe and the way to wear running shoes. The
shoe is supposed to absorb a part of the shock energy generated at each stride when the foot
contacts the ground. In order to reduce the pressure on the skin, the contact surface between
the sock and the shoe should be as large as possible; the shape of the shoe should be adapted
to the shape of the foot and the distance between the shoe and the skin should ideally be equal
to the thickness of the sock when it is not compressed. A lower distance will lead to an
increase of the contact pressure while a higher distance will increase the sliding distance of
the foot within the shoe.

Double jersey and yarn bridges structures were investigated because they were
supposed to be intermediates between the simple and the terry jersey structures. Their friction
behaviour shows they are intermediates but much closer to simple jersey. Moreover,
compression measurements revealed they are harder than the two other structures and

188
consequently less interesting for sports socks. To resume, terry and simple jersey structures
were found to give better results than double jersey and yarn bridges structures in terms of
friction and compression.

The conclusions drawn from this study will therefore only be about terry and simple
jersey structures. The thickness of terry jersey socks is a good, easy and inexpensive way to
reduce the pain people with bony prominences at specific areas of the foot or with pre-
existing blisters suffer from. The friction, compression and shock absorption experiments we
carried out showed that:
1 For a given normal load, terry jersey socks generate less friction than simple
jersey socks provided that the terries are compact, homogeneous and oriented
in the direction of the friction force.
1 Under progressive compression, for the same compression energy applied on
the sock, the simple jersey structure transmits less energy to the skin than the
terry jersey structure. Simple jersey socks consequently provide a better static
or quasi-static “damping effect” than terry jersey socks.
1 When a ball is dropped from a given height on simple and terry jersey
structures, the height of the first rebound of the ball is lower for a terry jersey
sock than for a simple jersey sock. The terry jersey structure thus gives a
higher dynamic “damping effect” or shock absorption than the simple jersey
structure.

We could thus conclude that:

1 When walking, the foot undergoes friction and progressive compression, the
confrontation of the friction and the compression results indicates that the best
knitted structure for hiking socks is the simple jersey structure. This can be
seen in the figure 1.
1 When running, the foot skin is first subjected to a shock and then to
progressive compression and friction. We do not know whether the shock, the
friction force or the compression energy has the greater influence on skin
irritations. The simple jersey structure is better in terms of friction and quasi-
static “damping effect” compared to the terry jersey structure. Therefore, if the
effect of the friction force and the compression energy is more important than

189
 the influence of the maximal compression force on the formation of blisters,
then we recommend wearing simple jersey socks when running.

0.8
Labonal terry jersey under 15 kPa
0.6
Labonal simple jersey under 10 kPa
0.4

0.2
Friction force (N)

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-0.2

-0.4

-0.6

-0.8

-1
Time (s)

Figure 88: Friction force of Labonal terry jersey under 15 kPa and Labonal simple jersey
under 10 kPa in contact with the Lorica® pin.

1 On the other hand, the terry jersey structure seems to have a higher dynamic
“damping effect” than the simple jersey structure. Consequently, if it is the
shock absorption that is more important than the friction force and the
compression energy regarding the formation of blisters, the knitted structure
recommended for running socks would be the terry jersey structure. Moreover,
this structure can be improved drastically with sufficiently compact and
homogeneous terries in contact with the skin and oriented towards the toe so
that the friction can be reduced when the foot slides towards the toe within the
shoe i.e. when the foot contacts the ground and is loaded with significant
normal and tangential forces. In that case terry jersey is better than simple
jersey in terms of friction.

190
 1 If a recreational runner has only terry jersey socks, it is recommended to
determine whether the terries are oriented towards the toes when worn in
contact with the skin. If it is the case, the socks should be worn with the terries
in contact with the skin. If the terries are oriented towards the heel when worn
in contact with the skin, which is the case for most of the terry jersey socks
which are commercialised, it is better to wear the socks with the terries in
contact with the shoe because they will be oriented, in this case, in the
direction of the friction force generated by the sliding of the foot in the shoe. It
can indeed be seen in figure 2 that when the terries are well oriented, the back
(terry side) of the terry jersey generates a lower friction force than the face
(non terry side) of the terry jersey whereas when the terries are inclined in the
wrong direction, it is the face of the terry jersey that gives less friction than the
back.

1.0
0.8
0.6
Friction force (N)

0.4 Simple jersey - back side

0.2
0.0 Terry jersey - back side
-0.2 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
-0.4 Terry jersey - face side
-0.6
-0.8
-1.0
Time (s)

Figure 89: Friction force of the face (non terry side), the back (terry side) of Labonal terry
jersey and the back of Labonal simple jersey under 15 kPa in contact with the Lorica® pin.

Moreover, we have seen that Labonal terry jersey friction force has a particular shape
that may be due to the rotation of terries during a friction cycle. It has been found that the
contribution of deformation is important in the case of the terry jersey structure. A discrete
model for the friction of textile surfaces recovered with terries is thus proposed that considers
the rotation and the rubbing of terries and neglects the adhesion. Hairs are supposed to first
rotate from an initial to a final position without sliding between them and the pin. Once in

191
their final position, the hairs are assumed to keep this position and be rubbed with a constant
friction coefficient. The modelled friction force is fitted to the experimental data. The fitting
is realised in both forward and reverse directions independently. The fitted parameter is the
maximal rotation angle in the studied direction. The maximal rotation angles in the forward
and the reverse directions are calculated applying the least squares method to an experimental
friction curve. After fitting, theoretical and experimental results are very similar.

The model may be improved by the introduction of the adhesion part of friction i.e. the
adhesion between the pin and the yarns and the attraction forces exerted by a yarn on the
neighbouring yarns around it. The terry yarns are simulated as independent of each other
while in reality, when the terry rotates, it will touch and deform the neighbouring terry. The
model may thus give better simulations by considering the compression force exerted by each
hair on the next one. The model may also be adapted to low and non hairy fabrics such as
yarn bridges and simple or double jersey. The height of the yarn can be measured for yarn
bridges fabric and may be estimated as equal to the yarn diameter for non hairy fabrics. Initial
angle may be assumed to be around 90° for non-hairy fabrics. In addition, the pin geometry
may be modelled by dividing the space into small areas which can be active or not depending
on the shape of the pin. The shapes of the foot and the shoe could thus be considered as these
two geometries define the contact area.

The pressure the sock exerts on the foot skin is different from one area of the foot to
another. The conclusions drawn by our study are given for vertical loads from 0.5 to 2.2N
corresponding to the plantar fascia zone. Complementary friction and compression tests under
higher normal loads can be carried out to investigate other areas of the foot.

Both TFA and LPMT’s tribometer can be improved in order to better simulate the
friction between the sock and the foot of a runner for example with the use of curved rubbing
elements that could simulate the contact of textile fabrics against skin overlying bony
prominences. The pin may also be designed in order to reproduce the shape of the tested foot
area and a shoe insole may be placed under the sock. Moreover, during the friction
experiments, no tension was applied to the sock whereas when it is worn the sock is subjected
to multidirectional tensions. Both the friction devices used are alternating tribometers while in
the shoe, the foot slides in a single direction. In addition, in the case of the LPMT’s

192
tribometer, the normal load is kept constant whereas for the TFA, it is the distance between
the pin and the sample that is fixed and we have seen that in-shoe conditions are probably
intermediate. A unidirectional tribometer with constant normal load and constant distance
between the pin and the sample may be developed in order to better simulate in-shoe friction.

Shock resistance tests may be carried out in order to determine which textile structure
best absorbs the shock energy when the foot contacts the ground. The ability of the different
socks to form folds which is related to their elasticity is also an important parameter to
investigate in future works as folds will increase sharply the friction and compression forces
the sock exerts on the foot. The ageing of the knitted structures is another important parameter
which can be studied. The simple jersey structure is thought not to change much with time
while for the terry jersey structure, the terries may be affected by repeated washings and we
have seen that the quality of the terries is very important in terms of friction i.e. compact,
vertical and homogeneous terries give reduced friction while heterogeneous and crushed
terries do not present interesting friction properties.

Finally, the friction and compression tests we carried out were realised in a dry
atmosphere. When running, the foot skin sweats and perspiration from other areas of the body
is accumulated in the shoe. The conclusions drawn by this study are qualitatively true for a
wet atmosphere but experiments on human subjects wearing well adapted socks (in terms of
fibre composition and lacing conditions) could be carried out in order to give quantitative
results.

193
 Dix chaussettes ont été étudiées, toutes en coton : quatre d’entre elles pour évaluer l’influence
de la structure tricotée sur le frottement (jerseys bouclette, simple, double et ponts de fil), trois sont
identiques au titre du fil près, et les trois dernières à la pointure près. Nous nous sommes attachés au
cas des phlyctènes apparaissant lors de la course à pied. Les zones du pied sujettes aux ampoules ont
été déterminées grâce à un sondage. La pression s’exerçant sur ces zones et la fréquence des foulées
ont été mesurées. Nous avons ainsi programmé une platine oscillante qui reproduit le frottement dans
la chaussure. Les mesures obtenues sur ce dispositif ont été comparées à celles du Textile Friction
Analyser de l’EMPA. En croisant ces résultats avec des mesures de compression et une première
approche de test d’absorption des chocs, nous avons observé que le jersey simple est plus indiqué
dans le cas de la marche car il exerce un frottement moins important que la bouclette et apporte un
meilleur amortissement quasi-statique. Dans le cas de la course à pied, il reste à déterminer si le choc
a un effet plus dommageable que le frottement et la compression exercée sur la peau. Si le choc est
primordial, le jersey bouclette sera recommandé, sinon le jersey simple. Nous avons développé un
modèle du frottement textile-peau en nous intéressant tout d’abord au cas des tricots bouclette. Le
mouvement des bouclettes lors d’un cycle de frottement est assimilé à une rotation entre un angle
maximal en sens bouclette et un angle maximal en sens inverse. Ce modèle donne des résultats très
proches de l’expérience.

Mots clés : contact peau - textile, chaussette, phlyctène, frottement, pression, modèle

Ten cotton-made athletic socks were studied: four in order to evaluate the influence of the
knitted structure on the sock - to - skin friction (terry, simple, double and yarn bridges jerseys), three
differ from the yarn count and three from the size. We focused our research on the case of skin
blisters generated during running. The areas of the foot where these skin affections may appear were
determined using a survey. The pressure exerted on these areas and the stride frequency were
measured whereby a linear alternative tribometer reproducing the friction in the shoe has been
programmed. The results obtained on this device were compared with the results obtained on the
EMPA’s Textile Friction Analyser. Complementary compression and shock absorption tests were
realised. It was observed that the best knitted structure for hiking socks is the simple jersey structure.
For running socks, if the effect of the friction force and the compression energy is more important
than the influence of the maximal compression force on the formation of blisters, then we recommend
wearing simple jersey otherwise we advise to wear terry jersey socks. A discrete model for the
friction of textile surfaces recovered with terries in contact with the skin is proposed. Terries are
supposed to rotate from an initial to a final position both in the forward and the reverse direction.
After fitting, theoretical and experimental results are very similar.

Key words: textile - to - skin contact, sock, skin blister, friction, pressure, model

194