POLISH RECOMMENDATIONS FOR GROUND IMPROVEMENT

WITH RIGID INCLUSIONS

- DESIGN, EXECUTION, QUALITY CONTROL

WS 5 Michał Topolnicki
RECENT ADVANCES Senior Technical Advisor, Keller Holding GmbH
IN RIGID INCLUSIONS Professor emeritus

ORGANIZATION
TC211 | GROUND IMPROVEMENT
Recommendations for Ground Improvement with
Rigid Inclusions - Design, Execution and Control

SCOPE:
Introduction
Chapters:
1 – General principles of rigid inclusion reinforcement
2 – Soil investigations for ground improvement
3 – Design
4 – Execution
5 – Quality control
Annex - Design aids
Details the methods and calculation procedures used to design
GI with RI and to dimension the reinforcing elements
Calculation examples
A – Road embankment on soft soil
B – Industrial floor on the ground reinforced with DSM columns
Literature

ISBN 978-83-01-22578-0 (12/2022; 398 pages)

[Link]
gruntowego-kolumnami-sztywnymi,988921148,[Link] 2
2
Concept for the monograph

 Notable lack of national regulations on the use of RIs;

 Joint initiative of Keller + Menard + IBDiM, supported by the national
federation of foundation contractors (PZWFS);
 Initial concept based on ASIRI (2010);
 Ultimately, the scope was revised and extended to include:
- DSM columns and dealing with soilmix material;
- design, execution and control of all system components, including
LTP and basal reinforcement;
- extensive Annex with design aids, providing also missing regulations
for traffic loads, allowable displacements, service life, etc. (90 pages);
- presentation and practical use of prEN 1997 and prEN 1990;
- two design examples (85 pages).

3
3
Scope of this presentation
 Geometrical constrains of RI systems
- recommendations for embankments,
- preliminary estimate of column spacing,
- layout of basal reinforcement,
- execution tolerances and related impacts,
 Modelling of RI systems
- representation of RIs in FE models,
- the use of unit cell models,
- relative soil stiffness and creep effects,
- verification of external bearing capacity,
 Design of basal reinforcement
- analytical vs FE models,
- PEBSTER research project,
 Execution risks
- observed column damages and causes,
- positioning of basal reinforcement relative to supports
 Conclusions
4
4
Example A

Embankment

Caps Basal
reinforcement

Rigid inclusions

5
5
 Geometrical constrains

Poland,
Road S7,
ca 2,500 km
of RI installed
2016-2017

6
6
Rigid Inclusions considered

7
7
Recommendations for embankments

For RI installed with a full displacement auger: lmin ≥ 4

8
8
Critical embankment height Hc

PEBSTER

McGuire et al., Critical height of column supported

embankments from bench-scale and field-scale
(d = dg = dp) tests, ISSMGE-TC211, IS-GI, Brussels, 2012

9
9
Preliminary estimate of column spacing Recommended embankment height Hd

·
·
·
·
·
·
·
·
·
·
·
·

· · ·

.
· · · · · · · · · · ·
· · · · ·

Example:
l = 1.7 m aw = 4.3% l’/dg=2.51 Hd /dg=4.2 Hmin = 1.32 m Hd = 1.68 m
H = 2 m, w = 0.25 m
dg = dp= 0.4 m (no caps) l = 2.3 m aw = 2.4% l’/dg=3.57 Hd /dg=6.1 Hmin = 1.88 m Hd = 2.44 m

10
10
Constrains related to basal reinforcement layout

Caps casted on site

Geogrid
Geotextile longitudinally
transversely

Prefabricated
caps

Rolls are typically 5 m

wide, and 100-300 m long
(narrower rolls possible,
wider unlikely)
Geogrid Geogrid
transversely longitudinally

11
11
Installation of geosynthetic reinforcement

Design axes

Design axes
Overlap

Overlap
Overlap

Caps Column heads

Overlap in longitudinal
reinforcement Overlap in transverse reinforcement

12
12
Installation of welded steel mesh reinforcement
w/o checking column positions

recommended
interlocking

C- hook

Sequence of laying

Sheet length
when checking
column positions Direction of placement
(from both sides)

0.5m 0.5m
0.5m

13
13
Impact of column installation tolerances on the design of basal reinforcement

column (as-built position)

l l = design column spacing

execution tolerance
l
Recommended tolerances for RI

. .

14
14
Impact of column installation tolerances on the design of basal reinforcement (Example A)

𝑅 , =𝜂 𝜂 𝜂 𝜂 𝜂 𝜂 𝑇 = 0.65 0.9 1 0.8 1 1 300 = 140.3 kN/m

Tk – the characteristic (short-term) tensile strength of the geogrid,

ηcr – effect of tensile creep due to sustained static load over the design service life
of the structure at the design temperature;
ηdmg – adverse effects of mechanical damage during transportation, installation and execution;
ηw – adverse effects of weathering;
ηch – adverse effects of chemical and biological degradation over the design service life
of the structure at the design temperature;
ηdyn – adverse effects of intense and repeated loading over the design service of the structure (fatigue);
ηjs – adverse effects of joints and seams between reinforcing elements.

,
Recommended: = 1.1
, ,

140.3
1.1 68.1 = 𝟕𝟒. 𝟗 ≤ = 91.1 kN/m
1.4 1.1

15
15
Unintentional rotation of a steel beam reinforcing the column

16
16
Impact of steel beam rotation on the column bending resistance X: design direction of bending
X X 𝛽: rotation angle
𝛽 𝐼 = 𝐼 𝑐𝑜𝑠 𝛽 + 𝐼 𝑠𝑖𝑛 𝛽
As
ℎ = (ℎ 𝑐𝑜𝑠𝛽 + 𝑏 𝑠𝑖𝑛𝛽)/2
ℎ Y Y
𝑊 = 𝐼 /ℎ
Ap
𝑏 𝜎, = ± ≤𝑓
/ ( )
dp
Example A: 𝑡 = 5 𝑦𝑒𝑎𝑟𝑠 𝛽= 0o 𝜎sd,max = 231 MPa
Column: Eb = 29 GPa
𝛽=10o 𝜎sd,max = 253 MPa
Nd = 208 kN
dp= 0.4 m; Ap ; Ip
Md = 32.3 kNm 𝛽=26o 𝜎sd,max = 310 MPa
Concrete C30/37:
Ec (5 years) = 29 GPa, 𝑡 = 50 𝑦𝑒𝑎𝑟𝑠 𝛽= 0o 𝜎sd,max = 267 MPa
Eb = 10.4 GPa
Ec,eff (50 years) = 10.4 MPa 𝛽=10o 𝜎sd,max = 292 MPa
Nd = 218.6 kN
IPE 180: Md = 35.5 kNm 𝛽=16o 𝜎sd,max = 311 MPa
b; h; As
𝑡 = 50 𝑦𝑒𝑎𝑟𝑠 𝛽= 0o 𝜎s,max = 278 MPa
Ix = 101 cm3
𝐸 =12.4 GPa
Iy = 1320 cm3 𝐸 , 𝐼 =𝐸 , 𝐼 −𝐼 +𝐸 𝐼 𝛽=10o 𝜎s,max = 304 MPa
Nd = 218.6 kN
Es = 205 GPa 𝐸 = 12,4 GPa Md = 37.5 kNm 𝛽=11o 𝜎s,max = 307 MPa
,
fyd = 310 MPa

17
17
Concrete cover X

IPE 180
Y
Example A
According to EN 1536
rp - rIPE ≥ 75 mm
99 mm > 75 mm
EC2 (2023), section 6.5 dp = 400 mm

The nominal cover shall be specified in the execution specification.

It is defined as a minimum cover plus an allowance in design for deviation:
cnom = cmin + Δcdev

For bored piles and for diaphragm walls the nominal cover values for durability and
bond of EC2 apply. The cover values according to EN 1536 and EN 1538 ... should be
checked additionally. The largest value acc. to EC2 and EN 1536 or EN 1538 applies.

The value for cmin shall satisfy the requirements for both bond and durability:
cmin = max {cmin,dur + ΣΔc; cmin,b; 10 mm}

18
18
Modelling of RI systems

19
19
Column modelling in FEM

What are the possibilities

for RI representation?

20
20
Column modelling in FEM
Second step: Full model
First step: Calibration
Normal axial force
Column length

Axisymm.,
volume elem.
Plane, EBR Plane, EBR Plane, plate
(default interface stiffness factors)
Plane, plate

Unit Cell system

 Caps complicate
calibration of EBR
and plate column
models 3D, standard mesh 3D, fine mesh
(RI modelled with VE) (RI modelled with VE)

21
21
Column modelling in FEM FE models
Calculation parameter Plane, Plane, 3D/standard 3D/fine
plate EBR mesh(1) mesh
Settlement of the road
109% 105% 95% 66 mm (100%)
surface
Selected results: col. 1 135% 100% 88% 17 mm (100%)
Horizontal
displacement col. 2 138% 106% 88% 16 mm (100%)
of the column head col. 3 154% 108% 92% 13 mm (100%)
col. 1 57% 79% 107% 121 kN (100%)
Normal force col. 2 76% 73% 112% 133 kN (100%)
at the head level col. 3 81% 79% 104% 156 kN (100%)
col. 9 89% 86% 99% 203 kN (100%)
(9) col. 1 98% 94% 97% 220 kN (100%)
Maximum axial
col. 2 105% 93% 98% 241 kN (100%)
(3)(2) force
(1) col. 3 105% 97% 95% 273 kN (100%)
in the column
col. 9 104% 95% 91% 341 kN (100%)
Positive bending col. 1 106% 87% 84% 22.1 kNm (100%)
moment in the col. 2 143% 74% 81% 18.8 kNm (100%)
column col. 3 159% 95% 82% 14.7 kNm (100%)
Negative bending col. 1 222% 16% 80% 18.6 kNm (100%)
moment in the col. 2 241% 19% 85% 14.1 kNm (100%)
column col. 3 267% 18% 83% 10.6 kNm (100%)
(1) Finite element mesh automatically generated by the programme

22
22
Road traffic loads for 2D and unit cell analyses

Load model prEN 1991-2 (= 40 - 9)

Direction x

pH = (pHx + pHy )/2

q1 = 9 kPa CUR 226. 2016: Design Guideline for Basal Reinforced Piled
Embankments, ed. S.J.M. van Eekelen and M.H.A. Brugman,
q2 = 40 kPa SBRCURnet and CRC Press, Delft, ISBN 978-90-5367-624-0,
The Netherlands.
Forsman, J.A, Korkiala-Tanttu, L.K., Uotinen V-M., Renewed
UC Finnish guideline on deep mixing, Proc. Deep Mixing – Best
Practice and Legacy, Int. Conference, Gdańsk, 2021.
(the depicted curve represents a relatively dense column
Topolnicki M., Road traffic loads for geotechnical analyses of spacing L/d = 2.17, and the stiffness ratio Ecol/Esoil = 15.5)
embankments, Ground Engineering, 6/2020.
23
23
Load distribution in a unit cell

Total load sharing:

A – directly on the supports; efficacy of arching A%=A/(A+B+C)·100%
(B+C) - on the ground between supports:
B - on the reinforcement, further transmitted to the supports
C - on the ground

24
24
Load distribution in a unit cell (selected analytical methods)

EBGEO Method

Prandtl Method Cone Method (Combarieu, 1998)

Hc

LTM
(Bohn, Vogt)
TASPIE+
([Link])
Klobe Method
CUR 226 Method
25
25
Load distribution in a unit cell (Example A)

Settlement [mm]

Comparison of system efficacy A% [%]

Method p = 12 kPa p = 22,2 kPa

Prandtl 73.6 75.1
Cone 81.7 84.4
EBGEO 62.0 same
Klobe 76.0 78.6
CUR 226 66.5 same
FE axisym. 73.6 76.9
(w/o geogrid)

FE axisymmetric model
26
26
Evaluation of the FE axisymmetric model (Example A)

Without With
geogrid geogrid
Characteristics Unit
p = 12 kPa p = 12 kPa p = 22.2 kPa
Average settlement of the road surface mm 81.3 71.8 76.5
Cap settlement mm 16.4 17.1 20.1
Ground settlement at the cap level mm 97.5 86.1 90.8
Column base settlement mm 15.8 16.5 19.3
Ground settlement at the column base mm 4.0 4.0 4.5
Normal force on the cap* kN 212.1 236.4 281.4
Max. axial force in the column* kN 316.0 329.7 379.7
Axial force at the column base level* kN 210.2 210.8 245.3
Post-construction settlement of the road:
- 260 to 2085 days (5 yrs.) 21.9 26.4
- 260 to 18510 days (50 yrs.) 33.5 38.1

* representative values

27
27
Relative stiffness of soft soil (Example A)
FE axisymmetric model

Absolute stiffness s Relative stiffness sδ

. .
For t = 50 years: , , . .
.

. .
For t = 0.7 year: , , . .
.

28
28
Are creep effects important? Field observations
H  6m Road embankment
Steel grid
Peat
Mud

Gyttja
9 to 12 m
Sand
Moraine clay
Pipes Ø100mm for the settlement probe
Concrete columns: Ø 0.36m, length up to 19.5 m, c/c 1.4 to 2.2 m

Soil Cone tip Moisture Organic Dilatometer

resistance, content content, modulus,
qc [MPa] wn [%] Iom [%] MD [MPa]
Peat 0 – 0.5 90 -150 30 – 50 1.0
Mud 0.5 – 1 60 - 90 10 - 20 1.0 – 2.0
Gyttja I 0.2 – 0.8
75 - 95 8 – 13 3 – 30
Gyttja II 1.0 – 2.0

Topolnicki, M. et al., 2019. Performance and modelling of road embankment supported on rigid inclusions and a transfer platform with steel
geogrid, Proc. of the XVII ECSMGE-2019, Reykjavik, DOI: 10.32075/17ECSMGE-2019-0811 29
29
Effects of creep
(Example A)

Fill
Creep
K-V
Org. model
silt
Creep
Ecm → Ec,eff
Creep
SSC model

Sand

Stages: 10 days (arbitrary starting time: WP, Columns, Caps); 5+30+5 days: Embankment;
220 days: Road commissioning, traffic loads; 5 & 50 years: Exploitation 30
30
Adopted FE model (Example A)
PLAXIS 3D
61,401 10-node
tetrahedral elements
(mean size 0.488 m)

31
31
Geogrid model Example A
EAshort = Jx,1day = 2500 kN/m
EAlong = Jx,100ys = 2025 kN/m
Tlong = 0.65 · 300 = 195 kN/m
Retardation time tr = 940 days

Standard model (Kelvin-Voigt)

PET geogrid

32
32
Effects of creep (Example A)
FE axisymmetric unit cell model

Mobilisation in time of:

Tensile force in the geogrid Normal force on the RI-cap and pore pressure

Maximum pore overpressure [kPa]

Normal force on the cap, A [kN]
Tensile force ZM [kN/m]

Cap edge, 0.45 m

(results for traffic load p =12 kPa)

33
33
Effects of creep (Example A)
Development of geogrid deformation (3D model)

34
34
External bearing capacity verification (for embankments)
,
Acc. to prEN 1997-3: RFA , MFA ,
, ,

 The verification of geotechnical limit states for individual inclusions may be omitted provided it is verified that the system
is able to redistribute loads without itself exceeding an ultimate or serviceability limit state
Recommendation: The ultimate bearing capacity of the column should be at least equal to the maximum design load
determined for the representative ULS verification method

Example A:
Parameter ULS Column number
(design verification
values) approach 1 2 3 4 5 6 7 8
Nhead,d EFA1) 120.0 221.0 233.1 263.4 365.0 349.7 212.5 130.1
kN MFA 87.6 154.3 168.1 190.6 279.1 256.7 153.4 93.4
Normal forces [kN]
Nmax,d EFA1) 208.4 309.0 316.8 351.1 457.9 440.0 300.2 218.6
kN MFA 146.7 214.9 226.5 251.9 341.9 317.9 214.0 154.4
Mmax,d EFA1) 32.5 24.7 16.1 7.6 3.4 15.4 27.4 35.5
kNm MFA 29.8 24.5 15.5 8.7 3.2 15.5 27.4 32.8
1) Values after multiplication by a partial factor γE = 1.35
Bending moments [kNm] Nd,max = 457.9 kN < Rri,ult = 481.3 kN
35
35
Horizontal column displacements (Example A)

max 29.9 mm < 30 mm

Recommendation: The horizontal displacement of the outer columns under

the embankment should not exceed 30 mm.

36
36
Horizontal column displacement
Field observations
H  6m
-10 0 10 20 30
0.0

I2
-5.0
Instrumented columns (I)
G2
and ground inclinometers (G)

G1
-10.0

I1
I1

-15.0

I2

Topolnicki, M. et al., 2019. Performance and modelling of road embankment supported on rigid inclusions and a transfer platform with steel
geogrid, Proc. of the XVII ECSMGE-2019, Reykjavik, DOI: 10.32075/17ECSMGE-2019-0811 37
37
Design of basal reinforcement

geotextile
transversely

filling with peat

protective
pads

38
38
Comparison of analytical and numerical methods (Example A)

Design tensile force in the geogrid [kN/m] for methods

Direction of
stretching EBGEO(1),(2) Klobe(1),(2) CUR 226(1),(2) Mean(3) FE (UC)(1) FE (3D)(4)
EFA(5)
Across embankment 93.3 56.8 46.1 65.4 54.0 66.7
(direction x)
Along embankment as for x(6) 47.0
(direction y)
MFA
Across embankment 89.6 57.7 48.2 65.2 53.6 68.1
(direction x)
Along embankment as for x(6) 48.3
(direction y)
(1) For unit cell load p =2.2 kPa
(2) For relative ground stiffness at the cap level ksδ = 250 kN/m3
(3) Mean for three analytical methods
(4) For traffic load including heavy vehicle
(5) Values multiplied by γ = 1.35
E
(6) In case of analytical and FE calculations for the unit cell the tensile force is the same in both directions.

39
39
Geogrid modelling in 3D FEM (Example A)

40
40
Resutanat tensile force in the basal reinforcement based on analytical calculations

Membrane action Earth pressure

EBGEO, BS8006, CUR 226

Recommendation

41
41
 Research project PEBSTER (Piled Embankments with Basal STEel Reinforcement)
Large scale test 5 × 5 m
Small scale tests 1.1 × 1.1 m Deformed steel mesh
top
steel plate
Level 4
water cushion for surcharge load
Level 3
0.60 m

sand
Level 2
steel grid
Level 1
Level 0
PVC pile

ppt
foam cushion tap
steel plate

bottom plate
Example Example
rigid pile flexible pile

Pressure cells
------- DFOS measurements
ppt = pore pressure transducer
Session D2, Monday 16:15, Paleocene
Load transducers
Mark Schneider (large scale test)
Red lasers to measure pile
settlement
+ paper on small scale tests

42
42
Execution risks

43
43
Column damages

44
44
Other construction works

LTP layer
Fill Wrong Tempor. Wrong
storage

LTP layer Fill

Direction of placement

Area of increased deformations

Temporary storage Right
Right
Fill
Fill
LTP layer

45
45
Positioning of basal reinforcement
Case 1

Case 2

46
46
Working platform subsidence

47
47
Positioning of basal reinforcement

Recommendation:

 Above the supports, on a protective layer of aggregate not exceeding 30

cm in thickness (for reasons of efficiency), if during construction and
operation of the embankment there will be no excessive settlement of the
transmission layer leading to reinforcement contact with the supports, and
when the protective layer will have the required thickness after it has been
compacted above and between supports and is inspected before the basal
reinforcement is placed;
 Directly on the supports, if contact between the basal reinforcement and
the column heads or caps cannot be excluded during construction and
operation of the embankment; this solution requires protection of the
reinforcement from mechanical damage (e.g. protective pads,
chamfering/rounding of the edges of the supports, more robust
reinforcement); in the case of steel or composite bar meshes, it is also
necessary to adapt the dimensions of the mesh and the mesh design to the
dimensions of the support in order eliminate column puncture.

48
48
 General conclusions

 The response to the monograph has been very positive, both in Poland and in other
countries.

 However, longer observation of practical applications is needed to assess whether

the authors' objective of improving the understanding of the RI system, the quality of
design, and the quality of execution is being achieved.

 We need to better understand and evaluate the limitations of RI applications, as I

believe we still see too many failures.

49
49
 Thank you for your
kind attention
and
good luck for your RI projects!

[Link]@[Link]

Referenced publications are available at: [Link]

50