CHAPTER 7.

FASTE N E R S AND STRUCTU RAL JOI NTS

7.1 I ntrod uction connections involving fittings, bolts, rivets, welding,

A complete airplane structure is manufactured from etc. No doubt main or primary fittings involve more
many parts. These parts are made from sheets, ex­ weight and cost per unit volume than any other part
truded sections, forgings, castings, tubes, or machined of the aircraft structure, and, therefore, fitting and
shapes, which must be joined together to form sub­ joint design plays an important part in aircraft struc­
assemblies. The subassemblies must then be joined tural design.
together to form larger assemblies and then finally A blanket factor of safety for all types of fittings or
assembled into a completed airplane. Many parts of load conditions is not logical. The manner in which a
the completed airplane must be arranged so that they load is applied to a joint often involves a dynamic or
can be disassembled for shipping, inspection, repair, shock load, for example, joints or fittings in landing
or replacement, and are usually joined by bolts or gear. Single pin connections often undergo rotation or
rivets. In order to facilitate the assembly and dis­ movement between adjacent parts, thus producing
assembly of the airplane, it is desirable for such bolted faster wearing away of material in operation. Repeat­
or riveted connections to contain as few fasteners as ed loads often present a fatigue problem. In an air­
possible. For example, a semimonocoque metal wing plane there are certain main fittings which, if they
usually resists bending stresses in numerous stringers failed, would definitely cause the loss of the aircraft .
and sheet elements distributed around the periphery Thus, the design fitting requirements of the military
of the wing cross section. The wing cannot be made as and civil aviation agencies involve many special or
one continuous riveted assembly from tip to tip, but higher factors of safety. This is particularly so in
must usually be spliced at two or more cross sections. designs involving castings.
These splices are often designed so that four bolts, for General Design Considerations
example, transfer all the loads across the splice. These Joints are perhaps the most common source of failure
bolts connect members called fittings, which are in aircraft structure and therefore it is most important
designed to resist the high concentrated loads and to that all aspects of the design are given consideration
transfer them to the spars, from which the loads are when making the structural analysis. Failures may
distributi,:d to the sheet and stringers. The entire occur for various reasons but generally because of
structure for transferring the distributed loads from some factor, such as secondary stresses due to eccen­
the sheet and stringers outboard of the splice into a tricities, stress concentrations, slippage of connectors,
concentrated load at the fitting and then distribute this excessive deflections, etc., or some combination of
load to the sheet and stringers inboard of the splice is conditions, all of which are difficult to evaluate to an
considerably heavier than the continuous structure exact degree. These factors not only affect the static
which would be required if there were no splice. strength but have a great influence on the fatigue life
Many uncertainties exist concerning the stress dis­ of the joint and the adjacent structure.
tribution in fittings. Manufacturing tolerances are such ( 1 ) Eccentricities and their effect on the part of the
that bolts never fit the holes perfectly, and small joint and the surrounding structures.
variations in dimensions may affect the stress distribu­ • If eccentricities exist in a joint, the moment
tion. An additional margin of safety of 1 5% is used in they produce must be resisted by the adja­
the design of fittings. This fitting factor must be used cent structure.
in designing the entire fitting, including the riveted, • When a joint contains a dihedral or anhedral
bolted, or welded joint attaching the fitting to the angle (such as wing structure), a rib should
structural members. The fitting factor need not be be provided at the vertex of the angle to
used in designing a continuous riveted joint. eliminate the eccentricity that would exist.
The ideal aircraft structure would be a single • A joint of a truss structure containing an
complete unit of the same material involving one eccentricity produces secondary stresses
manufacturing operation. Unfortunately the present which would be accounted for.
day types of materials and their method of working • Eccentrically loaded bolt and rivet patterns
dictates a composite structure. Furthermore, general may produce excessively loaded connectors
requirements of repair, maintenance and stowage if eccentricity is not considered.
dictate a structure of several main units held to other (2) Fatigue considerations.
units by main or primary fittings or connections, with (3) Mixed fasteners: It is not good practice to
each unit incorporating many primary and secondary employ both rivets and bolts in combination in

Airframe Structural Design 207

 a joint. Due to a better fit for the rivets, the ( 1 3) Carefully select interference-fit fasteners which
bolts will not pick up their proportionate share produce sustained tensile stress (stress corro­
of the load until the rivets have deflected sion crack in fastened material).
enough to take up the clearance of the bolts in ( 1 4) In fastener installation, a minimum edge dis­
their bolt holes. tance (see Fig. 7 . 1 .2) of two fastener diameters
This tends to overload the rivets and may plus 0.03 inch and a spacing of an approximate­
induce premature failure. If such a combination ly four diameter shall be considered. Edge
is absolutely necessary, it is advisable to use distance is measured from the center of the
close tolerance bolts in reamed holes. fastener hole to the closest edge of the sheet and
(4) Overall efficiency: Both sides of the joint should a shorter edge distance of l .5 d may possibly be
be considered. It is possible to design a perfect considered in a few applications.
joint on one side by paying a heavy weight and
production penalty on the other. A compromise
should be made to make the joint and surround­
I
ing structure the most efficient. It is the overall
efficiency which is the prime consideration. I d
(5) Splices in discontinuous members, which act in
conjunction with a part or parts which are
continuous past the splice, should be made as
@ Hole

rigid as possible using generous splice members

and close fitting attachments, thereby minimiz­
ing slippage which might overload the con­
tinuous material and cause premature failure.
(6) Insufficient rigidity of surrounding structure
may cause excessive deflections and consequent Edge distance (e)
changes in direction and magnitude of loads on
certain joints, such as those in a landing gear
installation. Fig. 7. 1 . 2 Fastener minimum edge distance.
(7) Do not use spot welds on either side at the
joggled area of a joggled member; use rivets at
the joggle. ( 1 5) The use of rivets involving tension only is poor
(8) Do not use a long string of fasteners in a splice. engineering practice and should be held to a
In such cases, the end fasteners will load up first minimum.
and yield early. Three, or at most four, fasteners When secondary tension loads are imposed
per side is the upper limit unless a carefully on a standard aluminum rivet (such as the
tapered, thoroughly analyzed splice is used. attachment of a diagonal tension web to a stiff­
(9) Carefully insure against feather edges in all joint ener), use the tension allowable of rivet and
designs (see Fig. 7 . 1 . 1 ). The thickness of coun­ sheet combination test allowable data.
tersunk sheet shall be equal to or greater than

=
1 .5 times the depth of the countersunk head of Basic Criteria of Fastener Strength Allowable
the fastener at the fatigue critical areas. For The allowable loads are based on the lowest values of
other applications, t � h + 0.020 inch shall be the following criteria:
( 1) Bearing load (protruding head only) Fb,dt
-
considered.
where
Fb, Allowable ultimate bearing allowable
stress of sheet material is based on
either MIL-HDBK-5D "B" value per
�
--..-:-
t( sheet Ref. 7 .6 or other sources
ckness)
Nominal shank diameter

=
d
Nominal sheet thickness

( Jl:2 )
t

(2) Shear-off load F,u

Fig. 7. 1 . 1 Feather-edge in countersunk sheet.

where
F,u - Ultimate shear allowable stress of
fastener material is based on MIL­
HDBK-5D per Ref. 7.6 or other

( 1 0) When possible use a double shear splice.

( 1 1 ) Maintain a fastener spacing approximately four
d - sources
Nominal shank diameter
(See Fig. 7 . 1 .3)
times the fastener diameter or more. (3) Countersunk fastener and sheet combinations -
( 1 2) Probably the most single important item regard­ the allowable ultimate and yield loads are estab­
lished from actual test data.

=
ing detail structural design is the matter of
equilibrium. If the engineer will show the load in (4) Yield strength to satisfy permanent set require-
equilibrium for every part of the assembly, most . ultimate load
. . 1 oad (I"im1t 1 oad
ments at hm1t ).
errors will be prevented. 1 .5

208 Airframe Structural Design

 Fastener Diameter 3/32 1/8 5/32 3/1 6 1 /4 5/1 6 3/8 7/ 1 6 1/2 9/ 1 6
Fastener F,u
Material (Ksi) Ultimate Single Shear Load (lb/fastener)

l l OOF 9
5056(B) 28 203 363 556 802 1 450 2290 3280
2 1 1 7-T3(AD) 30 217 388 596 862 1 550 2460 35 1 0
201 7-T3 l (D) 34 247 442 675 977 1 760 2790 3970
201 7-T3(D) 38 275 494 755 1 090 1 970 3 1 10 4450
2024-T3 1 (DD) 41 296 531 815 1 1 80 2 1 20 3360 4800
Monel 49 355 635 973 1 400 2540 4020 5730
7075-H75 38 275 494 755 1 090 1 970 3 1 10 4450
A-286 CRES 90 - -
1 726 2485 44 1 5 6906 9938
Ti-6Al -4V
& Alloy Steel 95 - -
1 822 2623 4660 7290 1 0490 1 4280 1 8650 236 1 0
Alloy Steel 1 08 - -
207 1 2982 5 300 8280 1 1 930 1 6240 21210 26840
Alloy Steel 125 - -
2397 3452 6 1 40 9590 13810 1 8790 24540 3 1 060
H- 1 1 Steel 1 32 - -
253 1 3645 6480 1 0 1 20 1 45 80 1 9840 25920 32800
(Refer to MIL-HDBK-5D per Ref 7.6)

Fig. 7. 1 . 3 Fastener shear-off allowable loads.

Fittings connected.
For structural economy, engineers in the initial layout (1) Fitting Factor
of the aircraft should strive to use a minimum number Under the conditions outlined below, the U.S.
of fittings, particularly those fittings connecting units FAA requires an ultimate fitting factor of 1 . 1 5 to
which carry large loads. Thus, in wing structure, be used in the structural analysis (military aircraft
splicing the main beam flanges or introducing fittings may not require the use of a fitting factor).
near the centerline of the airplane are far more costly • A fitting factor is required for joints which
than splices or fittings placed farther outboard where contain fittings when the strength of the fitting
member sizes and loads are considerably smaller. is not proven by limit and ultimate load tests.
Avoid changes in direction of heavy members such as • This factor shall apply to all portions of the
wing beams and fuselage longerons as these involve fittings, the means of attachment (connec­
heavy fittings. If joints are necessary in continuous tions), and the bearing on the members joined.
beams place them near points of inflection in order • In the case of integral fittings, the part shall be
that the bending moments to be transferred through treated as fitting up to the point where the
the joint to be kept to a small magnitude. In column section properties become typical of the
design with end fittings, avoid introducing eccen­ member.
tricities on the beam; on the other hand make use of The fitting factor need not be employed in the
the fitting to increase column end fixity, thus com­ following cases:
pensating some of the weight increase due to fitting • When a type of joint made in accordance with
weight by saving in the weight of the beam. For approved practice is based on comprehensive
economy of fabrication, the engineer should have test data.
good knowledge of shop processes and operations. • With respect to the bearing surface of a patt, if
The cost of fitting fabrication and assembly varies the bearing factor used is of greater magnitude
greatly with the type of fitting, shape, and the required than the fitting factor.
tolerances. Poor layout of major fitting arrangement • If a casting factor (for casting materials) has
may require very expensive tools and jigs for shop been used which is of greater magnitude than
fabrication and assembly. Fittings, likewise, add the fitting factor.
considerably to the cost of inspection and rejections (2) Hinge Factor
of costly fittings because of faulty workmanship or The U.S. FAA requires that control surface
materials. The stress analysis of most connections or hinges, except ball and roller bearings, incorpor­
fittings is more complicated than for the primary ate a special factor of not less than 6.67 with
structural members due to such things as combined respect to the ultimate bearing strength of the
stresses, stress concentrations, bolt-hole fit, etc., thus softest material used as a bearing, and that
an additional factor of safety is necessary to give a control system joints subjected to angular motion
similar degree of strength reliability for connections as in push-pull systems, except ball and roller
provided in the strength design of the members being bearing systems, shall incorporate a special factor

Airframe Structural Design 209

 of not less than 3.33 with respect to the ultimate lower than for comparable bolts or screws.
bearing strength of the softest material used as a High tensile loads may pull out the clinch, or
bearing. severe vibrations may loosen the fastening.
Riveted joints are normally neither watertight
7.2 Rivets ( permanent fasteners) nor airtight; however, such joints may be
attained, at added cost, by using a sealing
Rivets are low cost, permanent fasteners well suited to compound.
automatic assembly operations. The primary reason
_

• Riveted parts cannot be disassembled for

for riveting is low in-place cost, the sum of initial rivet maintenance or replacement without destroy­
cost and costs of labor and machine time to set the ing the rivet.
rivets in the parts. Initial cost of rivets is substantially • Rivets produced in volume are not normally
lower than that of threaded fasteners because rivets made with the same precision as screw-ma­
are made in large volumes on high-speed heading chine parts.
machines, with little scrap loss. Assembly costs are Use flush-head rivets only where:
low. Rivets can be clinched in place by high-speed • High aerodynamic efficiency is required, or
automatic machinery. • Head clearances are inadequate; i.e. where a
( 1 ) Advantages: protruding head might interfere with a mating
• Dissimilar materials, metallic or nonmetallic, part or the operation of adjacent mechanisms.
in various thicknesses can be joined. Any Solid rivets and threaded collar rivets such as Hi­
material that can be cold worked makes a Loks shall not be used for the attachment of
suitable rivet. control surface hinges, hinge brackets, or similar
• Rivets may have a variety of finishes such as parts which must be readily removable. Blind
plating, parkerizing, or paint. rivets shall be used only where access is limited to
• Parts can be fastened by a rivet, if parallel one side. Blind rivets shall not be used on control
surfaces exist for both the rivet clinch and surface hinge brackets, wing attachment fittings,
there is adequate space for the rivet driver landing gear fittings, fixed tail surface attachment
during clinching. fittings or similar heavily stressed locations. Also,
• Rivets can serve as fasteners, pivot shafts, do not use them in the design of hulls, floats, or
spacers, electric contacts, stops, or inserts. tanks where a fluid tight joint is required. In
• Parts that are painted or have received other structural applications, the upset head must be
finishes can be fastened by rivets. visible for inspection. Fig. 7.2. 1 illustrates the
(2) Disadvantages: conventional rivet material characteristics and
• Tensile and fatigue strengths of rivets are usage .

Rivet Driving
•

Material Characteristics Used in Material Recommended Use and Limitations

1 1 00-F Easy-soft Soft aluminum, fiber, Low strength - not recommended in

plastics large dia. ( 1 / 1 6- 1 /8 inch)
5056(B) Medium Aluminum alloys Used if the strength of 5056 is adequate
and lower driving force is desirable
(3/32-3/1 6 inch)
2 1 1 7-T3 Medium-hard Aluminum alloys For use in sealing application in integral
(AD) fuel tanks ( 1 1 1 6-5/32 inch)
20 1 7-T3(D) Hard Aluminum alloys For general use
2024-T3 1 Hard Aluminum alloys High strength, limited use due to
(DD) refrigerated cond. (3/ 1 6- 1 /4 inch)
Monel(M) Very hard Copper alloys, steel Driving force too high for large dia.
and CRES
A-286 Very hard Titanium, steel and Up to 1 /4 inch dia. - Used in pylon, hot
CRES areas and corrosion environment
Titanium Very hard Steel and CRES Up to 3/1 6 inch dia. - Used in hot areas
B 1 20(K) and corrosion environment
7075-H75 Medium-hard Aluminum alloys Do not use in fatigue critical areas

Fig. 7. 2. 1 Conventional rivet material characteristics and usage.

2 1 0 Airframe Structural Design

Semi-tubular rivets surface. The COMP-TITE blind fastener overcomes
Semi-tubular rivets shall not be used in any structural this problem by forming a large and uniform bearing
application. They are normally used in soft materials surface that will not crush or delaminate the com­
such as rubber, leather, fabric, etc., with the upset posite material during installation. As illustrated in
head against a harder surface, such as a washer or Fig. 7.2.4, a coiled washer element is driven over the
back-up strip. (See Fig. 7.2.2). tapered end of the nut, expanding to its final diameter.
The formed washer is then seated against the joint
surface by the continued advance of the sleeve and
corebolt. This fastener form a flat washer surface
without damaging the mating surface.

/ Sleeve
Fig. 7. 2 . 2 Semi tubular rivet.

Blind rivets
Blind rivets, or fasteners, are designed to be installed
where access to both sides of a sheet assembly or
structure is not possible or practical. The blind rivet
usually consists of a tubular sleeve in which a stem washer

having an enlarged end is installed. The heads of such

rivets are made in standard configurations such as (1) Using NAS 1675-type installation

brazed, universal, and flush. The rivet and stem are

tooling, the nut is restrainedfrom
turning while the corebolt is driven.
inserted in a correct size hole, and the stem is drawn
into the sleeve by means of a special tool. The bulb or
other enlargement on the end of the stem expands the
end of the rivet and locks it into the hole. A typical
blind rivet is the Cherrylock shown in Fig. 7.2.3.
Fig. 7.2.4 illustrates a COMP-TITE blind fastener,
which is a large-bearing blind fastener developed for
use with advanced composite materials. Existing blind
fasteners do not provide sufficient blind head expan­
sion and they may damage the composite structure
during installation. This damage is caused by either
excessive clamp load for the available bearing area or
by radial fastener expansion within the fastener hole,
delaminating the composite plies on the blind-side
(2) The advance of the corebolt forces the
washer and sleeve over the taper,
expanding and uncoiling the washer to
its maximum diameter.

(3) Continued advance of the corebolt

draws the washer and sleeve against the
Rivet joint sutface, preloading the structure.
sleeve
At a torque level controlled by the
break groove, the slabbed portion of
Inse rted Expanded Finished
the corebolt separates, and installation
is complete.

(Installation of Cherry blind rivets)

Fig. 7. 2.4 COMP- TITE blind fastener installation
Fig. 7.2 . 3 Cherry blind rivets and huck blind rivets. sequence.

Airframe Structural Design 2 1 1

Blind Bolts where they can provide adequate strength. The hi­
A blind bolt, like a blind rivet, is one that can be shear rivet is made of steel and employs a swaged
completely installed from only one side of a structure aluminum collar to hold it in place. The collar is
or assembly. The blind bolt is used in place of a blind driven on to the end of the rivet by means of a special
rivet where it is necessary to provide high shear tool in conventional pneumatic rivet gun. A hi-shear
strength. The bolt is usually made of alloy steel, rivet is shown in Fig. 7.2.6.
titanium, or other high-strength material. A typical
blind bolt is shown in Fig. 7.2.5. Hi-Lok fasteners
The Hi-Lok fastener is basically a high strength fas­
Hi-shear fasteners tener which combines the best features of a rivet with
Hi-shear rivets are designed for quick, permanent those of a bolt and nut. It consists of two parts, a
installations where it is desired to reduce weight and threaded pin and a threaded collar. The pin is a
installation time. Such rivets or bolts can be used only straight shank, precision threaded pin which is in-

Sleeve assembly Core bolt

_____£___
M a n d rel
Protrusion External serrations prevent
W orkpiece

E
before rotation of nut
pull-up
Nut portion Indented
of sleeve locking

_:: places)
device
' _

-------1111 �
----------------... N
�·

Lip formed into locking

Internal threads S l eeve groove of sleeve
After pull-up
Before pull-up (Completed installation with corebolt installed)

Sleeve inserted in hole, readyfor pull­ Following the installation of the sleeve,
up by mandrel of installation gun. After the corebolt is inserted into the the sleeve
setting the nut, the mandrel is and torqued to the limits specified on
withdrawn from sleeve. the individual standards.

Fig. 7. 2 . 5 Installation of t h e Hi-Kor blind bolt.

ll=ij
DIIl
8
2 3

��fi:-.1 0 6
Hole prepared. Pin inserted
into hole.
Collar slipped . .
over in end. Hi-shear set and Ri veting starts. Excess collar Groove in pin filled.
~ Collarforming
sucking bar Collar begins material starts Collar trim cut off completed.
readied. to form. to trim. Groove by pin trimming edge. Trim lodged in set
-----'-------' starts to fill
before being ejected.

Fig. 7. 2.6 Hi-shear rivet.

2 1 2 Airframe Structural Design

stalled in a straight walled hole, drilled normally, at Taper-Lok fasteners
0.002 to 0.004 diametral interference. The installation Taper-Lok fastener is an interference-fit, lightweight
of the Hi-Lok fastener is completed on one side of the fastening system with self-sealing feature. The system
assembly after the bolt has been inserted through the comprises a close-tolerance tapered-shank bolt and a
hole from the other side. The hexagonal wrench tip of companion washemut assembly which is both self­
the installing tool is inserted into a recess in the bolt, centering and self-locking. The washernut assembly,
which holds the pin while the tool turns the collar. As as the name implies, consists of a nut with a free­
the collar is tightened to the design torque level built spinning captive washer. The Taper-Lok system and
into the collar, the hex portion of the collar is sheared installation sequence are shown in Fig. 7.2.8.
off automatically by the driving tool. This leaves the
installation with the correct amount of torque and
preload. They are normally used where:
• High shear strength is required, or Close-tolerance Taper-Lok drill
• Impact riveting of solid rivets is prohibited and
they cannot be squeezed, or
• Expansion of the shank would cause undesirable
effects, or
• High clamp-up is desired for sheet pull-up or Integral countersink
faying surface sealing requirements dictate. Radius former
The Hi-Lok fastener is illustrated in Fig. 7.2.7.
Cutting flutes
Hex recess
Hi-Lok adaptor
socket Installation seq uence

1 Special, close-tolerance tapered

drills are usedfor Taper-Lok hole
preparation; drilling, reaming and
countersinking are accomplished in
one operation.
High surface finish

Hex wrench tip

J
Head protrusion ----
Insert the hex wrench tip of the power
The tapered shank bolt is then
T firmly in place by hand pressure.
2
driver into the pin's hex recess.
inserted in the hole, and seated
Headprotrusion (in thousandths
of an inch) above the structural
material, m ultiplied by 0. 048
gives th e interference value
between the bolt and the hole.

Finnly press the power driver against 3 Full contact along the entire shank
the collar, operate the power driver of the bolt and the hole prevents
until the collar's wrenching device has rotation of bolt while tightening the
been torqued off. washernut. During tightening, the
nut spins freely to the locking point,
Installed but the washer remains stationary
col lar and provides a bearing surface
against the structure.

4 Torquing of washernut by
conventional wrenching methods
produces a controlled inteeference
fit, seats the bolt head, and creates
an evenly balanced pre-stress
condition within the bearing area of
Collar wrench ing the structuraljoint.
device ___ _..

This completes the installation ofthe

Hi-Lok fastener assembly.

Fig. 7. 2. 7 Installation of the Hi-Lok fastener. Fig. 7. 2 . 8 Installation of tapered fastener.

Airframe Structural Design 2 1 3

 Fig. 7.2.9 illustrates a LGP fastener which is a 7.3 Bolts and Screws ( Remova ble
specially designed version of the HUCK lightweight Fasteners)
titanium lockbolt system. It is specifically for graphite
composite applications. It is an all titanium system A bolt is an externally threaded fastener designed for
with a Ti-6A l -4V pin and a flanged commercially insertion through holes in assembled parts, and is
pure titanium swage on collar. The fastener comes in normally intended to be tightened or released by
both a pull stem version for installation with common torquing a nut. The aircraft high-strength steel bolts
pull tools or automatic drill/rivet machines such as a and screws are shown in Fig. 7.3 . 1 and high-strength
DRIVMA TIC machine. Another LGP fastener is steel nuts are shown in Fig. 7.3.2.
called LGP sleeved fastener which is designed to be To ensure satisfactory service life of bolts and nuts
installed into interference fits in graphite composite used in primary structure in addition to required
structure without causing any installation damage design strength is the type of thread form and thread
while gaining the improved structural fatigue, elec­ relief, plus material grain structure of thread forms.
trical continuity, and water/fuel tightness with an The basic form of thread used in aircraft design is the
sealant (see Fig. 7.2.1 0). 60° American National Form of thread as specified in

Pin Pin .

Collar

Fig. 7. 3. 1 Standard bolts and screws.

(Courtesy of Standard Pressed Steel Co. )
C o l l ar

Fig. 7. 2 . 9 HUCK- COMP fastener.

(Courtesy of Huck Manufacturing Co . )

Pin Sleeve Pin

Fig. 7. 3.2 Standard nuts.

(Courtesy of Standard Pressed Steel Co.)

MIL-S-77 42 which does not require a specific radius

Col l a r in the root of the thread. In MIL-B-7838 the mini­
mum root radius is specified and thread is rolled after
Fig. 7. 2. 1 0 HUCK-TITE interference fit fastener. heat treat and produces a grain flow in the threads
{Courtesy of Huck Manufacturing Co. )
which is continuous and follows the thread contour.

2 1 4 Airframe Structural Design

 Other rolled thread improvement over the MIL-B- large shear or tension loads from one structural
7838 thread form is the MIL-S-8879 thread form. Its member to another. In designing or strength checking
root radius is a smooth uniform radius which blends a multiple bolt fitting, the question arises as to what
to the thread flanks with an uninterrupted surface and proportion of the total fitting static load does each
is a larger radius than found in any other standard bolt transfer. This distribution could be affected by
form. Fig. 7.3.3. shows the three thread forms and many things such as bolt fit or bolt tightness in the
root stress concentration. Additional fatigue strength hole, bearing deformation or elongation of the bolt
improvements available in the rolled thread fasteners hole, shear deformation of the bolt, tension or com­
pertain to the bolt head to shank fillet radius. This pressive axial deformation of the fitting members and
radius is coldworked to the desired finish requirement the member being connected, and a number or other
after heat treat of the bolt and is free of seams or minor influences. Since aircraft materials such as
inclusions. widely used aluminum alloys have a considerable
degree of ductility, if the fitting is properly designed,
the loads on the bolts approach their maximum value
and will tend to be in proportion to the shear strength
of the bolt. That is, if the shear strength of the bolts is
not the critical strength (bolt shear strength is stronger
than the local material bearing strength), the yielding
of the fitting material in bearing, shear and tension
will tend to equalize the load on the bolts in propor­
tion to their shear strengths. For stresses below the
elastic limit of the fitting material the bolt load dis­
tribution no doubt is more closely proportional to the
bearing area of each bolt; on the other hand, the load
distribution on each bolt is the function of fitting
material modulus, cross-section, fastener pattern
(spacing, rows and diameter), fastener deflections, and
materials (see later section of this chapter). Since the
primary interest is failing strength (based on design
R o l l e d thread g ra i n ultimate loads), the bolt load distribution in propor­
flow tion to the bolt shear strength is usually assumed. Fig.
7.3.4 shows several high strength tension bolts.
Mach i n ed thread Rolled th read Nuts
Fig. 7.3.5 illustrates different standard steel nuts. Nut
material should be more ductile than bolt material;
thus when the nut is tightened the threads will deflect
to seat on the bolt threads. Therefore, it is important
to select the proper nuts to match the tension bolt or,
otherwise, it could decrease the capability of the
tension bolt load. Usually refer to design handbook or
other documents from fastener manufacturers.
Self locking nuts are widely used in the aircraft
industry. The use of the self locking nut reduces
assembly costs as it eliminates the bothersome cotter
pin which takes extra operations and is very difficult
to install on the nut in the many joints and comers of
an aircraft. However, the castellated hexagonal nuts
with cotter pins are mandatory on installations subject
Photo-stress comparison to bolt or nut rotation on any rotational hinge design.
There are two basic types of nuts: tension nuts
such as 1 2 point nuts for tension bolt applications and
Fig. 7. 3.3 Comparison of thread forms and root stress shear nuts for shear bolt applications which are
concentration. primarily loaded in shear.
Aluminum alloy nuts are not used on bolts de­
signed for tension. Bolts used in places where the nut
In addition to standard bolts, screws and nuts, should be aluminum alloy are not allowed. Aluminum
there are many special fasteners that have been de­ nuts are not recommended.
veloped to join parts or structure where the more Since honeycomb panels generally have thin
common fasteners are not usable or do not provide facings, the shear and tension loads in particular
adequate strength. Other special fasteners have been should be transmitted to the entire honeycomb struc­
designed to provide for quick removal of inspection ture wherever possible. For this reason, most fasteners
plates and cowlings. The use of special fasteners is not go through the entire honeycomb or are bonded to
discussed here. the opposite facing and core by means of a potting
adhesive as shown in Fig. 7.3.6. These fasteners will
Bolts and Screws provide the best structural strength since the adhesive
Aircraft bolts are used primarily to transfer relatively bonds the fastener to both facings and core. They add

Airframe Structural Design 2 1 5

weight to the panel and are more time consuming to used in composite structure and areas with tem­
install. peratures above 450°F.
• No threads shall be in bearing in a fitting as shown
Detail design considerations in Fig. 7.3.7.
• Avoid joints incorporating bolts or screws in con­ • Critical applications requiring the control of torque
junction with upset shank rivets or interference fit of bolts iµid screws shall be indicated on the
fasteners. Due to the better fit of the other fast­ drawing such as preloaded tension bolts for
eners, the bolts or screws will not pick up their tension fittings.
share of the load until the other fasteners have • Bolts smaller than 0.25 in diameter shall not be
deflected enough to take up the clearance between used in any single bolted structural connection
the bolt and its hole as previously mentioned. including primary control systems or any applica­
When this type of joint cannot be avoided, the tion where failure would adversely affect safety of
other fasteners should be able to support the entire flight.
shear load. • Shear bolts with thin heads or shear nuts shall not
• Cadmium-plated bolts and screws shall not be be used where bolts are subjected to bending loads.

Companion Temperature

·· MS 2�
Tensile Shear Style and Part N u mber Description Material Locknut Application

44
{ksi) To

1 60 96 1 1 1 1 1- �su20024
I nternal
Wrenc h i n g
A l loy Steel
2 FW
2 FLW
450°F

1 60 96
1 1 1 1 1 1 111!1 • MS 21262 S e l f - L oc k i n g , Cap A l loy Steel -
250°F

1 60 96
1 1 1 1 1 1 111.:1 e �E:::: D r i l l e d H e a d , Cap A l loy Steel 32FL 450°F

563
m11 1 1 1 1 1 1 1 1 111 • 572 4
NAS 42FW
1 60 -
thru H e x , F u l l T h read A l loy Steel 450°F
2 FLW

1223
NAS

1 60 96
1 «1 1 1 1-11 • 1235 NAS
thru
H e x H ea d ,
S e l f Loc k i n g
A l loy Steel -
250°F

1266
NAS

1 60 95 1 1 1 1 1 1-11 • 1270 NAS

thru
NAS
Hex Head
6Al-4V
T i ta n i u m
48FLW
FN T20 500° F

300 1 80 111 1 1 -lli EWB 30 1 2 - Point A l loy Steel FN 30 450°F

300 1 80 111 1 1 _1_! 930

EWB 1 2 - Po i n t A l loy Stee l FN 930 900°F

260 1 56 111 1 1 -- · 26 EWB 1 2 - Po i n t A l loy Steel

FN 26
FNF 26 450°F

260 1 56 111 1 1 -111 . 926 EWB 1 2 - Po i n t A l loy Steel FN 926 900°F

220 1 32 1 1 1 1 _1_ . EWB 22 1 2 - Point A l loy Steel

EWN 22
FNF 22
FNH 22
450°F

220 1 32 111 1 1 _1_ 22

LWB 1 2 - Po i n t A l loy Steel
FN 22
450°F

220 1 32 111 1 1 -1 11 • EWB TM9 1 2 - Po i n t A l loy Steel EWN TM9 900°F

220 1 32 11111111-1 11 • 922 LWB 1 2 - Po i n t A l loy Steel FN 922 900"F

Fig. 7. 3.4 Several tension and shear bolts.

( Courtesy of Standard Pressed Steel C o . )

2 1 6 Airframe Structural Design

 av't/lnd Tensile
Style and Part Na. Description Material Size Range
Equipment (ksil
450"F and 500"F

EWSN 26 Shear Nut Alloy Steel I 1 0 thru 1 .0 inch 200-220

II

• EWN 22 High Tensile Alloy Steel 1/, inch thru 1 '4 inch 200-220

a FN 22 Featherweiaht Alloy Steel I 1 0 lhru 1 '4 inch 200-220

a FNF 22
FNH 22
Featherweiaht Alloy Steel I 1 0 thru 1 11, mch 200-220

a FN 26 Featherweight Alloy Steel I 1 0 thru 1/1 inch 260

• FNF 26 Featherweight Alloy Steel I 1 0 thru 1 inch 260

a FN 30 Featherweight A lloy Steel I 1 0 thru 'le inch 330

750"F and SOO"F

,_,
�""""....
"Eiiif': ..
59FT Thin Height NAS 1022C CRES Steel I 6 thru I '/, inch 60

• 97 lightweight
MS 2104U
NAS 679
CRES Steel I 4 thru 1/i. inch 1 25

-
MS 21046C
Reaular
59 fll Height
A N 363C CRES Steel I 4 thru 1 11. inch 1 25
NAS 1021C

Fig. 7. 3.5 Several standard steel nuts.

(Courtesy of Standard Pressed Steel Co. )

• Bolts of 1 80 ksi heat treat and above shall be Bolt or screw

procured to specifications which meet or exceed
the requirements of MIL-B-88 3 1 ( 1 80 ksi ) . MIL­
B-8906 (220 ksi) and MIL-B-8907 (260 ksi ) .
• Bolts highly loaded in tension shall be assembled
with washers under both bolt head and the nut. Fitting

Potting holes

(Avoid)

Bolt or screw

Honey comb
panel Fitti ng

( G ood)
potting
compound
Fig. 7. 3. 6 Typical insert. Fig. 7. 3. 7 No threads in bearing.

Airframe Structural Design 2 1 7

 which are determined for each part but are the gen­
7 .4 Fastener Selection eral ranges over which the entire aircraft is expected
In making a fastener selection, the designer must write to operate.
down all the conditions to be encountered by the Fig. 7.4.1 shows the fastener characteristics in­
overall design. These are not specific requirements fluencing fastener selection.

Fastener Advantages Limitations

Solid rivets Good clamp-up Non-removable
(Fig. 7 .2. 1 ) High rigidity Limited static shear
Lowest cost Low tension & shear
Lowest weight Low tension fatigue
Good static & fatigue in shear joint High noise level during installation
Lockbolt High clamp-up Non-removable
(Fig. 7.2. 1 0) High shear strength Moderate cost
High reliability Limited tensions
Low noise level during installation Low fatigue,
Moderate weight
Hi-Lok Same as Lockbolt but somewhat lower Same as Lockbolt but somewhat lower
(Fig. 7.2.7) cost and weight cost and weight
Taper-Lok Same as Lockbolt except Same as Lockbolt except
(Fig. 7.2. 8) high joint shear fatigue high cost
capability
High regidity
High shear High shear strength Non-removable
rivets Low cost Low clamp-up
(Fig. 7.2.6) Low weight Low tension
Poor fatigue
High noise level during installation
Blind rivets Locked spindle Low tension & shear
(Fig. 7.2.3) Moderate rigidity Poor fatigue
Low cost Moderate clamp-up
Low weight Moderate reliability
Blind installation Non-removable
Blind bolts Locked spindle Low tension
(Fig. 7.2.5) High shear strength Poor fatigue
Blind installation Moderate clamp-up
Moderate reliability
High cost
Higher weight
Non-removable
12 point High static tension High cost
tension bolt High tension fatigue High weight
(Fig. 7.3.4) High torque No flush surface
High reliability
Removable
Hexagon bolt High tension (long thread) Low tension (short thread)
High static shear Low fatigue
Low cost Moderate weight
Removable Moderate torque
No flush surface
1 00° Flush bolt High static shear Low tension
Low weight Low fatigue
Removable Low torque
Flush surface Moderate cost
Fig. 7. 4. 1 Fastener characteristics comparison.

2 1 8 Airframe Structural Design

 After the designer determines the type of joint to tact with any of the many fluids used in aircraft.
be used, he then determines whether the fastener is • Availability - Is there more than one source for
loaded axially or in shear. If tension fatigue is the the fastener? Competition is good because of its
principal load condition, items such as thread form influence on price and delivery schedules. Procure­
rolled threads (see Fig. 7.3.3) and rolled head-to­ ment costs are also important in fastener selection.
shank fillets with special design tension heads are • Storage - Do the manufacture already have stocks
used. of needed fasteners in the stores? Shelf life may
For static tension loading, the same enlarged root also be a problem; for example, 2024-T3 1 (DD)
radius threads are not essential. For joints with rivets have a short shelf life after removal from the
incidental tension loadings, nearly any fastener can be refrigerator.
considered, including rivets, blind rivets, and shear • Fastener installation - It should be considered
head fasteners. whether there is equipment for installation, in
Shear loading requires consideration of fasteners plant, availability for purchase, or whether it must
such as bolts or swaged collar fasteners with conven­ be designed. Is the installation manual or auto­
tional or shear heads. Shear heads offer the higher matic? How is reliability, fastener-fit, applied
strength-weight ratios. For shear fatigue loadings, the torque on a nut and bolt combination, etc?
designer selects fasteners capable of interference fit or In the design process, the designer considers the
of developing high residual compression stresses in relationship between cost, weight, and function in
the structural material around the holes. selecting from the previously chosen fasteners, the
The process of fastener selection is usually accom­ optimum combination of these factors should be care­
plished in the following steps: fully considered. In some military applications,
• Static strength is determined by material, diameter, ground support equipment and facilities may be
head size, and thread size and length. required. In such cases, great consideration to cost
• Corrosion resistance - Fastener materials with and little or no attention to weight is required. When
good corrosion resistance are not acceptable for removeability is required, the design case limits the
many applications because of high cost or galvanic choice to bolts or screws. Inaccessibility, when com­
corrosion they may cause in the dissimilar metals plete, requires the use of blind fasteners unless pre­
used in structure. Cadmium plating is the most viously installed nut plates can be used with bolts or
commonly used finish on steel fasteners, but it screws.
leaves much to be desired. In blind rivets, alloy 7.5 L u g Desig n and Analysis
steel stems have been replaced by corrosion resist­
ing steel stems to avoid unsightly rust or expensive (Shear Type)
paint touch-up. A method of analysis (Ref. 7 .5) for the design of lugs
• Material compatibility - Fastener materials, made of aluminum or steel alloys, with static load
finishes, and lubricants must be sel_ected so as to be applied axially, transversely, or obliquely (as shown in
compatible with other materials which they contact Fig. 7 .5 . 1 ) is presented herein.
and with the functions and life of the assembly or
equipment in which they are installed.
• Fatigue (Tension) - Tension fatigue failures are
complex problems; therefore, very careful selection p
of materials, configuration, manufacturing pro­
cesses, finishes, lubricants, mating parts, and instal­
lation procedures are required. Vacuum melt mate­
rial is generally used to improve life of fasteners.
Threads per MIL-S-8879 (with enlarged root Axial
radius) reduce stress concentrations and improve
fatigue life. Rolling of threads after heat treatment
and cold working of head-to-shank fillet induce
compressive stresses which prolong fatigue life.
• Fatigue (Shear) - Many fasteners are used in
applications where the primary loading is shear;
however, fastener fatigue failures are induced by
secondary loads in tension or bending. In such
cases, many shear fasteners are used in tight or Transverse
interference-fit holes.
• Elevated temperature strength is required and this
p

is a function of the material as to strength at

temperature and strength after exposure to tem­
perature.
• Magnetic permeability - In or near certain types
of equipment (compasses or directional control
devices), magnetic permeability of fasteners is a
critical characteristic.
• Fluid compatibility - The finish or lubricant on a
fastener or the fastener material must be carefully Obliq ue
considered when the fastener will come into con- Fig. 7. 5. 1 Three different applied load cases on lug.

Airframe Structural Design 2 1 9

 tio !!. also affects the shear-bearing efficiency

-Da ,
t
D
br • but only for very large values of - or
t
sually be ignored.
')Wable ultimate load for tension failures
(7.5.2)
Solid rivets
(Fig. 7.2. 1 )

Net section
·ension failure

S hear
tea rout

r ig. 7. 5. 2 Lug net section tension and shear tearout

failure under axial load.
D

I
3.0
.,,,,,,.
.,,.,..,.,..

�v ,_..,--
2.8 .3
.,,,...- �
...,,,...,,.

�v
2.6 4
:,...-- �
.---

� ��
...... ..,.....,,... - --
..,.....,,...
2.4
i--- 5

�
--

J��
�
2.2 � -- j-

Cunie @ is a cutoff to be usedfor all -

. - 6

� �--
2.0 Loa
- aluminum alloy handforged billets when
the long transverse grain direction has -::: 7

Fa � 11 ��V'" -- --
1 .8
the general direction G in the sketch.

� v:
•

Curve @ is a cutoff to be used/or all

� �� � -
I � - 8
1 .6 - D
Kh, T1 aluminum alloy plates, bars, and hand

A� ::::--�
1 .4 forged billets when the short transverse = 9
....- grain direction has the general - 10

)e. � - - -"®
I I direction G in the sketch, andfor die -
1 .2 - .I forgings when the lug contains the
i- - - parting plane in a direction

I/'
1.0 approximately normal to the direction -

@
o/G.
- 15

/
0.8 -
:,.......-- provided by Curves @ and
-- Note: In addition to the limit ns
, in no _

I / :,,...-- __,
� 20
0.6 .
event shall a K,,, greater than 2.00 be -

-
usedfor lugs made from 0.5 in. thick or - 25

% :::::::-
I
I - 30

u�
0.4 thicker aluminum alloy plate, bar, or
handforged billet.

' �
0.2

0
0.6 0.8 1 .0 1 .2 1 .4 1 .6 1 .8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
a
D
Fig. 7. 5. 3 Shear-bearing efficiency factor, Kb, ·

220 Airframe Structural Design

where Fryx Tensile yield stress of lug material
Allowable ultimate load for tension across grain

=
P,u
failure Ultimate tensile strength of lug material

=
F,ux
K, Efficiency factor for tension, see Fig. across grain
7.5.4. (Pu ) min The smaller of Pbru or P,u

=
A, Minimum net section for tension, A, Determine allowable yield bearing load on bushing,
( W - D) t
F,u Ultimate tensile strength of lug material, (7.5 .4)
refer to Ref. 7 .6
Yield axial load attributable to shear-bearing is given
by Allowance yield bearing load on bushing
Compressive yield stress of bushing
material
(7.5.3) The smaller of the bearing areas of
bushing on pin or bushing on lug. (The
where latter may be the smaller as a result of
�
Allowable yield load on lug external chamfer on the bushing).
Yield factor, see Fig. 7.5.5. Finally, calculate margins of safety.

Legend L, LT and ST indicate grain in direction F

in sketch:
Aluminum designation:
14S-2014
24S-2024
75S- 7075

Cune G>
4130 steel
14S-T6 and 75S. T6 plate � 0.5 in (L, LT)
75S- T6 bar and extrusion (L)
14S-T6 handforged billet � 144 sq in (L)
14S-T6 and 75S-T6 die forgings (L)

Cune ®
14S-T6 and 75S-T6plate > O.S in , � Jin
75S- T6 extrusion (LT, ST)
75S- T6 handforged billet � 36 sq in (L)
14S-T6 handforged billet > 144 sq in (L)
14S-T6 handforged billet � 36 sq in (LT)
14S-T6 and 75S-T6 dieforgings (L T)

Cune ®
24S-T6plate (L, L T)
14S-T4 and 24S-T42 extrusion (L, L T)

Cune ©
0.3 1----1 24S-T4 plate (L, L T)
24S-T3 plate (L, L T)
14S-T6 and 75S-T6 plate > J in (L, LT)
24S-T4 bar (L, LT)
0.2 75S- T6 handforged billet > 36 sq in (L)
75S-T6 handforged billet � 16 sq in (L T)

0.1 Cune @
75S- T6 handforged billet > 16 sq in (LT)
J4S- T6 handforged billet > 36 sq in (LT)

0 L..�..J..�--l��.L-�-L.�---"L-�...L..�-.L�� Cune @
1 .0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 Aluminum alloy plate, bar, handforged billet, and die
forging ( ST)
w
75S-T6 bar (L T)
D
Curve ([)
18-8 stainless stee� annealed

Curllt! ®
1 1 3
18-8 stainless steel, full hard. Note: for 4 , 2 and 4 hard,

interpolate between Cunes <J> and ®.

Fig. 7. 5.4 Efficiency factor for tension, K1 •

Airframe Structural Design 22 1

 A tension efficiency factor is introduced to account The ratio !!. also affects the shear-bearing efficiency
for stress concentration effects, and curves are given t
for predicting such failures in lugs of steel and alumi­ D a
factor Kb, , but only for very large values of - or - ,
num alloys. t D
The method, as described in this Section, had a and can usually be ignored.
theoretical basis, and its validity has been verified by The allowable ultimate load for tension failures
comparison with many test results. This method also
contains the evaluation of pin bending moments. (7.5.2)
Lugs Loaded Axially - Case I
Three modes of failure: (a) net section tension; (b)
shear tearout, assuming all the load to be transmitted
on "40 degree planes"; and (c) bearing as shown in Net section
Fig. 7.5.2. It is indicated that modes (b) and (c) could tension failure
be regarded as a single mode of failure and that for
this mode, the allowable ultimate load for shear-bear­
ing failure is

where:
Pbru = Kb,A b,Ftu (7.5 . 1 )

Pbru = Allowable ultimate load for shear-bearing

T
failure w

j_
Kb, Shear-bearing efficiency factor, see Fig. Shear
7.5.3 tea rout

Ab, Projected bearing area (A b, = Dt)

F,. Ultimate tensile strength of lug material
with grain (refer to Ref. 7.6)
Over a range of lug shapes covering all lugs
a Fig. 7. 5. 2 Lug net section tension and shear tearout
found
. in practice, Kb, depends only on the ratio - . failure under axial load.
D

3.0

2.8

� vi.--
2.6 4
��
�

�� i:-
2.4
5
.l-- ...... - � � �
�v-
�__,____.__.__,, ...
2.2
�1
.__ ,___
6
Curve @
_
_ i.- =-i- _

2.0 Load --" is a cutoj/to be usedfor all -

- l�J....oo"""' L..-- aluminum alloy handforged billets when
��
� V' �L-- � the general direction G in the sketch.
1 .8 the long transverse grain direction has -- 7
...j t.
__..

: F '* lo':� � '<,

8

Curve @J is a cutoff to be used/or all

� �C:::-
1 .6 - B D
K.,
0 ,.....
aluminum alloy plates, bars, and hand
1---,.----,.----.--'
J /.
.:, �-±=4---t=� forged billets when the short transverse 9

��--
1 .4 w��-�-�--1-.,..u � �,,�
�,,,,. =
---- grain direction has the general -
10
1 .2 V'"'1 , direction G in the sketch, and/or die _

� _.. � forgings when the lug contains the

�- - - - - --•-.[b parting plane in a direction

,
� -
1 .0 approximately normal to the direction

Note: In addition to � limita/i:ns

o/G.
- 15

�I-drl"'f!.�,=--f_..
0.8

0.6 J-l-__jV =�-4==J:=:::j:=:::j:==t event shall a greater than be -

-
�1 provided by Curves 0J and � , in no 20
�� ___.. ..,._-- _
__ Kb, 2.00
I used/or lugs madefrom in. thick or - 0.5 25
i 1�
_- _..
I 1

/� �c:;;..-
0.4 J'. thicker aluminum alloy plate, bar, or -
30

,�
,__
'
handforged bil1et.
0.2 --

0 �L-__,....___.___.___.___..__.__.__._-1.._-1.._-1..__.___.___.__...._....
.. _ .._ J
0.6 0.8 1.0 1 .2 1 .4 1 .6 1 .8 2.0 2.2 2.4 2.6 2.8 3.0 3.2 3.4 3.6 3.8 4.0
a
D
Fig. 7. 5. 3 Shear-bearing efficiency factor, Kbr ·

220 Airframe Structural Design

where Fiyx Tensile yield stress of lug material
Allowable ultimate load for tension across grain

=
P,u
failure Ultimate tensile strength of lug material

=
F,ux
K, Efficiency factor for tension, see Fig. across grain
7.5.4. ( Pu ) min The smaller of Pbru or P,u

=
A, Minimum net section for tension, A, Determine allowable yield bearing load on bushing,
( W - D) t
F,u Ultimate tensile strength of lug material, (7.5 .4)
refer to Ref. 7 .6
Yield axial load attributable to shear-bearing is given
by Allowance yield bearing load on bushing
Compressive yield stress of bushing
material
(7.5.3) The smaller of the bearing areas of
bushing on pin or bushing on lug. (The
where latter may be the smaller as a result of
�
Allowable yield load on lug external chamfer on the bushing).
Yield factor, see Fig. 7.5.5. Finally, calculate margins of safety.

Legend L, LT and ST indicate grain in direction F

in sketch:
Aluminum designation:
14S-2014
245-2024
75S-7075

Curve <D
4130 steel
14S-T6 and 75S. T6 plate :;. 0.5 in (L, LT)
75S-T6 bar and extrusion (L)
14S- T6 handforged billet :>. 144 sq in (L)
14S-T6 and 75S-T6 die forgings (L)

Cut11e ®
14S-T6 and 75S·T6plate > 0.5 in , :>. lin
75S-T6 extrusion (LT, ST)
75S· T6 handforged billet :>. 36 sq in (L)
14S- T6 handforged billet > 144 sq in (L)
14S-T6 handforged billet :>. 36 sq in (LT)
14S-T6 and 75S·T6 dieforgings (LT)

Curve @
24S-T6 plate (L, L T)
0 .4 1--.+-_JL- _ _.L__.i_�rr---;--+�:-1 24S-T4 and 24S·T42 extrusion (L, LT)
I

Curve ©
0.3 24S-T4plate (L, LT)
24S· T3 plate (L, LT)
14S-T6 and 75S-T6 plate > 1 in (L, LT)
24S-T4 bar (L, LT)
0.2 75S· T6 handforged billet > 36 sq in (L)
75S-T6 handforged billet :>. 16 sq in (LT)

0:'1 Curve @
75S-T6 handforged billet > 16 sq in (LT)
14S-T6 handforged billet > 36 sq in (LT)
0 L.. � ..l- -1..
� -L
� � --1��.L-�-'-�-1- �
-:"" Curve ®
1 .0 1 .5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
Aluminum alloy plate, bar, handforged billet, and die
forging (ST)
w
75S-T6 bar (LT)
D
Curve (!)
18-IJ stainless steel, annealed
Cur11e ®
1 1 3
18-IJ stainless steel, full hard. Note: for 4 2 and 4 hard,
,

interpolate between Curves (!) and®.

Fig. 7. 5. 4 Efficiency factor for tension, K, .

Airframe Structural Design 22 1

 ""' '\.
1.1

1 .0

""'�
""'
0.9

0.8

�
c

0.7

0 .6

0.5
0 0.5 1 .0 1.5 2.0 2.5 3.0

Fig. 7. 5. 5 Yield factor, C. Fig. 7. 5. 7 Lug shapes.

Lug Loaded Transversely - Case II changes of cross section), an equivalent lug

( 1 ) Obtain the areas A 1 , A2, A3 and A 4 as follows: should be sketched as shown in Fig. 7.5.8
(a) A 1 , A2 and A 4 are measured on planes indi­ and used in the analysis.
cated in Fig. 7.5 .6(a) perpendicular to the ax­ Obtain the average of these areas using
ial centerline, except that in a necked lug A 1
and A 4 as shown in Fig. 7.5 .6(b) should be
measured perpendicular to the local center­
line. 6

A3 least area of any

( 1 ) ( 1J ( 1J ( 1J
1 + + +

(7.5.5)
radial section
I
I (d) Compute:
,
I
and
A b,
(e) Determine allowable ultimate load

==
(7.5.6)
where
P,,. Allowable ultimate load as deter­
mined for transverse loading

==
p
�) (b) K,,. Efficiency factor for transverse
load (ultimate), (see Fig. 7.5 .9)
A b, Projected bearing area
Fig. 7. 5. 6 Locations of cross-sectional areas of A1, A2, F,ux Ultimate tensile strength of lug
A3 and A4. material across grain.
In no case should the allowable transverse
load be taken as less than that which could
(b) A 3 is the least area on any radial section be carried by cantilever beam action of the
around the hole. portion of the lug under the load (see Fig.
(c) Since the choice of areas and the method of 7.5. 1 0). The load that can be carried by can­
averaging has been substantiated only for tilever beam action is indicated very approxi­
lugs of the shapes, as shown in Fig. 7.5 .7, mately by curve A in Fig. 7.5.9; should K,,.
thought should always be given to assure that be below curve A, a separate calculation as a
the areas A 1 , A2, A 3 and A4 adequately ref­ cantilever beam is warranted.
lect the strength of the lug. For lugs of un­ (f) Determine allowable yield load of lug
usual shapes (for example, with sudden (7.5.7)

222 Airframe Structural Design

 (a) (b)

Fig. 7. 5. 8 Equivalent lugs of unusual shapes.

Curve <D
14S-T6 and 75S-T6plate :f, 0.5 in

Curve ®
14S-T6 and 75S-T6 plate > 0.5 in , :;;, 1 in
75S- T6 extrusion
1.7
1 1 Hr - � /
14S-T6 handforged billet :fit 36 sq in
14S-T6 and 75S-T6 die forgings 1 50
1 .6 --- 4 1 30 Steel
I/ �
Curve @
24S-T6 plate H.T. 1 25 ksi
=

1.5
:/
1'

-
24S-T4 and 24S-T42 extrusion

1 .4 Curve ©
14S-T6 and 75S-T6plate > 1 in
/K,,, for all
75S-T6 handforged billet � 16 sq in.
aluminum and steel

L ksi )v
1.3
Curve @

i /
75S-T6 handforged billet > 16 sq in
1 .2 �.r. - 1
�
14S-T6 handforged billet > 36 sq in "'"'

1.1
� v- ../ v
Note:

-
These curves do not apply to aluminum alloy v ...-- ......

l/
lugs having the short grain direction (ST)
1.0
t· ,,<

v v I\
:.:: ....in the plane ofthe lug. v '

-
�
0:
0.9
v vv
<:s Above aluminum designations: J
"

o:5 in -1/
14S - 2014 \
:.::
24S - 2024 2024-T3-T4 plate
v
0.8
/J v/
75S - 7075 t .;;

0.7 - /

I 0� --
CD
--/
0.6
.-- \
�/ ,
-

I �
-
�
�
�

0.5
l0�
I
/ �
.,.--
\
" 2024-TJ or -T4 /
--e
-

./ -- 356-T6

I� y /
0.4 /
plate > 0. 5 iri_A
0
2 2 4-T4 bar ©
0.3 /

��
/ "
/
I/ Curve A- approximate
/
-
cantilever strength,
0.2
� ,,.
/ if K,,. is below this curve, -
se� Fig. 7 j 5.10. 1
0.1
/
.... I
--
---
-
-

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 .0 1.1 1 .2 1 .3 1 .4

Fig. 7. 5. 9 Efficiency factor for transverse load, Kw

Airframe Structural Design 223

 === =
where load) divided by [ Py (Eq. (7.5.3)J

l
P Allowable yield load on lug R,, [Transverse component of applied
K,ry Efficiency factor for transverse yield load divided by [ PY Eq. (7.5.7))
A b,
Fryx =load (yield), see Fig. 7.5.9
Projected bearing area
Tensile yield stress of lug material
across grain
(g) Determine allowable yield bearing load on
(b) Determine allowab e yield bearing load Pbry on
bushing by Eq. (7.5.4).
Pin Design
In general, static tests of single bolt fittings will not
bushing by using Eq. (7.5 . 4 ) . show a failure due to bolt bending failure. However, it
(h) Calculate margins of safety. is important that sufficient bending strength be pro­
vided to prevent permanent bending deformation of
the fitting bolt under the limit loads so that bolts can
be readily removed in maintenance operations.
Furthermore, bolt bending weakness can cause
peaking up a non-uniform bearing loads on the lugs,
thus influencing the lug tension and shear strength.
The unknown factor in bolt bending is the true value
of the bending moment on the bolt because the
moment arm to the resultant beaqng forces is difficult
to define. An approximate method for determining
the arm to use in calculating the bending moment on
bolt is given below:

=
moment arm, b 0.5 t1 + 0.25 t2 + b
= (7.5 .9)

=
If it is no gap ( b 0) between the lugs, the moment
Fig. 7. 5. 1 0 Cantilever beam action of the portion of the arm is taken as
lug under load. (7.5 . 1 0)
b 0.5 t1 + 0.25 t2
If it is desired to take account of the redllction of
Lug Loaded Obliquely - Case Ill pin bending that results from peaking as shown in Fig.
(a) Resolve the applied load into axial and transverse 7.5 . 1 1 , the moment arm is obtained as follows:
components and obtain the lug ultimate and yield (a) Compute the inner lug r - use Db instead of D,
margins of safety from the interaction equation: if bushing is used

where
R a

for ultimate load,

1
- ( 1 .6 + 1 .6)0.62 s - 1
M. s. -
Rtr
(7.5 .8)
j(�) - i] n f2
[Axial component of applied ultimate (b) Take smaller of Pbru and P,u for inner lug as

=
R.
IoadJ divided by [Smaller of Pbru (Eq. ( [Link]
(7.5 . l ) and P,u (Eq. (7.5 .2)) ( Pu)min and compute -
A
-F.-
b, 1u.r
R,, (Transverse component of applied
ultimate loadJ divided by [ P, (Eq. · ( PJmin . th e
and r to ob tam
(c) Enter Fig. 7 . 5 .1 2 with A

=
·

br F,ux
ru

(7.5 .6))
--

for yield load, reduction factor for peaking, and calculate tl\e
R. [Axial component of applied yield moment arm

Load from pin peaks t,

near shear planes 2
--.
t,
J.
p
-

2
p
-·

�
t,
p -4--'"----
2 --1.

Fig. 7.5. 1 1 Pin moment arm for determ ination of bending moment .

224 Airframe Structural Design

 . MD ' .
resu 1ts from M, assummg an d 1stn'b ut1on (where
( 7.5 . 1 1 ) U
I = moment of inertia of pin) and its margin of safety
The maximum bending moment in pin from the as
equation as

( �)
b F
M.S. = --- -1
M =
P u ( 7.5 . 1 2 ) MD
21

and calculate the bending stress in the pin that where Fb =

the modulus of rupture deter-

o . �-i; . .-,
1 .0

�o�
.�
.- --
0.9

/O· o- ....- --
--- - �
:__.. � .- �---

I' I/ j ,.o ':' o'.'" _..,."C,.. _,,,- _.,,-- i.,...-- �i-- t:::: :::::::
I ,o,� ,,.
0.8 ov. �
-- -- - ,.

o ·� o '"" '?lo ,.;I�

::;;;;;"

J / _,- � ....,... r:,.....- :;::::;;. �

-
0.7
� �
-
,.

/ V V o�
_ ..
-
/�

I I I / / / / ,.,0 ·o .&."o
0.6 ti · �
�
-
i.--"": ,.,,,.
/ ! / j//�/� �� � o�� ,
' o.0- i,..---
I 1 i j �.,. -'

0.4 : ;
0.5 -- - - - --
-
--r
)' . ..... D

I / r,
II"' / �:.. �
- · - · -·
,
_i_ D
II I
0.3 f tf
i--
' I a - - - ---

�--·- -·� :
a- -
L
I '� v
0.2 : ' I I I
I � . -,- 2

, _ (D/2) _ ( !)
' , , 1--
I .
t

',J,
I
I>"'
.
a- _! !! -
I I _
J
t D 2 t

>/
0. 1 I I

0
0 0. 1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 .0 1.1 1 .2 1 .3 1 .4 1 .5 1 .cl 1 .7 1 .8 1 .9 2.0

Fig. 7. 5. 1 2 Peaking factors for pin bending (dash lines indicate region where these theoretical curves are not
substantiated by test data).

Bolt Dia. Bolt Area Moment of Single Shear Tension Bending

Inertia
(in) (in 2) (in 4) (lb) (lb) (in-lb)
1/4 0.049 1 0.000 1 92 3680 4080 276
5/ 1 6 0.0767 0.000468 5750 6500 540
3/8 0. 1 1 05 0.00097 8280 1 0 1 00 932
7/ 1 6 0 . 1 503 0.00 1 797 1 1 250 1 3600 1 480
1/2 0. 1 963 0.00307 1 4700 1 85 00 22 1 0
9/ 1 6 0.2485 0.0049 1 8700 23600 3 1 40
5/8 0.3068 0.00749 23000 30 1 00 4320
314 0.44 1 8 0.0 1 5 5 3 33 1 50 44000 7450
7/8 0.60 1 3 0.02878 45000 60000 1 1 850
1 .0 0.7854 0.049 1 58900 80700 1 7670

Note: AN steel bolts, F, u = 1 2 5 ksi, F, u = 75 ksi

Fig. 7.5. 1 3 A N steel bolts allowable strength (F1• - 1 2 5 ksi).

Airframe Structural Design 225

mined from Ref. 7 .6 or other sources Example
or Given: Materials -
Ultimate bending moment (allowable) Lug: Aluminum extrusion 7075-T6 (see Fig. 4.3.6)
MS = 1 Bushing: Steel with H.T. - 1 50- 1 80 ksi
M Pin: Steel with min. H.T. 200 ksi==

The ultimate allowable bending moment for a = 1 .875 in

F,. = 1 25 ksi AN steel bolt can be obtained from W - 3.75 in
Fig. 7.5 . 1 3 . Db - 1 .25 in
t (or '2) = 0.75 in
Bushing and Bearing t , - 0.375 in
It is customary to provide bushings in lugs with single Pin diameter = D 0.75 in
bolts or pin fittings subjected to reversal of stresses or
==

to slight rotation. Thus, when wear and tear takes (L) g rain direction
place, a new bushing can be inserted in the lug fitting.
Steel bushings are commonly used in aluminum alloy

T
single bolt fitting lugs to increase the allowab�e bear­
ing stress on the lug since the bushing increases the
bearing diameter. If bushings are not used on single
bolt connections, sufficient edge distance should be W - 3.75 Bus hi ng

1
provided ·to ream hole for the next size bolt in case of P - 75 kip
excessive wear of the unbushed hole. If considerable
rotation occurs a lubricator should be provided for a
plain bushing or an oil impregnated bushing should be
used.
Plain spherical bearings usually consist of two
pieces: an outer race with a spherical inside diameter
swaged over a spherical inner ball, as shown in Fig.
7.5. 14. Joint rotation takes place between these two
pieces with a ball-and-socket feature providing a small
amount of unrestricted alignment.

Pin

01 (a) Compute:
!._ .... 1 .875
Db 1 .2 5
w 3.75
- 1 .5

S pherical = -= 3.0
ball --..m.,..,.,..,.,.,.-r7r.t Db 1 .25
!!b_ 1 .25 = 1 .67
.______ Lug t 0.15
Ab, == Db t - 1 .25 (0.7 5) -= 0.938
A, - ( W - Db ) t - (3.75 - 1 .25) 0.75 - 1 .875
Fig. 7. 5. 14 Plain spherical bearing.

(b) Allowable shear-bearing failure load

The advantage of a plain spherical bearing over an [Ref. Eq. (7.5.l)J.
anti-friction bearing lies in the much greater load
carrying capability of its area contact as opposed to Enter Fig. 7 .5.3 with � - 1 .5 ; 0_ 1 .67,
the point contact or line contact of the ball. Disadvan­ Db t
tages of the plain spherical bearings include: obtain Kb, 1 .26==

• Very much higher friction Pbru-= Kbr (Abr) (F,uz) = 1 .26 (0.938) (74 )
• The necessity of frequent lubrication ... 87.5 kip
Poor reliability compared to plain bushing
(c) Allowable tension failure load: [Ref. Eq. (7.5.2)J
•
• Usually higher cost.
Because of these factors, mechanism designs Enter Fig. 7.5.4 (curve <D) with _! _ 3.0,
should specify the plain spherical bearing for joints Db
where the load is high (such as powered controls, obtain K, ""' 0.922
landing gear joints, etc.); anti-friction bearings should
be used elsewhere. Highly loaded joints where the P,. == K, (A, ) (F,. ) ... 0.922 ( 1 .875) (8 1 )
= 1 40.03 kip
aligning feature of spherical bearings is not required
should utilize plain bushings. (d) Allowable yield load: (Ref. Eq. (7.5 .3)J.

226 Airframe Structural Design

 1. ( Pu )min = Pbru = 87.5 kip 1 40.03
Enter Fig. 7.5.5 with 1 = 0.62
. 1 . 1 5 (75)
( P,.) min 87. 5 . Yield margin M.S. = -1
= l . I 5. obtam C = 1 . 1 p
Abr ( F,ux l 0.938 (8 l) limit load
!!E. ( Pu)min -1
85.6
2. Py - C F. 75
= 0.49
'"" l . 1 5 <u )
(72) (87. 5)
I.I
= = 85.6 ki p
81 2. Pin:
(e) Allowable yield bearing load on bushing: Ultimate margin in bending
(Ref. B.q . (7.5.4)J
Pbry = 1 .85 Fey x A brb M.S. = _!l_ 1
l.15P
_

= 1 .85 ( 1 45) (0.563) = 1 5 1 .02 kip 1 27. 1 6

where Abrb = Dt = 0.75 (0.75) = 0.563 in 2 = - 1 = 0.47
1 . 1 5 (75)
Fey = 1 45 ksi (From Ref. 7.6)
Ultimate margin in shear
(f) Allowable pin bending load (Ref. Eq. (7.5 .9)J
Assume there is no gap between t1 and t2 , i::
M.S. = -·- - 1
therefore 6 = 0 l . 1 5P
b = 0.5 t1 + 0.25 t2 + 6 1 1 0. 5
1 = 0 .28
= 0.5 (0.375) + 0.25 (0.75) + 0 = 0.375 1 . 1 5 (7 5 )

pin moment, M = ( Pu )min (�) 3. Bushing:

Yield margin M.S. = P,,,y -1

( �7 )
limit load
o. 5
= 87 . 5 = 1 6.4 in-kip 1 5 1 .02
--- - 1 = 2 .02
75
From Fig. 7.5 . 1 3, the allowable ultimate bending 1 .5
moment is 7.45 in-kip (D = 0.75 in), and for H.T.
= 200 ksi steel bolt and allowable pin bending Conclusion: The least margin of safety as shown

( ��� )
loa
�:
above is 0.27, which is greater than requirement of a
- 2 (7.45) - 23.84 in-kip minimum margin of safety of 0.20 (recommended).

( �� � ) 7.6 Welded and Adhesive Bonding

- 4
Pb = ( P. }min (: ) = 87 .5 Joi nts
.
Welded Joint
= 1 27.2 kip Since the overall structure of an airplane cannot be
(�) Allowable pin shear load fabricated as a single continuous unit, many struc­
tural parts must be fastened together. For certain
From Fig. 7.5 . 1 3, the allowable ultimate shear materials and types of structural units, welding plays
load is 33. 1 5 kip (D = 0.7 5 in), and for H.T. = an important role in joining or connecting structural
200 ksi (Fsu = 1 2 5 ksi, from Ref. 7.6) steel bolt units. Research is constantly going on to develop
and total allowable pin shear load is better welding machines and welding techniques and

P, = 2 (33. 1 5) ( \; )2
= 1 1 0.5 kip
also to develop new materials that can be welded
without producing a detrimental strength influence on
the base or unwelded material.
Welding is used extensively for steel tube truss
(h) Margins of safety structures, such as engine mounts, general aviation
1 . Lug: airplane fuselages, and for steel landing gears and
fittings. The strength of welded joints depends greatly
Ultimate margin in shear bearing on the skill of the welder. The stress conditions are
usually uncertain, and it is customary to design
M.S. =
Pbru
-1 welded joints with liberal margins of safety. It is
l.15P preferable to design joints so that the weld is in shear
87.5 or compression rather than tension, but it is frequent­
- 1 - O.O I ly necessary to have welds in tension. Tubes in tension
1 .1 5 (75)
are usually spliced by "fish-mouth" joints, as shown in
Ultimate margin in tension Fig. 7.6. 1 , which are designed so that most of the weld
is in shear and so that the local heating of the tube at
P,.
M.S. = -- - 1 the weld is not confined to one cross section.
l.15P To weld or use fasteners: here are a few points to

Airframe Structural Design 227

consider when deciding whether welding or a me· without using the 1 0% reduction.
chanical fastener would be best for a particular When the structural engineer considers the possi­
design. bility of welding or brazing as a method to satisfy the
• Is the joint to be permanent or temporary? If it is joining requirements of a aircraft, several questions
temporary, welding is obviously out of the ques­ immediately arise:
tion. • Are the alloy compositions, conditions, and forms
• Is air tightness or leak tightness required? If so, being used readily weldable?
welding may be the easiest and least expensive way • To what degree is the design depending on the de­
to achieve this. velopment of the full design strength in the joint?
• Is access to both sides of the part available? This is • To what degree are the joints repetitious, and in
often essential with mechanical connection, but not what total quantity?
generally required with welding. • Can heat treatment and/or machining be per­
• What equipment is needed? What does it cost? formed on the welded assembly?
Sometimes expensive drilling and assembly systems • Can the weld be formed manually, by machine with
are needed to give a bolted or riveted part the operator, or automatically?
same production rate as welding. This is especially The designer of welded structures in steel can
true if there are long seams, or if many holes must greatly help the welder obtain good joints or connec­
be drilled. tions by adhering to the following general rules:
• Is it practical to design the joints with overlapping • It is much easier to obtain a good weld when the
edges? This is necessary with bolts and rivets, but parts being welded together are of equal thickness.
not usually essential with welding. It is general design practice to try and keep thick­
• Does available labor have the necessary skills to ness ratio between the two welded parts less than 3
use the assembly method properly? Some welding to 1 . Some designers try to keep within a 2 to 1 ra­
methods require highly trained personnel. tio in order to eliminate possibilities of welders
Tests show that plain carbon and chromemolyb­ burning the thinner sheet.
denum steels suffer very little in loss of tensile • Steel tubes often have walls as thin as 0.035 in,
strength due to welding. For cold rolled sheet or and the welder must control the temperature to
tubing the refinement in grain due to cold working is keep from overheating the thin walls and burning
lost in the material adjacent to the weld which lowers holes in them.
the strength to a small degree. Welding, however, does • A weld should not encircle a tube in a plane
produce a more brittle material which has lower perpendicular to the tube length.
resistance to shock, vibration and reversal of stress. • In general, avoid welds in tension since they
Thus, it is customary to assume an efficiency of weld produce a weakening effect. In some connections it
joints less than 1 00%. Ref. 7.6 gives the allowable is difficult to avoid all tension loads on welds; thus
ultimate tensile stress for alloy steels for materials weld stresses should be kept low and if possible
adjacent to the weld when the structure is welded incorporate a fishmouth joint or finger patch to put
after heat treatment. For welding members subjected part of weld in shear.
to bending, the allowable modulus of rupture for alloy • Tapered gusset plates should be incorporated in all
steels when welded after heat-treatment should not important welded joints to insure gradual change in
exceed that as specified in Ref. 7 .6. For weldable stress intensity in members. These gussets lessen
materials heat-treated after welding, the allowable the danger of fatigue failure by reducing stress in­
stresses in the parent material near a welded joint may tensity.
equal the allowable stress for the heat-treated mate­ Welding consists of joining two or more pieces of
rial; in other words, no reduction for welding. How­ metal by applying heat, pressure, electron beam, etc.,
ever, it is good design practice to be conservative on with or without filler material, to produce a localized
welded joints; thus a reduction of 1 0% of the heat­ union through fusion or recrystallization across the
treated properties is often used in calculating the joint interface. Examples of common welding pro­
tensile or bending strength in the member adjacent to cesses include: fusion, resistance, flash, pressure, and
the weld, or properly test the designed component friction. Several terms used in describing various

Not m o re than 30°

Not more than 45°

(�
(0 1-- -
(Butt weld) (Fish mouth weld)

Fig. 7. 6. 1 Welded joints of butt weld and fishmouth weld.

228 Airframe Structural Design

sections of a welded joint are illustrated in Fig. 7 .6.2. (2) Resistance Welding (spot, roll and seam welding):
(1) Fusion Welding: Resistance spot welding is produced by the resis­
Fusion joining is accomplished by heating the tance to the flow of electrical current through the
joint zone to a molten state and weld is formed plates being jointed. The roll and seam welding
when adjacent molten surfaces coalesce and are basically spot welding process.
freeze. A filler metal may be added to the welded In general, aluminum alloy spot welded joints
joints. All fusion welds must be classified as should not be used in primary or critical struc­
follows: tures without the specific approval of the military
• Class A - a vital joint (failure would jeo­ or civil aeronautic authorities. The following are a
pardize the aircraft safety.) few types of structural connections in aircraft
• Class B l - a secondary structure (failure where spot welding should not be used.
would not jeopardize the aircraft safety) • Attachment of flanges to shear webs in
• Class B2 - nonstructure stiffened cellular construction in wings.
The fusion weld applications are summarized in • Sonic fatigue critical areas such as flaps, slats,
Fig. 7.6.3. wing fixed trailing and leading panels, etc., if
the jet engines are mounted underneath the
wing.
• Attachment of shear web flanges to wing sheet
Parent metal h eat covering.
affected zone or • Attachment of wing ribs to beam shear webs.
Attachment of hinges, brackets and fittings to

\.,!.I '"' "';

•

supporting structure.
• At joints in trussed structures.

hI \�
• At juncture points of stringers with ribs unless
a stop rivet is used.
• At ends of stiffeners or stringers unless a stop
Parent Weld zone Parent rivet is used.
meta l metal
• On each side of a joggle, or wherever there is a
possibility of tension load component, unless
stop rivets are used.
Fig. 7. 6.2 Schematic diagram of weld and parent • In general most aluminum and aluminum alloy
material. material combinations can be spot welded.

Type of Welding Characteristics

Gas tungsten-arc Tungsten electrode with inert gas shielded. The most popular welding
(or called T l O or heliarc) process in aerospace.
Hot wire gas and tungsten-arc A faster welding and improved distortion.
Pulsed current gas and Better depth-to-width ratio, narrower heat affected zone and lower
tungsten-arc porosity content.
Plasma-arc Permits higher welding speeds and improved process control.
Gas metal-arc For thick materials. ,
Vacuum electron beam (E.B.) Welding heat is generated by impingement on the metal joint by high
velocity electrons. Vacuum chamber size and depth of welding are the
limiting factors. It is a relatively costly process and requires close
tolerance machining and straight line joint surfaces. This welding process
is widely applied on titanium material (some for tension joint
applications).
Non-vacuum electron beam Least costly of the E.B. welding process and high speed welding of sheet
gages with constant thickness and straight line joints.
Pulsed laser Small function spotwelds by a concentrated coherent light beam.
Generally for electronic parts and instrumentation.

Fig. 7. 6.3 Summary of fusion welding.

Airframe Structural Design 229

(3) Flash Welding: Brazing
Flash welding is an electric current passing Brazing consists of joining metals by the application
through a joint causing a flashing action which of heat causing the flow of a thin layer, capillary
heats the metal to the fusion point and the end thickness, of nonferrous filler metal into the space
pressure is required to complete the process. between pieces. Bonding results from the intimate
This welding process is usually applied to joining contact produced by the dissolution of a small amount
tubular and solid cross-sections. of base metal in the molten filler metal without fusion
(4) Pressure Welding: of the base metal.
Pressure welding is produced by a radially oscil­ Soldering
lating oxyacetylene ring torch to heat the butting Soldering joints are accomplished by applying a low
joint under pressure. This process is similar to a melting point filler metal between joined surfaces.
flash welding. Soldering is not to be recommended as a means of
(5) Friction Welding: structural attachment due to its inherent low resis­
Friction welding joint is created through heat tance to creep.
generated by forging under rotation force, see
Fig. 7.6.7. This method is also referred to as Adhesive Bonded Joints
inertia welding and offers a wide range in joining Light weight materials can often be used with ad­
dissimilar metals. This process is more economi­ hesive bonding rather than with conventional fasten­
cal than either flash or pressure welding. ing, simply because the uniform stress distribution in
the joint permits full utilization of the strength and
rigidity of the adherends. Payloads in aircraft can be
increased.
The following are possible advantages of properly
designed adhesive bonded joints:
Flywheel • Adhesive bonds provide airtight joints .
• E lectrochem ical corrosion is reduced or e l i m inated.
In a bonded joint adhesives are electrically insu­
lated from each other; there are no holes to expose
base metal, and cladding, anodize and other corro­
sion protection surfaces are not destroyed.
• Higher fatigue life of joints may possibly allow
reduced sheet gages.
• Residual strength of damaged structure which is
adhesively bonded can be large. In a laminated
Kinetic energy of rotating flywheel is panel, cracks may grow for some time in only one
predetermined by selectio11 of moment layer; in stiffened panels, cracks may grow more
of inertia. slowly across a bonded area than across a rivet
line.
• Aerodynamic surface smoothness of adhesively
bonded structure is excellent.
Adhesively bonded joint may suffer from the
following limitations:
• Assembly may be more expensive than for conven­
... p tional joints.
• Very extensive process control over the entire
bonding procedure is required.
• Curing temperatures of some adhesives may
degrade other components.
• Service degradation of bonds is difficult to check.
Axialforce is applied and energy stored
Bonded joints shall not be used in any application
in which a complete bond failure or obvious partial
in flywheel becomes converted to heat
at thejoint fa::es.
failure could cause loss of the aircraft. Bonded joints
shall be classified for structural application as follows
as a means for designating levels and types of inspec­
tion and peel strength requirements. The appropriate
class shall be designated by the engineering depart­
ment.
7. 7 Fatigue Design Considerations
( Mecha nical fastened joints)
The purpose of this section is to introduce a rational
Welding and upset occurs just before method of analysis; the stress severity factor concept
rotation ceases. (Ref. 7. 1 9) that emphasizes the fatigue characteristics
of the structural joint rather than it's static strength.
The stress severity concept requires a detailed analy­
Fig. 7. 6.4 Principle of inertia weiding. sis of the load distribution on each critical fatigue

230 Airframe Structural Des ign

fastener within the structure. The fatigue performance •Use stiffer fasteners (material change)
of a joint is largely influenced by hole preparation, •Use stiffer doubler material
fastener type, and installation technique. Residual (b) To reduce the "peaking effect", that is the
stresses from cold working the hole and from inter­ large fastener loads developed at the ends of
ference between the fastener and hole must be the doubler
accounted for. The load distribution can be calculated • Taper the doubler thickness
by structural finite element model analysis. • Use tapered width at both ends
Once the load distribution has been found, the life • use more flexible (or smaller) fasteners at
estimates of the structural joints may be directly the ends
obtained by combined use of the severity factors, (c) In order to insure all fasteners loading up
fatigue load spectrum, and material S-N data. efficiently, and also more consistent results,
( 1 ) Basic intent of the design criterion the doubler should be installed using close
The basic intent of this joint criterion is to realize tolerance or reamed holes when interference
an equivalent fatigue life within the joint equal to fit fasteners are used.
or greater than the basic panel life. The joint (3) General guide for splice design
configuration is designed such that the peak The main effort is to keep the length of the splice
stresses within the joint are equal to or less than as short as possible. Within this limit the "peaking
the peak limitation stress as defined by the basic effect" can be dealt with as outlined in the fore­
panel. There are two essential requirements going items (b) and (c) which also apply to
necessary before the analysis can be made: (a) the splices.
basic panel design allowable stress immediately (4) The fastener load distribution
adjacent to the joint as defined by the fatigue (a) The fastener load in splices can be made to
analysis and (b) the fastener load distribution approach a somewhat uniform distribution
along the length of the splice or doubler being efficiently since they are all acting in one
analyzed. Once these two requirements are satis­ direction (unless unusual intermediate
fied, the stress concentration factor can then be applied loads are present).
determined to define the joint peak stresses by (b) In a doubler, however, the fastener loads
using curves derived from conventional stress form two groups acting in opposite direc­
concentration factor data. tions to lm&d and unload the doubler. Thus,
(2) Fastener load distribution method of analysis the fastener loads will be larger at the ends
The imporance of defining the point of maximum and vanish at the center where the relative
stress concentration in fatigue design warrants a displacement between members is zero. This
detailed definition of joint fastener load distribu­ will not, efficiently, approach uniformity as
tion. The net stress level at the fastener hole in the case of the splice. These facts are
coupled with the fastener load dictates the stress illustrated in Fig. 7.7. 1 .
concentration to be expected. The load distribu­
tion along the splice or doubler length is a
<£:
t
function of the area distribution, fastener pattern Doubler , D
(spacing, rows and diameter), fastener deflections P --. ' '. : : : : : : : - '.' ! ! ,
and materials. Excessive unconservatism can be Plate

m
expected in assuming equal distribution of load
on a given pattern. This method of analysis is P,
limited to stress levels within the elastic limit.
Since all fatigue damage should be within the
elastic range, this method is readily adaptable. Load in Mem ber D

(+)� �
Comparison of Doublers and Splices
It is helpful to keep in mind that there are two basic t

differences between doublers and splices.

( 1 ) Different purposes
• A splice's function is to transfer a given load.
l� ( - ) (+) � 1( +)
Fastener loads Fasten e r loads

It is kept as short as possible to accomplish Fig. 7. 7. 1 Comparison of internal loads in typical

this. doubler and splice.
• A doubler's function is to pick up load (and
relieve another member). In order to do this
efficiently, it must have some considerable (c) Fastener fit
length, although this is kept to a minimum. The fatigue life of a joint loaded in fluc­
Therefore, doublers are, by nature, relatively tuating shear is greatly influenced by the
long members compared to splices. fastener fit in the hole. A close fit or an
(2) General guide for doubler design interference fit hole must provide long term
(a) To increase the load picked up by the fatigue life.
doubler (d) Fastener hole preparation
• Increase the doubler planform width Drilled and punched holes have equal fatigue
• Increase the doubler thickness characteristics whereas reaming produces
• Increase the length of the doubler slight improvement. Coining process (Ref.
• Increase the number of fasteners 7.33) is the addition of a concentric groove
• Increase the size of fasteners around a hole that induces a high compres-

Airframe Structural Design 23 1

 40

�
� 30
ti S pacematic
2:'
c;:; drill only
�
20

0 No. of cycles
1 03 1 0• 1 0• 1 06 1 07

Fig. 7. 7. 2 Fatigue life comparison of drilled vs. expanded holes.

sive circumferential stress in the material • The thickness of machined countersunk sheets
around the hole edge. Under cyclic loading must be 1 .5 times the depth of the fastener coun­
of a structure, the residual compressive stress tersunk head (see Fig. 7.7.3).
lowers the applied mean stress at the hole
and hence the fatigue life is increased. Fig.
7.7.2 illustrates the improved fatigue charac­

'
10
teristics obtained . by the use of mechanical
cold work such as by expanding the hole
diameter slightly with a mandrel. 8
' �
\�
Reducing Stress Concentration by
6
Careful Detail Design
Serious fatigue troubles have, in the majority of cases, K,
been the result of unusually high stress concentration 4
factors existing in vital tension members or joints. The
relatively simple changes in detail design may improve
service life by a factor of from between 1 0 and 1 00. 2
When designing for fatigue, one should not be too
concerned about testing scatter in the order of 2
because far more potent factors are under the de­ 0
0 0.5 1 .5 2 2.5
signer's control in detail design. However, the detail tlh
design is the most important single factor affecting the
fatigue life of a structural joint and an airplane as well Fig. 7. 7. 3 Variation of stress concentration factor
(see Chapter 1 5 .0 for further discussion). with countersunk fastener hole.
General rules and information:
• Symmetrical joints should be used wherever possi­ • The fastener edge distance should not be less than
ble. 2 times the fastener diameter (metal structures).
• Joints should be designed to distribute the loading • Avoid single rows of fasteners at all splices; use
evenly between the parts. double or three rows of fasteners whenever possi­
• Local eccentricity should be kept to a minimum in ble (Fig. 7.7.4).
joints. • Use double shear joints wherever possible.
• Provide adequate increased thickness on machined • Shotpeening the plate surfaces.
members at all riveted and bolted joints. • Structural shims applied properly may reduce the

232 Airframe Structural Design

 :!
local fastener I oad transfer to improve . J'oint life
Avoid abru �t ch e in cross section.
:�::
•
• Avoid supenmp g stress conc ions.
• A void feather edges and sharp c s when spot-
facing. .
Design with sufficient fastener clamped force
)
•

(fastener
. t<;> rqu e see Fig. 7.7.5.
• Des1gn with sufficient . fastener spacing, usually
���6) .
4 and 8 times the fastener diameter (see
• �00 oounte �1,:fu .
oint> ("" Fig. 7. 7. 7).
i�
' m sec i
. . , ( "". F>g.
�ubjoct to
D un p le countersu
.
9 ) . nking
•

�
.

• Avoid d1mp 1 e .oounte"'

First fastener w ; ; rnlieved w•.
bending ("'e F 7 7 . or
I ----�;-
� ��
0 • long sphce
d
·

0
No. ofrows
____.____ 2 3 double< by "· her tapering . �he oe ° ' doubk•
plate, or by using thin •=hary oubler (see Fig.
7. 7 . 1 0). . . .
Fig. 7. 7. 4 Variation
. · of stres S Concentration factor • Wing spanw1se 1· oint design should follow Fig.
with no. of rows of !'asteners.
'
7.7.1 1 .

700

600

!_ must be > 1 ·5
1 06 h
Life cycles, N

.
Fig. 7. 7. 5 joint . . ed countersun k J'oints.
z, , comparzson between clamped and
I"e Fig. 7. 7. 7 Mach m
unclamped !'•asteners.

1 30 1
f, - Gross are a stress .
Not recomme nded m fa t'gue
1 critic al 1· oints
. ·

.
Fig. 7. 7. 6 Stress d1strz'bution between two holes at
JD-spacing. Fig. 7. 7. 8 Dimple countersun k j'oints.

Airframe Structural Design 233

 � Ii �
Dimple countersunk First First
row row

r===--@
Very thin at first fastener �
(a) By tapering the splice plate

Smaller and/or

/
softer rivet

��,
Very thin _i
auxiliary
doubler

(b) By using thin auxiliary doubler

Fig. 7. 7. 9 Dimple countersinking subject to bending. Fig. 7. 7. 1 0 Two typical designs of splice and doubler.

(Avoid) (Preferred)

(a) Integrally stiffened panel

(Avoid) (Preferred)

(b) Skin-stringer panel

(Preferred)

Fig. 7. 7. 1 1 Typical wing spanwise joint designs.

Fastener Load Distribution and By-pass Load lations justifying the use of the finite element model­
It is considered that most fatigue damage occurs ing analysis. Before simulating the structural model­
under loading conditions for which the fasteners ing, the fastener spring constant which can be con­
behave as linear elastic members. Even for very verted to equivalent structural beam member is
simple structures the treatment of the fasteners as obtained by the following hypothetical equations (see
flexible members result in rather elaborate calcu- Ref. 7.8).

234 Airframe Structural Design

 r.
Case VI: For aluminum plates and titanium fasteners
J_ (see Fig. 7.7. 1 2)

: ; �t t0vEbb { ( 1av ) 2 [ ( 1av ) 2 ]

�
l
8
C= _ OJ 3 3 2.06 +

1
d d

r. (Plate)

t _ 2t, + r.
Fastener
t, (Strap)
+
}
1 .242 (7.7.5)
av
2 Case VII: For aluminum plates, titanium straps and
titanium fasteners (see Fig. 7.7. 1 2)

t0vEbb { ( 1•v ) 2 [ ( 1•v ) 2 ]

where C Fastener constant (in/lb)
-

K, Fastener spring constant (K, 1 IC) 8

C= o . 1 325 2.06 +
- -

E00 Fastener material modulus of elasticity

- -

d d

}
Fig. 7. 7. 1 2 Average thickness of plate and straps.
+ 1 . 1 1 25 (7.7.6)
Case I: For steel plates and fasteners (see Fig. 7.7. 1 2)

t.v�b { ( 1; r [ ( 1; r ] }
It is convenient to simulate the structural finite
element modeling by a simplified model as demon­
c= 0.1 3 2.12 + + 1 .0 strated by the following joint illustrations as shown in
Fig. 7.7. 1 3.
(7.7. 1 ) When there is more than one fastener in a row
Case II: For aluminum plate and straps and aluminum (normal to the loading, or in the axial direction) the
fasteners - Same as Case I. spring constants of the individual fasteners in the row
can be simply added together and considered as one
Case III: For aluminum plates and steel fasteners (see fastener.
Fig. 7.7. 1 2) Frequently, however, in the case of doubler instal­

t.v�b { ( 1; r [ ( ; r] }
lations there are too many rows of fasteners for any
c= 0. 1 3 2. 1 2 + 1 + 1 .87 analysis to include all of them, and it is necessary to
group, or "lump", two or more rows together as one
(7.7.2) row, or one fastener. Since the end fasteners are the
most highly loaded it is best to do the least grouping
Case IV: For aluminum plate, steel straps and steel at the ends and the most at the middle. Fig. 7. 7 . 1 4
fasteners (see Fig. 7.7. 1 2) illustrates how this i s carried out.

t.v�b { ( ; r [ ( ;r ] }
Fig. 7. 7 . 1 5 through Fig. 7. 7 .22 illustrate fastener
C= 0.1 3 2.12 + + 1 .43 load distribution of several typical joints designs and
can be used as a guide for preliminary sizing purposes
(7.7.3)
Case V: For aluminum plate, steel straps and alumi­
num fasteners (see Fig. 7.7. 1 2)

c= t.v�b { ( 1; r [ 0. 1 3 2.12 + ( 1; r ] + 0.84 }

(7.7.4)
0
o o o
o o o
c o g
0
c o c o
0 0 c o o o
0 0 0 0
o o o o o o
o o o � o
0 0 0 0 0
0
o
0 0
o
o o o 0
o

(a) Actual doubler

[!" 0 0 0 o o 0 0 0 00 0 0 0000 0 0 00 0 0 0 0 0 0 :2J

(a)
Actua l structure
(b) Con1111rsion of rows to single
equi""1entfasteners

[!: o o o O 0 0 0 0 0 0 0 �
Ana lysis model
(Descriptive) (c) Grouping offastenersfor
analysis purposes

Fig. 7. 7. 1 3 Structural model of four fastener lap joint. Fig. 7. 7. 14 Grouping of fasteners.

Airframe Structural Design 235

 I
"'I
..,..__ �
I

I
p K K K
I { --..
- Fastener spring constant
K K K p

K - AEll (plate)
K,

0.4P

]
t
.,
\l! 0.3P 1-----i---���--'-�---L---.J
�

0. 2 P L-----....L.-
. -----1----__JIXl------L----__JL--___---1

Fig. 7. 7. 1 5 Fastener load distribution of three rows of fasteners.

"4--- t
�
S
I
p -4-- K K 'K K
I �
- Fastener spring constant
K 'K K s - P

- AEI I (plate)
K, .
0.4P 1-----IL...---- K

0.1 P ,_______j_______ 1.- ______ ....1...______ _.1

____ ....J

Fig. 7. 7. 1 6 Fastener load distribution offour rows of fasteners.

236 Airframe Structural Design

 f--- I
--!
I
p

I
..,_ <
I I s�
K K K K K
I
p
K K K K K

K1 - Fastener spring constant .

0.4 P
K - AE/l (plate)

Fig. 7. 7. 1 7 Fastener load distribution of five rows of fasteners.

--I I-- �
l � � K ---+
14-- l l
P - �..�
._ , .5 K ,. K
_---l--
I
K, - Fastener spring constant
1 .5K I
K ----J>--- K . � -- P
p 2/3K _ ...
K
I
K, - Fastener spring constant
0.4P 1----1-----�� 0.6P
K - AEI l (plate) K - AE/l (plate)
� �

�
.la .la
� �
5 0.3P 1---+----"'"'"=--h�---l 0.5P 1---+-----=::::!o..�--1--1
1:i
� �
0.2P '----'-----"--' K, - 0.4K
0.4P '-----..L.-----'--'

Fig. 7. 7. 1 8 Fastener load distribution of three rows Fig. 7. 7. 20 Fastener load distribution of two rows of
of fasteners with different plate thickness. fasteners with tapered plate thickness.

�l �
\5K I i.�K£�£s:�s:+ KI s- P
I- t -l

l
P -l

I I \- P
""
P- 2.5K 2K K '-
K, - F a ste n er s p ri ng co n sta nt 2.5K 2.5K 2.5K
0.4P �--1-_____,�_____,____....___ _--+____,
K - AE/l (plate) 0.5P >-----+- K, - Fastener spring constant

_
K - AE/l (plate)
l 0.3P �--1----+---r--1 l

�
.la .la 0.4P >------"',,___, __._______,.______,

�
_

� 0.2P 1---+-->.--+-----++--1
ll'

� 0.3P 1---t-----'�'-�=::::zlf.---i

0. 2P .____., _____,________,.____,
K, - 2K

Fig. 7. 7. 1 9 Fastener load distribution of five rows of Fig. 7. 7. 2 1 Fastener load distribution of three rows
fasteners with different plate thickness. of fasteners with

Airframe Structural Design 23 7

 '"
p - -
4.5K 4K 3K 2K
'
K I
I ' K 2K 3K
'
4K 4.5K - p
'

I
0.3P I

I
Fastener load curve

0.2P ' "' ,. ,.

'"<:!
..
.l:!
�
l!!
..,
<l! 0.1 P
Note: For the axial spring rate (K) distribution given
the load in allfasteners is the same and

I I
independent of the fastener spring constant (K,J

0 I
Fig. 7. 7. 22 Fastener load distribution of five rows of fasteners with stepped plate thickness.

Severity Factor (SF) Concept

( 1 ) Concept (Ref. 7.7)
The severity factor SF, which is a fatigue factor
that accounts for:
• Fastener type, method of installation, inter­
ference, hole preparation, etc.
• Detail design
• Fastener load distribution to avoid "peaking
effect"
• Minimization of the stress concentration
caused by both local load transfer at a fastener
and bypass load (see Fig. 7.7.23).
P Bypassing load in the
=

element (2) Local Stresses

P + t;. P
The maximum local stress in the considered
6.P - Load transfer through the element,

H
fastener

P+ AP f AP �� - P
Local stresses caused by load
where a1
transfer LJP, see Fig. 7.7.24(a)
Local stresses caused by pypass
Fig. 7. 7. 23 Load fastener load transfer and bypass load. load P, see Fig. 7.7.24(b)

Faying surface

1
o, - K,0 dt (J
t:, p
(a) Local stresses caused (b) Local stresses caused by
by load transfer, 6.P bypassing load, P

Fig. 7. 7. 24 Local peak stresses caused by load transfer and bypass load.

238 Airframe Structural Design

K,b =
Stress concentration factor, bear­ gross area stress, see Fig. 7. 7 .26
ing stress, see Fig. 7.7.25 (J Bearing distribution factor, see
K,8 =
Stress concentration factor, bypass Fig. 7.7.27

J
I
:: I : I
3.0
1 -- 0
,__

] ! I
-

- ---
v
-;
T -

-9�
- -0 )

I
d
-
w I

l om•x
-
I
,__

Krb - Oma.1 °

v
2.0

/
v
/

.... v
�
1 .0 ----
0 0.1 0.2 0.3 0.4 0.5 0.6
dlw

Fig. 7. 7. 25 Bearing stress concentration factor - K,b.

--- -
3.6 - e
---
o- - o

3.5 ---
-
om..,
3.4
K,9 - [Link]
K,, 3.3

3.2

r
c
Fig. 7. 7. 26 Stress concentration factor - K,8 •

Airframe Structural Design 239

 6.0

Single shear

5.0

4.0

3.0

2.0

1 .0
0 1 .0 2.0 3.0 4.0 5.0 6.0 7.0

-
tld - (plate thickness)/(fastener dillmeter)

Fig. 7. 7. 2 7 Bearing distribution factor 9

Dividing the maximum local stress by the

reference gross area gives a stress concentration a
factor SCF for the considered element: Standard hole drilled 1 .0
Broached or reamed 0.9
Omax Cold worked holes 0.7-0.8
SCF -

-
Orcr

As mentioned previously, the fatigue perfor­ Fig. 7. 7. 28 Hole condition factor a.

mance of fastener joint is largely influenced by
fastener type and method of installation. These
effects are expressed by a factor which is applied
to the previously derived SCF in order to get a
dimensionless factor expressing the fatigue design {J

( PJ
quality: Open holes 1 .0
Lock bolt (steel) 0.75
a{J t::,.p Rivets 0.75
SF - - K,b e - + K,8 - (7.7.7)
o,.1 dt wt Threaded bolts 0.75-0.9
Taper-Lok 0.5

...
where a - Hole or surface condition factor, see Hi-Lok 0.75

-
Fig. 7.7.28
{J Hole filling factor, see Fig. 7.7.29
(a and {J are empirical factors derived from Fig. 7. 7. 2 9 Hole filling factor {J.
fatigue tests)

A condition of an unreinforced open hole (not

cold work) in a wide sheet, the severity factor SF (3) Application of Severity Factors
- 3.0. For a hole tilled by a well driven fastener, Procedures:
with no load transfer, SF < 3.0. In case of a load Perform a load tnp1sfer analysis.
Evaluate a, {J, 9, K,b and K11•
•

transfer, usually, SF > 3.0. •

240 Airframe Structural D�ign

 • Calculate the severity factory SF. Because the Severity Factor is, after all, a
direct measure of the severity of the state of
stress at a particular location in a structural joint,
relative values of Severity Factors at all fastener
Example I: hole intersections in a joint give an excellent
indication of which locations are the most critical
Consider a plate with an open hole subjected to a from a fatigue standpoint. Fatigue tests have
tensile load shown that specimens almost always fail at the
,r Plate location of the highest Severity Factor.
A typical joint will have many fastener hole
a- Open hol'!, �
p
intersections. The overriding problem in Severity
Factor analysis is, therefore, that of obtaining a
detailed stress distribution of the entire joint
region. Once the stress distribution has been
found, the Severity Factors may then be deter­
mined as per Eq. ( 7.7.7).
a 1 .0 (see Fig. 7.7.28)
{3 1 .0 (see Fig. 7.7.29) Structural Joint Life Prediction
a1 0 (no load transfer) The basic formula for predicting the fatigue life of a

--
structural joint is that of the severity factor SF. In this
p p equation (J, a and {3 vary with the joint material,
a2 = Ab K,g = 3.0 -- fastener type, and. installation process. For preliminary
ypass A bypass design, Fig. 7.7.27, Fig. 7.7.28 and Fig. 7.7.29 can be
Thus, used to estimate the values of (J, a and {3 respectively.
However, for detail design, coupon test specimen
should be required to achieve more accurate results.
In addition, the digital computer is strongly recom­
( t:: )
Uref
mended as a useful tool for analyzing member stresses
( 1 .0) ( 1 .0) 0 + 3.0 and fastener load distribution in a joint. Ideally, it is
= =
3.0 desired to achieve evenly distributed loads on each
p fastener. This is especially important on the end
fasteners as they are most sensitive in joint life
analysis. Once the load distribution is obtained and 8,
The result, of course, is simply the stress a and {3 are determined, the severity factor SF can be

concentration factor for an open hole. The SF, calculated. Then, knowing the defined constant am­
therefore, may be imagined as a stress concentra­ plitude loading, the predicted fatigue life of the
tion factor with reference to some hypothetical proposed (baseline) joint can be obtained using the
stress which would exist if the entire local area of SF as calculated above in conjunction with material
interest were replaced with a single uniform piece constant-life diagrams (or S-N curves).
of structure. The SF is not referenced to the
nominal stress at the particular hole location. Example

Example II:
Given a critical support structural joint:
Consider a simple lap joint
t
T
\._ ......
_.f.�....__ I ! -- p 7.2 i n

1
P - ___ __ �

p ......_ 5d � p

.--
---- -
24.07 kip
1
---- �1----
-
�
�..
9.04 kip

�
...;.
t ......
t
1 7.54 kip

... 2.48 kip

a = 0.9 (see Fig. 7.7.28)
{3 = 0.75 (see Fig. 7.7.29) (a) Actural structures r- 0.3 in
(J = 2.0 (see Fig. 7.7.27)
K,b = 1 .26 (see Fig. 7.7.25)
Thus

[ ]
24.07 kip ------1
(0.9) (0.75 ) ! 0 = 8.5
SF = (2.0) ( 1 .26 ) + Finite element
p dt model number
(b) Equivalentfinite element model output data
5dt

Airframe Structural Design 24 1

 ==
Material- 2024-T3 plate with t = 0.3 inch prediction, then the SF value should be revised by the
Fastener Diameter = �� (or 0.6875) inch following equation (also see Fig. 7. 7 .30).
K a + b (SF) (7.7. 8)
- Threaded bolts where K Fatigue quality index (see Chapter 15 .0)

= =
All loads at this joint come from the finite element a and b = constants obtained from test results,
model, a = 0- 1 .5 and b = 1 .0- 1 .5
24.07 K = SF, if a 0 and b 1 .0
a,.r = = 1 1 .14 ks1.
7.2 ( 0.3 )
LJ P = 9.01 - 2.48 = 6.53 kip
a = 1 .0 (from Fig. 7.7.28)
f3 = 0.855 (from Fig. 7.7.29) I
K,b = 1 .25 (from Fig. 7.7.25)
/
,
K,8 = 3.0 (from Fig. 7.7.26)
(J = 1 .4 (from Fig. 7.7.27) 10

From Eq. ( 7.7.7) /'

( )
9 '

SF = af3 K,b fJ
LJ P+ K,8 .!._

/ i/
'
a,.r dt

(
wt 8

1 .0 (0.855) 6.53
I 11"
�
= 1 .25 ( 1 .4) 2 (0.6875 ) (0.3)
1 1.14
+ 3.0
1 7.54
7.2 (0.3)
) I
i
I

;
v
II�

= 0.0769 ( 27.7 + 24.36)

= 4.0 )" ; ;
4
Given structural joint fatigue loading was defined ; -'

as a constant amplitude loading of 24.07 ± 24.07 3

;�
kip, then '

K - a + b (SF )
2 (24·07) 2
Smax = 22.29 ksi '
Value : a - 0.0 - 1 .5
7.2 (0.3) b - 1 .0 - 1 . 5

Sm;n = O 2 3 4 5 6 7 8
Severity factor (SF)
(S S
S = max - m;n) = (22.29 - 0) 1 1 .145 ksi
a
2 2 Fig. 7. 7. 30 Severity factor (SF) vs fatigue quality
(S + Sm;n ) (22.29 + O) index (K).
Smean = max = 1 1 . 145 ksi
2 2
(For definitions of Smax > Smin • s. and Smean • see Chapter
1 5.0). 4.3 2.8 2.8 3.6 2.9
From Fig. 1 5 .4.6 of Chapter 1 5.0, obtain the pre­
dicted life for this joint as 46,000 cycles.
After the predicted life of the baseline configura­
tion is complete, a small component test specimen
should be built to measure the strain gage reading at
the most critical location. Each specimen should
represent the typical as well as the critical part of the (a) Baseline design
entire joint structure. In addition to a strain gage
survey at some critical cross section, it is highly riffener
Skin _____. 2 . 1 2.4 2. 0 2 . 2 2.2 2.9 2.6 2.6 2 . 7 2.4 Scu l ptured

:24.03.1 ;_/ 2:1 2:3 2.13.3�.1 2.1 2'.6?6

desirable to perform a stress coat, or photo stress Doub ler F 1 .4 1 .5 3i0 3.3 2.8
measurement of the test specimen. A structural finite Ir
element modeling analysis should also be performed,
taking into account the method of loading and sup­
porting the test specimen. Flange
St
:r
5i · I
�-f333- �-�-�-� - --- - -- -- -- J
•
-
2. 0 2. 2 3. 1 3.4 3.3
By correlating the final analysis with the measured
strain data, a high level of confidence will be obtained (b) Improved design
that an accurate stress distribution has been achieved.
When the fatigue test of the specimen is completed Fig. 7. 7. 3 1 SF comparison between baseline and
and the test life is significantly different from the improved design.

242 Airframe Structural Design

 Once this is accomplished, improvements of the
baseline configuration can easily be evaluated with
just a few minor changes in the finite element model­
ing analysis as illustrated in Fig. 7.7.3 1 . Fig. 7.7.32
illustrates the comparison of five different joint
designs from the best design of SF = 3.2 to the worst
design of SF = 1 3.
7.8 Shim Control and Requ i rement
Shims used where they could have been avoided are
3 costly and wasteful, shims not used where needed, or
used improperly, can be far more costly. Engineering
can reduce both cost and waste by controlling shim
usage through design and specifications as shown in
(a) Double-scar/joint specimen (SF = 3.2)

1/2 stud

(b) Double-shearjoint specimen (SF = 4.1) (c) Stepped double shearjoint specimen (SF = 4.3)

P=== c , � r·==::rJ. 7 '!,

5°43
2 "!... -:I
-:I

(d) Plain-scarfjoint specimen (SF - 4.1 - 8.1) (e) Single-shearjoint specimen (SF = 13)

Fig. 7. 7. 32 SF comparison offive different joint designs (all dimensions are in inches).

Fig. 7.8. 1 . Another improvement design is that of the

wing joint which eliminates the need for shimming as
illustrated in Fig. 7.8.2.
Shims are used in aircraft production to control
structural fit-up, to maintain contour or alignment,
+ + +- _
+ 1• and for aesthetic purposes. In attempting to control
- shim usage by specification, the engineer now
depends on the maximum shim thickness in the parts
-
-
--
- - - - list and the note "shim if and as required". The
mechanic must then decide when to shim, what the
shim taper and thickness should be, what gap to allow,
and whether the gap should be shimmed or pulled up
Fig. 7. 8. 1 Wing joint design (with shims). with fasteners.

Airframe Structural Design 243

 An examination of improper shimming practices The responsibility for shims belongs to the en­
reveals some unsatisfactory results. These include a gineer who must evaluate each of the designs to deter­
great number of unnecessary shims, waste of manu­ mine proper shimming requirements. It must be speci­
facturing manhours, and increased production costs. fied when, where, what, and how to shim. Above all,
the engineer must not delegate the responsibilities to
the mechanic.
Estimates of Tension Stress from Pull-up
Each problem is unique and requires a study of
deflection. Some example analyses are included in Fig.
+ + 7.8.3.
After the tension residual stress (or sustained
tensile stress) is estimated, do not allow this stress to
t exceed the allowable (as shown in Fig. 4.7.7).
Outer wing
surface Example of pull-up
(a) Sequence of assembly may cause gap as shown
Fig. 7. 8. 2 Wing joint design (without shims).

111 inc�
Clamping sll'f!sses (of cross-section fZ2ZZZ2ZI .Lt)
.......
Case Loading condition Stress

.---=-= - - --.., ...L

1 .5 E t} t

(1
��·� L -�4� r. - -L- ,-

��-t
6
E b i' --
M - -- E 8 i'
6L' - 1 2L
----�
f.0 - � - �
�J �
L' 2L
( L ��
· ,._��
�--i-----------+---------1

�
3E t} t
'· - �

f. _ 6E t5 t
•
L'

�----==�'* M - E b i' 4E t} t
(5) �
1 .5L2 '· - �

(6) �-t-
� f
6.4E b i'
w - --
L'
f. -
b
[Link] t} t
--
L'

�l
�r
32E t} i' 16E t} t
f.b - --
L'
w - --
L'
(7) � L �
P L t}
t ��t f. -
b
1 2E t} t
--
L'
(8) r L ·I

� +
2.4E t} t
f. - --

�1J-� -+-8
b
L'
(

Fig. 7. 8. 3 Pull-up (or clamp-up) stresses.

244 Airframe Structural Design

 Gap 1 .1 in How to Minimize Shims
To minimize shims the engineer shall:
(a) Calculate part tolerances
2024-TJ
(b) Check sequence of assembly
(c) Define max. gap for pull-up
1' (d) minimize stress concentration
1 + + + (e) Use materials which exhibit good stress corrosion
resistance
I
....._ _
(t) Do not allow residual tensile stresses to exceed
I the allowables specified
I (g) Use stress relief heat treating and surface cold
working (such as shot peening) to remove resi­
•

•
=
(b) Define maximum gap for pull-up
Maximum allowable tensile sustained stress F,
50 ksi (refer to Fig. 4.7.7)
Maximum pull-up gap [see case (3) of Fig.
dual surface stresses. Removal of residual tension
stresses is especially important for high heat treat
steel and aluminium alloys
(h) Use control of tolerances to reduce assembly fit­
7.8.3J up stresses
(i) Mimimize stresses in the short-transverse plane
{> - F,L2 - 50 ( 1 . 1 ) 2 - O 02 1. (gap) G) Necessary shims:
-
3 ( 1 0.4 ) ( 1 0 3 ) (0.1 ) 0 Specify shim material and size.
3 Et •
·

• If taper is required, minimum shim thickness

Fatigue Consideration should not be less than 0.03 inch.
Fatigue performance depends on alternating stresses • Specify structural shim or non-structural shim,
and mean stresses. While pull up rarely affects the see Fig. 7.8.6.
alternating stress in a given part, it will increase the
mean stress. The significance of this increase is
illustrated in Fig. 7.8.4. The tension residual stress
due to pull-up is 60,000 psi at point A. The normal Shim

31 I 11 I %'
operating mean stress has been established at 45,000
psi from design condition. When the residual stress is
added to the design mean stress and evaluated on the
material S-N curve, the part life is significantly re­
duced at all alternating stress levels as illustrated in
Fig. 7.8.5.
(a) Non-structural shim

ffi ! 11 I J I : 1 £ 1
�
Shim

1 .01 i n

(b) Structural shim

Fig. 7. 8.4 Pull-up stress due to the mismatch between
two structures.

Fig. 7. 8. 6 Differences between structural and non­

structural shims.

s.

Refere nces
7.1. Ritchie, Oscar: 'Design Trade-offs that Determine
1 0,000 h r 60,000 hr Fastener Selection.' SAE paper No. 670886. SAE
Fig. 7. 8. 5 Reduction of structural life due to pull-up Transactions, ( Vol. 76, 1 967), 280 1 -2807.
stress. 7.2. Ramsey, C.L. and Ingram, Jr., J.C.: 'Structural Fast-

Airframe Structural Design 245

 eners for Extreme Temperature Utilization.' SAE tion, Wheaton, Ill.6 0 1 8 7 .
paper No. 6 70887. SAE Transactions, (Vol. 76, 1 96 7 ) , 7 . 3 0 . Cole, R.T.: 'Fasteners for Composite Structures.'
2808-28 1 5 . Composite, (July, 1 98 2 ) .
7.3. Hills, J.F.: 'Advantages of Some New Fastener Sys­ 7.3 1 . Phillips, J.L.: 'Fastening Composite Structures with
tems.' SAE paper No. 680206. HUCK Fasteners.' Technical Paper of HUCK Manu­
7.4. Smith, C.R.: 'Effective Stress Concentrations for facturing Co. ( 1 984 ) .
Fillets in Landed Structures.' Experimental Me­ 7.32. Phillips, J.L.: 'Fatigue Improvement by Sleeve Cold­
chanics, (Apr. 1 97 1 ) , 1 6 7- 1 7 1 . working.' SAE Paper No. 730905. SAE Transactions,
7.5. Melcon, M.A. and Hoblit, F.M.: 'Developments in the (Vol. 82. 1 97 3 ) , 2995-30 1 1 .
Analysis of Lugs and Shear Pins.' Product Engineer­ 7 .33. Speakman, E.R.: 'Fatigue Life Improvement Through
ing, (June, 1 95 3) , pp. 1 60- 1 70. Stress Conining Methods.' Douglas Paper No. 5506.
7.6. MIL-HDBK-5D: 'Metallic Materials and Elements for Douglas Aircraft Company, ( 1 96 9) .
Flight Vehicle Structures. U.S. Government Printing 7.34. Kho!. R.: 'Fasteners That Fight Fatigue.' Machine
Office, Washington, D.C. 1 98 3 . Design , ( Feb. 20, 1 97 5 ) .
7.7. Jarfall, L.E.: Optimum Design of Joints: The Stress 7.35. Boucher, R.C.: 'Table Speeds Calculation of Strength
Severity Factor Concept. The Aeronautical Research of Threads.' Product Engineering, (Nov. 27, 1 96 1 ) .
Institute of Sweden, 1 967. 7.36. Baumgartner, T.C. and Kull, F.R.: 'Determining
7.8. Tate, M.B.: 'Preliminary Investigation of The Loads Preload in a Bolted Joint.' Machine Design , (Feb. 1 3,
Carried by Individual Bolted Joints.' NA CA TN 1051, 1 964) .
( 1 946 ) . 7.37. Anon.: SPS Threaded Fasteners - Section 1, Reference
7.9. McCombs, W.F.: 'Analytical Design Methods for Guide to Bolts and Screws . Standard Pressed Steel
Aircraft Structural Joints.' AFFDL - TR-67-184, Co., Santa Ana, Calif. 92702. 1 966.
( 1 96 8 ) . 7.38. Anon.: SPS Threaded Fasteners - Section II, Refer­
7 . 1 0. Peterson, R.E.: Stress Concentration Design Factors. ence Guide to Self-Locking Nuts . Standard Pressed
John Wiley and Sons, 1 95 3 . Steel Co., Santa Ana, Calif. 92702. 1 966.
7.1 1 . Hartmann, C.C.: 'Static and Fatigue Strengths of 7.39. Hopper, A.G. and Thompson, G.V.: 'How to calcu­
High-Strength Aluminum-Alloy Bolted Joints.' late and Design for Stress in Preloaded Bolts.' Prod­
NA CA TN 2276 , ( 1 95 1 ) . uct Engineering, (Sept. 1 4, 1 964) .
7 . 1 2. Rosenfeld, SJ.: 'Analytical and Experimental Inves­ 7.40. Cobb, B.J.: 'Torque and Strength Requirements for
tigation of Bolted Joints.' NA CA TN 1458, ( 1 94 7 ) . Preloading of Bolts.' Product Engineering, (Aug. 1 9,
7. 1 3. Grover, HJ.: 'Fatigue of Aircraft Structures.' 1 96 3 ) .
NA VAlR Ol -lA -13, ( 1 966 ) . 7.4 1 . Leftheris, B.P.: Stress Wave Riveting. Grumman
7 . 1 4. Anon.: Fatigue and Stress Corrosion Manual for Aerospace Corp., Bethpage, new York 1 1 7 1 4. 1 972.
Designers. Lockheed-California Company. 7.42. Cornford, A.S.: 'Bolt Preload - How can you be sure
7. 1 5 . Deneff, G.V.: 'Fatigue Prediction Study.' WADD TR it's right?' Machine Design , ( Mar. 6, 1 975 ) .
61-153, ( 1 96 2 ) . 7.43. Dann, R.T.: 'How Much Preload for Fasteners;.'
7 . 1 6. Niu, Michael Chun-yung ( 4- * ::J ): 'L- 1 0 1 1 Fastener Machine Design , (A ug. 2 1 , 1 97 5 ) .
Handbook.' Lockheed-California Company. Report 7.44. Whaley, R.E.: 'Stress-Concentration Factors for
No. CER 51 -013, (Jan. 24, 1 97 3 ) . Countersunk Holes.' Experimental Mechanics, (Aug.
7. 1 7. Anon.: 'Fastener Technology Catches Up.' A viation 1 96 5 ) .
Week &Space Technology, (Dec. 1 4, 1 987 ) , 89-93. 7.45. Harris, H.G. and Ojalvo, I.U.: Simplified Three­
7 . 1 8. Bruhn, E.F.: Analysis and Design of Flight Vehicle Dimensional A nalysis of Mechanically Fastened Joints.
Structures. Tri-State Offset Company, Cincinnati, Grumann Aerospace Corp., Bethpage, New York
Ohio 45202. 1 96 5 . 1 1 7 1 4. 1 974.
7. 1 9. Niu, Michael Chun-Yung ( '*- * ::J): Structural Joint 7.46. Thrall, Jr. E.W.: 'Fatigue Life of Thick-Skinned
A nalysis Handbook (Fatigue). Lockheed-California Tension Joints.' A eronautical Engineering Review,
company, 1 97 3 . (unpublished) . ( Nov., 1 95 3 ) , 37-45.
7.20. HANDBOOK H28: 'Screw-Thread Standards for 7.47. Brilmyer, H.G.: 'Fatigue Analysis of Aircraft Bolts.'
Federal Services (Part I, II, & III ) .' U.S. Department of Aeronautical Engineering Review, (July, 1 95 5 ) ,
Commerce, National Bureau of Standards (1 95 7). 48-54.
7.2 1 . Sines, G. and Waisman, J.L.: Metal Fatigue. McGraw­ 7.48. Anon.: Adhesive Bonding AL COA A luminum .
Hill Book Company, Inc., New York, N.Y. 1 959. Aluminum Company of America, Pittsburgh, Pa.
7.22. Kimball, D.W. and Barr, J.H.: Elements of Machine 1 96 7 .
Design. John Wiley and Sons, Inc. new York, N.Y. 7.49. Trembley, W.H.: 'Fastening Analysis t o Optimized
1 93 5 . Design.' SA E Paper No. 730309. SAE Transactions,
7.23. Kuenzi, E.W.: 'Determination of Mechanical Proper­ (Vol. 82, 1 97 3 ) , 1 066- 1 074.
ties of Adhesives for Use in Design of Bonded Joints.' 7.50. Silha, C.W. and Schwenk, W.: 'Fasteners with a
U.S. Dept. Agriculture Forest Service, Prod. Lab., Memory.' SAE Paper No. 730900. SAE Transactions ,
Report No. 1851 , ( 1 95 7 ) . (Vol. 82, 1 97 3 ) , 2982-2987.
7.24. MIL-A-5090B: Adhesive: A irframe Structural; Metal 7.5 1 . Gill, F.L.: 'A Corrosion Inhibiting Coating for Struc­
to Metal. Government Printing Office, Washington, tural Airframe Fasteners.' SAE Paper No. 730902.
D.C. 1 95 5 . SAE Transactions, (Vol. 82, 1 97 3 ) , 2 9 88-2994.
7.25. McClure, J.G.: 'The Prestressed Bolt.' Machine 7.52. Robinson. G.l.: 'Influences of Grain Flow on the
Design (Sept. 1 5 , 1 960) . Strength of Lugs.' A ircraft Engineering, ( Sept., 1 9 5 1 ) ,
7.26. Stewart, W.C.: Bolted Joints - ASME Handbook. 25 7-260.
McGraw-Hill Book Company, N.Y. 7.53. Anon.: Shur-Lok Specialty Fasteners for Industry, (9th
7.27. Anon.: 'Fasteners Reference Issue.' Machine Design , edition) Shur-Lok Corp., P.O. Box 1 95 84, Irvine,
( 1 987 ) . CA. 927 1 3.
7.28. Parmley, R.0.: Standard Handbook of Fastening and
Joining. McGraw-Hill Book Company.
7.29. Anon.: Assembly Engineering. A Hitchcock Publica-

246 Airframe Structural Design