JOURNAL OF THE

AUSTRALIAN
INSTITUTE OF STEEL
CONSTRUCTION
(ABN)/ACN (94) 000 973 839

CONSTRUCTION

An Engineering Insight into the Fundamental

Behaviour of Tensile Bolted Joints
Design of Steel Roof & Wall
Cladding Systems for Pull--out Failures

VOLUME 35 NUMBER 1
MARCH 2001
ISBN 0049--2205
Print Post Approved
pp 255003/01614
 AISC DETAILER MEMBERS -- The best in Steel Fabrication
AISC disseminates information on up--to--date steel design and fabrication technology, and this information flows onto its
detailer and fabricator members. When considering fabricated steelwork, it makes sense to deal with those detailers and
fabricators who share the institute’s resources. Steel detailing company names, addresses and phone numbers are shown
below. Steel fabricator members contact details can be found on the inside back cover.

New South Wales & ACT Wilds Drafting Service Pty Ltd
Ahaust Steel Detailers 119 Loridan Drive Brinsmead, Cairns 4870 07 4034 1435
111 Best Road, Seven Hills 2147 02 9831 6511 South Australia
Cadraw Pty Ltd Sasteel Drafting Service P/L
10/33 Windsor Rd, Kellyville 2155 02 9629 4976 33 Maxwell Road, Pooraka 5095 08 8349 9622
Centreline Drawing Services Pty Ltd Steel Drafting Pty Ltd
6/21 Oaks Ave, Dee Why 2099 02 9981 4432 6/42 Quick Rd, Mitchell Park 5043 08 8357 4405
Elmasry Steel Design And Detailing Steelplan Australia P/L
88 Arthur Street, Strathfield 2135 02 9764 6660 1/27 College Road, Kent Town 5057 08 8362 7603
Enterprise Drafting Company P/L, Suite 2, Warradale Drafting Service P/L
Level 1 163 King St, Newcastle 2300 02 4929 6910 1 Boulder Court, Woodcroft 5162 08 8322 5533
Evan Swan Pty Ltd
7 Rutledge Ave, Dapto 2530 02 4261 8763 Tasmania
Monaro Drafting Abel Drafting Services
PO Box 299, Cooma 2630 02 6452 2337 PO Box 579, Wynyard 7325 03 6445 1411
Production Line Drafting P/L Hills Project Services
3/131 Howick St., Bathurst NSW 2795 02 6334 3500 269 Wellington St, Launceston 7250 03 6343 0122
R & B Elston Victoria
6 Park St, Ingleburn 2565 02 9605 2668
Southtech Bayside Drafting Pty Ltd
PO Box 270, Moruya 2537 02 4474 2120 Cnr Skye Road & Farrell Street, Frankston 3199 03 9781 4011
Supadraft BDS Technical Services
PO Box 716, Brookvale 2100 02 9975 1777 Level 1, 240 Bay Street, Brighton 3186 03 9596 6500
Engineering Design Resource
Northern Territory 68 Hotham St, Traralgon 3844 03 5174 0255
Precision Drafting Australia Fabcad Drafting P/L
11 Malabar Court, Larrakeyah 820 08 8941 2666 68 Hotham St, Traralgon 3844 03 5174 9026
Structec Pty Ltd Flexsteel Drafting Service
PO Box 4381, Darwin 801 08 8981 9472 3 Monterey Cresc, Donvale 3111 03 9842 1737
Queensland Haddow Drafting P/L
1/14 Summerhill Rd, Glen Iris 3146 0419 584 962
4D Steel Detailing Pty Ltd, PO Box 193
Fortitude Valley, Brisbane 4006 07 3839 8075 PM Design Group
17b Julia St, Portland 3305 03 5521 7204
BDS Technical Services
80 Tribune Street, South Brisbane 4101 07 3844 8093 Precision Design Pty Ltd
Level 1 75--89 High St, Cranbourne 3977 03 5995 2333
Brice Engineers Pty Ltd
7--8 Brice Court, Mt Louisa 4814 07 4774 8322 Structural Steelwork Solutions
3/237 Glen Eira Rd, Caulfield North 3161 03 9550 0330
Cadtech Queensland Pty Ltd
7 Albany Creek Road, Aspley 4034 07 3862 7655 Universal Steel Detailers
PO Box 129, Wendouree 3355 03 5339 9690
DKL Drafting
5 Thor Street, Victoria Point 4165 07 3207 7769 Western Australia
Endraft Pty Ltd Barry Carnegie & Assoc P/L, Unit 3
PO Box 256, Fortitude Valley 4005 07 3252 5467 46 Hasler Road, Osborne Park 6017 08 9244 1311
Hempsall Steel Detailing Pty Ltd Cadstruction Drafting
Suite 1/67 Redcliffe Pde, Redcliffe 4020 07 3284 3020 Suite 4 First Floor, East Victoria Park 6101 08 9380 6069
Mount Isa Mines Ltd Mod Drafting
Design Office, Mount Isa 4825 07 4744 3725 Unit 2/5 Edwards Cresc, Redcliffe 6104 08 9454 3753
Online Drafting Services Qld, Unit 6 Pacific Chambers Multiplan
3460 Pacific Highway, Springwood 4127 07 3299 2890 Unit 12 4 Queen St, Bentley 6102 08 9356 5993
Pacific Computing Pty Ltd Perth Drafting Company (WA)
80 Tribune Street, South Brisbane 4101 07 3844 7088 48 Kishorn Road, Applecross 6153 08 9364 8288
Paul Anderson Drafting Service Pty Ltd Steelplan Australia Pty Ltd
39 Lurnea Crescent, Mooloolaba 4557 07 5478 0186 15/885 Albany Highway, East Victoria Park 6101 08 9362 2599
Q E I Pty Ltd Universal Drafting
361 -- 363 Montague Road, West End 4101 07 3844 2772 7/175 Main St, Osborne Park 6017 08 9440 4750
Steelcad Drafting Pty Ltd Westplan Drafting, Unit 3/11,
PO Box 1456 Coorparoo DC 4151 07 3844 3955 Robinson Road, Rockingham 6168 08 9592 2499
Steeltech Steel Detailers P/L
24 Curzon Street, Tennyson 4105 07 3848 6464 New Zealand
Tregar Engineering Services Ormond Stock Associates Ltd
800 Ann Street, Fortitude Valley 4006 07 3252 7104 PO Box 1048, Palmerston North, NZ. 64 6 356 1088

While every effort has been made and all reasonable care taken to ensure the accuracy of the material contained herein the Authors, Edi-
tors and Publishers of this publication shall not be held responsible in any way whatsoever and expressly disclaim any liability or respon-
sibility for any loss or damage costs or expenses howsoever incurred by any person whether the purchaser of this work or otherwise in-
cluding but without in any way limiting any loss or damage costs or expenses incurred as a result of or in connection with the reliance
whether whole or partial by any person as aforesaid upon any part of the contents of this publication. Should expert assistance be required,
the services of a competent professional person should be sought.

Editor: Peter Kneen
STEEL CONSTRUCTION is published quarterly by the
Australian Institute of Steel Construction -- a national body
whose purpose is to promote the use of fabricated steel
through engineering research and the dissemination of
knowledge.
Its services, which are available free of charge to financial
corporate members, include technical information and
advice and a library which contains local and overseas
publications.
For details regarding AISC services, readers may contact
the Institue’s offices, or visit the website: www.aisc.com.au
AISC CONTACTS
Head Office -- Sydney
Level 13, 99 Mount Street
North Sydney NSW 2060
(PO Box 6366, North Sydney NSW 2059)
Tel: (02) 9929 6666 Facsimilie (02) 9955 5406
Email: enquiries@aisc.com.au
Website: www.aisc.com.au
New South Wales & ACT
State Manager -- NSW & ACT
Scott Munter -- Mob 0418 970 899
Tel (02) 9929 6307 Fax (02) 9955 5406
Queensland & N.T.
State Manager -- Queensland & NT
John Gardner -- Mob 0418 788 870
Tel (07) 3243 8311 Fax (07) 3422 1888
Victoria & Tasmania
State Manager -- Victoria & Tasmania
Leigh Wilson -- Mob 0417 353 364
Tel (03) 9529 1799 Fax (03) 9529 1744
Western & South Australia
State Manager -- Western & South Australia
Rupert Grayston -- Mob 0419 922 294
Tel (08) 9480 1166 Fax (08) 9226 2355
1 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001
EDITORIAL
This issue presents two papers dealing with fasteners.
The first paper describes some fundamental aspects
of bolted joint behaviour and in particular the
importance of the preload in the bolts of a tensile
connection. The importance of preload and dynamic
loads are discussed.
The second paper presents results of research done
into the failure modes around profiled sheet metal
cladding under simulated wind uplift conditions.
When thin steel roof and wall cladding systems are
subjected to wind uplift/suction forces, local
pull--through or pull--out failures occur prematurely
at their screwed connections. During high wind
events such as storms and cyclones, these localized
failures then lead to severe damage to buildings and
their contents. In recent times, the use of thin steel
battens, purlins and girts has increased considerably,
which has made the pull--out failures more critical in
the design of steel cladding systems.
Disclaimer: Every effort has been made and all reason-
able care taken to ensure the accuracy of the material con-
tained in this publication. However, to the extent per-
mitted by law, the Authors, Editors and Publishers of this
publication: (a) will not be held liable or responsible in
any way; and (b) expressly disclaim any liability or re-
sponsibility for any loss or damage costs or expenses in-
curred in connection with this Publication by any person,
whether that person is the purchaser of this Publication or
not. Without limitation, this includes los, damage, costs
and expenses incurred if any person wholly or partially re-
lies on any part of this Publication, and loss, damage, costs
and expenses incurred as a result of the negligence of the
Authors, Editors or Publishers.
Warning: This Publication should not be used without
the services of a competent professional person with ex-
pert knowledge in the relevant field, and under no circum-
stances should this Publication be relied upon to replace
any or all of the knowledge and expertise of such a person.
CONTRIBUTIONS
Contributions of original papers or reports on steel
design, research and allied technical matters are
invited from readers are invited from readers of
Steel Construction, for publication in the journal.
The views expressed in these papers are those of
the authors and do not necessarily reflect the views
of the AISC. Additionally, the editor also invites
readers to submit letters, comments and discus-
sions on these papers, which may be subsequently
printed within a forum section of the journal.
Submissions should be in two columns in electronic
form including diagrams and equations together with
a clean camera ready printout.
 AN ENGINEERING INSIGHT TO THE FUNDAMENTAL BEHAVIOUR OF
TENSILE BOLTED JOINTS
Dr Saman Fernando
Ajax Technology Centre
76--88 Mills Road, Braeside Vic 3195
1. INTRODUCTION
The threaded fastener (nut and bolt) has played a
significant role in the industrial revolution even
though the exact date of its conception is not known.
The concept of a helical thread was first introduced
by Archimedes in the 3rd century B.C. Some
archeologists argue that the threaded fastener was in
existence even before Archimedes at the “Hanging
Gardens of Babylon”. It is accepted that the common
forms of threaded fastener assemblies have been in
existence for at least 500 years. Threaded fasteners
are probably the best choice to apply a desired clamp
load to assemble a joint, at a low cost, with the option
to disassemble if and when necessary. Furthermore,
the simplicity of its mechanism of developing and
maintaining the desired clamp force made it very
popular and it has become one of the most accepted
engineering products. In a negative sense, this
simplicity may have made some users complacent
and therefore to disregard some important issues
associated with a bolted joint.
2. BOLTED JOINT
Based on the service loads there are two types of
bolted joints. In tensile joints the bolts are loaded
parallel to the bolt axis while in shear joints the bolts
are loaded predominantly perpendicular to the bolt
axis. For example the connection of two flanges of a
pressure vessel constitute a tensile joint while the
connection of a beam to a column can be considered
as a shear joint.
In a typical shear joint the bolt acts as a shear pin. The
analysis of a shear joint is quite straightforward. The
bolt does not need to maintain a specific tensile load.
In this case the tensile load is applied only to prevent
STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001
the nut from loosening. When the shear load on the
joint changes the corresponding stress field in the bolt
also changes. Under dynamic loading this can lead to
possible fatigue failure of the bolts.
Consider however that, a shear force can be
transmitted with the help of friction forces
perpendicular to the bolt axis, which are created by
the tensile force on the bolt and friction between the
plates/bolt/nut etc. Even though the joint supports a
shear load, in this instance, it is considered as a tensile
joint or more specifically as a friction joint.
Essentially, for this joint the body of the bolt need not
touch the joining members. In a friction joint a
variation in the shear force does not cause a variation
on the tensile force on the bolt. As a result for
dynamic loading situations a friction joint will
eliminate possible fatigue failure.
In most situations involving dynamic loading the
tensile joint becomes a requirement and more
attention is needed in the design of this joint.
Therefore the rest of the discussion will be focussed
on the tensile joint.
In order to shed some light on to the behaviour of a
generic tensile bolted joint a comprehensive 3D
Non--linear Elasto--Plastic Finite Element model
analysis has been conducted. Results of this work will
be published in the near future. Although most of the
complexities of a generic tensile bolted joint can be
addressed with the above comprehensive approach,
cost and the effort requires does not qualify it as a
generic engineering tool for wider applications.
Therefore, an attempt has been made to develop a
simple analytical method using fundamental theory
and first order approximations. Although the
approach presented here can be substantiated and
calibrated by the aforementioned FEM analysis and
experimental data, the emphasis made in this
publication is to qualitatively highlight the
importance of the various critical parameters
associated with a generic tensile bolted joint.
The optimum pre--tension of a bolt in a joint has been
a subject of confusion.
3. IMPORTANCE OF PRE--TENSION
In order to emphasize the importance of pre--tension
or pre--load on the bolts in a bolted joint, the
following first order analysis based on fundamental
engineering principles is carried out. In a typical
2
bolted joint one of the main functions of the bolt is to Type A Joint:
maintain an adequate positive clamping force during
the service life of the joint in order to prevent leaks, Fa
relative movement, wear and fretting, etc. To achieve δ, k a
a particular service life requirement for a bolted joint Fc Fb
it is very important to understand the effect of bolt
pre--tension ( F i ) and the applied load ( F a ) on the
When F a = 0
Clamping force ( F c).
kc kb Fb = Fc = Fi

For simplified analysis purposes, two types of tension

joint load configurations may be considered:
Fc Fb
a) An external load is applied at the surface adjacent
to the nut and the head (Type A, Figure 1(a)), Fa

Figure 2. Mechanical representation of

b) An external load is applied at the jointed interface
Type A joint
(Type B, Figure 1(b)). These two types share the
applied load differently between the fastener and the The bolt is represented by a tension spring of stiffness
clamping force. However, in reality a typical joint k bwhile the joint is represented by a compression
will be a combination of the above Type A and Type B spring of stiffness k c. With initial pre--load ( F i) and
Joints. no external load applied, the spring system will be in
equilibrium with,
Fb = Fi (1)
Fa Fc = Fi (2)
where F b is bolt tension and F c is clamping force.
Fc
Fb Ab Eb
kb = (3)
Lb
Fc Fb
where A b is the effective stress area of the bolt, E b
Young’s modulus of bolt material and L b the
effective length of the bolt.
Fa
Ab ≅ π D2 (4)
4
where D is the nominal diameter of the bolt.
(a): Type A Joint Ac Ec
kc = (5)
Lc
where A c is the effective stress area of the joint
members, E c Young’s modulus of joint material and
L c the effective length of the joint.
If D j < D b
Fa Fc Fa
Fb A c = π (D j 2 − D h 2) (6)
4
Fa F Fb Fa
c where D j the joint diameter, D b bolt under head
/washer bearing diameter and D h the hole diameter.
If joint thickness t < 8D and D b < D j < 3D b,

 D5L + 100  2
Dj L
A c = π D j − D h  + π
2 2 b g g
−1
4 8 Db

Figure 1. (b): Type B Joint Eq (7)

3 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001

If D j > 3D b then, The Figure 3 shows the relationship between
F i, F b, F c and F a for bolted joint in the elastic range.

Ac = π
4
 Db +  −D
Lg
10
2

h
2
 (8) Load
k bF a
Fb = Fi +
where L g is the grip length of the joint. kb + kc
Fi
In general, due to the larger stress area ( A c > A b),
k c > k b (Eqn (3), (5)). The applied force F a will k cF a
generate an overall displacement δ as shown in the Fc = Fi −
kb + kc
Figure 2. This displacement imparts an additional
load of k bδ on the bolt.
The new bolt tension may now be represented by, Fa
0 Load F
F b = F i + k bδ (9) F a0
Figure 3. Variation of Fb and Fc with Fa in
The same displacement relaxes the compression
the Elastic region
force on the joint members by k cδ resulting a new
clamping force of: From this graph it can be seen that F c = 0 when
F c = F i − k cδ (10) (k b + k c)F i
F a0 = (17)
kc
For the equilibrium of forces:
This relationship proves that as high as possible
F a = ∆F b − ∆F c = k bδ + k cδ (11) pre--tension F i will provide the best load carrying
capacity for the joint. As discussed earlier, one of the
The resultant overall joint stiffness k a can be defined main functions of a fastener is to keep the joint
as; together. Therefore, it can be considered that the joint
is failed when the applied load reaches F a0 .
F a = k aδ (12)
The bolt tension when F c = 0 is (combining (15),
By substituting (12) in (11); (16) and (17)),

k aδ = k bδ + k cδ (k b + k c)F i
F b0 = = F a0 (18)
kc
ka = kb + kc (13) shall be smaller than the breaking load of the fastener.
Combining (12) and (13) For clamping load to become zero before the bolt
reaches yield;
Fa Fa
δ= = (14) Fy kc
ka kb + kc Fi < (19)
(k b + k c)
by substituting (14) in (9) and (10) respectively;
where, F y is the yield strength of the fastener.
k Fa
Fb = Fi + b (15) In order to obtain a feel for the relative magnitudes of
kb + kc the above parameters the following example is given;
3.1. Example 1:
k cF a
Fc = Fi − (16) Property Class 8.8 (Property Class X .Y is defined as
kb + kc
Ultimate Tensile Strength UTS =X *100 MPa and
Equation (15) confirms that only a component of the Proof Strength YS= 0.1* Y * UTS)
applied load is contributing to increase the tension of UTS = 8 * 100 = 800 MPa
the bolt. Typically k c is larger than k b and hence, the YS = 0.1 * 8 * 800 = 640 MPa.
increase in the bolt tension will be less than the Effective Area = 245 mm2
decrease in the clamping force. Therefore the Proof Load = 147kN
parameter k c∕k b has a significant impact on the Breaking Load = 203kN
performance of the joint. Bearing Diameter (D b) = 40mm

STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001 4

Effective Grip Length = 100mm Now, let’s investigate what happens if the fastener
Young’s Modulus = 200GPa was under tensioned. For example only 30% of the
k b (eq.(3)) =245*200/100 kN/mm yield load, Fi2= 0.3*147 = 44.1kN.
=490 kN/mm The force at separation is (eq.(18));
Joint: (k b + k c)F i2
Young’s Modulus =200GPa F a0 =
kc
Joint Diameter (D j) > 120mm
F a02 = (490+3167)*44.1/3167 = 50.9kN
Hole Diameter (D h) =22mm
Length =100mm This is a significant reduction from 147kN as
Effective Area (eq(8)) =1583.5mm2 calculated earlier. By reducing the pre--load from
k c (eq.(5)) =1583.5*200/100 kN/mm 127.3kN to 44.1kN (86.6% yield to 30% yield) the
=3167kN/mm load carrying capacity of the joint has fallen from
147kN to 50.9kN. However, it is important to notice
Note: In reality the bolt effective--length will be that the bolt will yield at a joint load of 147kN even
slightly larger than the joint effective--length. Joint though the joint has already failed by that time due to
effective area is based on using a high tensile washer separation.
on either end.
A graphical representation of the above process is
Now, k c∕k b = 3167/490 = 6.46. shown in Figure 5.
For most common applications this number is Load
between 4 and 8.
Now maximum clamp load F imax is (eq. (19)); k bF a
Fb = Fi +
Fy kc Fi kb + kc
F imax =
(k b + k c) Fb
= 147 3167 = 127.3kN k cF a
(490 + 3167) Fc = Fi −
Fb kb + kc
This gives the maximum pre--tension load for this F i2
joint as 127.3/147*100% = 86.6% of the proof load.
The corresponding maximum applied load for the Fc
separation of the joint is (eq.(18)); Fa
F a02 F a0
(k + k c)F i
F b0 = b = F a0 = 147kN Figure 5. Effect of reduced pre--load.
kc
i.e., the proof load of the fastener. This confirms that As shown in Figure 5, when the pre--load is reduced
the tensile load applied on the joint at separation will from F i to F i2 the separation load is reduced from
be equal to the load on the fastener. F a0 to F a02. In this case, when the joint has failed the
bolt is still far from its yield load. If the applied load is
The Load vs Displacement graph for the above case is
increased after separation of the joint, the total
shown in Figure 4.
applied load will then be transferred to the bolt. The
bolt load will increase at the same rate as the increase
Load of applied load. As the joint is already separated, this
F b = F i + k bδ will lead to further failure mechanisms such as bolt
bending, fretting, joint wear, fracture etc. It is now
F imax kb clear that this is not the most economical way of using
a bolt.
kc F a = k aδ
As discussed earlier in eq. (15,16) the ratio k c∕k b is
an important parameter for a bolted joint. This
F c = F i − k cδ determines the contribution of applied load to the bolt
load. The larger this factor, the smaller is the effect on
ka = kb + kc the bolt. This implies that the thinner and longer the
bolts are better it is. However, it should be kept in
mind that the load carrying capacity of a bolt is
Figure 4. Load Displacement graph in the proportional to the square of the bolt diameter and
elastic zone. therefore a reduction in diameter will have some

5 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001

negative effects. In general, a larger number of small alternating and mean stresses on fatigue life is shown
diameter bolts are better than a small number of larger in Figure 7.
diameter bolts, especially under dynamic loading.
1000
Sy
4. EFFECT OF DYNAMIC LOADING: Axial loading stress in MPa
800
Most mechanical connections are subject to dynamic S′ a ≈ 10 3
600
loads. Rotating and reciprocating machinery ≈ 10 4
generates significant cyclic loads. One of the main ≈ 10 5
failure modes associated with cyclic loading is 400 k
≈ 10 6
fatigue. The fatigue life of a bolt can be estimated by
200
a combination of S--N and S ′a vs S m diagrams, where
S′ a is the fluctuating stress, S m is mean stress and N j Su
is number of cycles for failure. This theory is well
established and reported elsewhere. Without going to 200 400 600 800 S y 1000
Sm
too much detail, a simple example is presented here in
order to highlight the importance of pre--load on the Figure 7. S′ a vs Sm curve for class 10.9
fatigue life of a bolted joint.
steel.
Figure 6 shows the effect of the peak fluctuating
stress ( S ′a ) on the lifetime of the product. The To understand the effect of pre--load on fatigue life,
lifetime is given as the number of cycles that the the following example is considered:
product can undergo before fatigue failure. As we 4.1. Example 2:
decrease the magnitude of the fluctuating stress, in
this case less than 421MPa, the life time will Bolt: M20, Class 10.9, eff. grip length 100mm
approach infinity as it will not subject to fatigue S u = 1000MPa k b = 168kN/mm
failure. Similarly, if we increase the peak alternating S y = 900MPa k c = 1050 kN/mm
stress the lifetime will decrease. In this example, as
the peak alternating stress approaches 772MPa,
F y = 203kN F a = 58 kN (mean)
failure will occur around 103 cycles. F u = 255kN F′ a = 81.2kN (peak
alternating)
In general, wind and earthquake dynamic loads will
have irregular frequency and amplitudes. Although Case 1:
the same theory stated here can be used to estimate the Pre--Load = F i = 60% F y = 121.8 kN
fatigue life the treatment would be somewhat
complicated. k bF a
Fb = Fi +
Log S′ a (k b + k c)
772
Axial loading stress in MPa F b (mean) =121.8+490*58/(3167+490)
634
= 129.6kN
517
= 100*129.6/203 % F y
421 = 63.8% F y
S m = 0.638 * 900 = 574MPa
Log N
(k b + k c)F i
10 3
10 4
10 5
10 6
F a0 =
kc
Figure 6. Peak alternating stress S′ a vs Life F a0 = (490+3167)*121.8/3167 = 140.6 kN
N cycles graph (S--N Curve) for Class 10.9
Steel F ′b = 81.2*490/(490+3167) = 10.9kN = 5.4% F y

Figure 7 shows the effect of peak alternating stress S ′a = 0.054*900 = 48.6 MPa
and the mean tensile stress on fatigue failure. As can Case 2:
be expected, when the mean stress reaches ultimate
strength ( S u ), the sample will fail without any Pre--Load = F i = 25% F y = 0.25*203= 50.75kN
fluctuating load. On the other hand, when the peak
kb .Fa
alternating stress is 772 MPa and mean stress is zero, Fb = Fi +
it will fail around 103 cycles. The effect of combining (kb + kc )

STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001 6

F b (mean) = 50.75+490*58/(3167+490) fluctuating bolt load of 80.7kN resulting a peak
= 58.5 kN = 28.8% Fy fluctuating stress S′ a of 357.3MPa. This point is
shown as “k” in Figure 7. This point is clearly within
Sm = 0.288 * 900 = 259MPa
the life span of 106 cycles limit. This implies that the
(kb + kc ).Fi bolt will fail around 106 cycles.
Fa0 =
kc This example clearly identifies importance of
properly tightening the bolts in dynamic situations.
F a0 = (490+3167)*50.75/3167 = 58.6 kN Reduction of pre--load from 60% Yield to 25% Yield
will alter the joint from no fatigue failure to fatigue
F ′b = (58+81.2--58.6)+58.6--58.5
failure.
= 80.7kN= 39.75%Fy
A similar problem may occur if the bolts are over
S ′a = 0.3975*900 = 357.75 MPa tightened. The following example shows the effect of
The above parameters for Case 1 and Case 2 are over--tightening the bolts.
shown in Figure 8. For Case 1, a pre--load of 60% F y is Case 3:
applied. The mean applied load of 58kN result in a
mean bolt load of 129.6kN. An applied load of If the initial pre--load is set up at 95% of the yield load
140.6kN will separate the joint. ( F i=192.9kN) in Case 1 of the previous example, the
bolt load at mean applied load (58 kN) is;
Load F b = 192.9+490*58/(3167+490)= 200.7kN
10.9 When the fluctuating component of 81.2kN is applied
k Fa as shown in Case 1, this will impart a fluctuating load
129.6 Fb = Fi + b
kb + kc of 10.9kN on the bolt. Now the total applied load of
121.8 80.7
211.6 kN (200.7 + 10.9) will exceed the yield load
Fb (203kN) of the bolt and will be subject to plastic
deformation. When the fluctuating component is
k cF a
Fc = Fi − released momentarily the pre--tension of the bolt is
kc + kb lost due to plastic deformation. The loss of
58.5
50.75
pre--tension makes it similar to Case 2 and leads to
failure by both joint separation and fatigue.
Therefore it is crucial that the bolt pre--tension has to
Fa be within a very specific range to achieve correct and
140.6 optimum performance of the joint.
58.0 58.6
81.2
5. TORQUE TENSION RELATIONSHIP:
Now that the importance of the bolt pre--tension is
Figure 8. Loading diagram for example 2.
established it is important to investigate how this can
The peak fluctuating component of the 81.2kN be reliably achieved.
applied load imparts a fluctuating bolt load of Torque has been considered synonymous with
10.9kN. As the total maximum applied load tension in the past with the unavailability of an
(58+81.2=139.2kN) is less than 140.6kN the joint economical and reliable bolt tensioning method.
will not separate under applied load conditions. The Several approximations has been used in the design
mean ( S m ) and peak fluctuating stress ( S′ a) for Case of bolted joints at varying success and confident level
1 are 574MPa and 48.6MPa respectively. This point in order to relate the torque to tension.
is shown as point “j” in Figure 7 and is in the area The Nut Factor approach is the most commonly used.
where no fatigue failure will occur. The simplified torque tension relationship;
In Case 2, the pre--load is only 25% F y=50.75kN.
T = KDF
This may be a result of tightening error of the bolt. As
a result the clamp separation load is reduced to only where K is the nut factor, D is the bolt diameter and F
58.6kN. That means when the joint load of is the bolt tension.
58+81.2kN is applied, the joint will separate (this
may be considered as failure) and the load excess of This formula can be further expanded to;
58.6kN will be transmitted directly to the bolt. Under T = F D (K 1 + K 2 + K 3)
the mean load of 58kN, the bolt will experience a
mean load of 58.5kN resulting a mean stress S m of where K 1, K 2 and K 3 are contributions due to bolt
259MPa. The fluctuating load of 81.2kN imparts a stretch, thread friction and under head/nut bearing

7 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001

friction respectively. The following chart describes the fraction of torque spent on overcoming thread
these parameters. friction and K 3D the fraction of torque spent on
overcoming under head friction.
K
For a M12 bolt;
Pitch p = 1.75mm
K1 K2 K3
Thread friction μ t = 0.15
D p μt D rt α D μb rb Thread radius r t = 6mm
Thread angle α = 30degrees
Under head friction μ b = 0.15
Using energy balance principals a first order Effective under head radius = 8mm
relationship between the torque (T) and Tension (F) Now;
can be derived as follows; T = F (0.28 + 1.04 + 1.2)
In a rotation of the nut by δθ; K 1 : K 2 : K 3 = 0.28 : 1.04 : 1.2
work done by torque = T δθ = 11:41:48 %
work done by tension = Fp δθ∕2π From this simple analysis it is evident that typically,
work done by thread friction around 10% of the effort is going to the stretch of the
= Fr t μ t δθ∕ cos α bolt and compression of the joint, 40% of effort is
work done by under head friction going to overcome thread friction and the remaining
= Fr b μ bδθ 50% is going to overcome bearing friction. This
where, p thread pitch, α thread flank angle, r t implies that approximately 90% effort is going to
effective thread radius, r b effective bearing radius, overcome friction while only 10% is doing useful
and μ t and μ b are thread and bearing friction work.
coefficients respectively. There are a large number of parameters such as,
surface finish, hardness, lubricants, among other
Now for energy balance;
things, that can alter the friction coefficients
Fp rtμt associated with a bolted joint. A 10% reduction in
Tδθ = δθ + F cos α δθ + Fr b μ bδθ friction contribution (from 90% -- 81%) will increase
2π
the bolt stretch--joint compression contribution from
T = FD  p
+
rt μt
2πD D cos α
r μ
+ b b
D
 10% -- 19% which is a 90% increase.
As such, it shall be understood that the torque tension
relationship is not a reliable way of ensuring the
T = FD(K 1 + K 2 + K 3) tension of the bolt under most situations. The value of
Dh K can vary from approximately 0.2 to 2.0 depending
on the condition of the bolt and the bolted joint.
D
Any irregularity or damage to the thread can also be
seen as an increased friction hence adding to the
overall variability of the friction coefficient.
Therefore, if the torque is used as a measure of
rb
tension it shall be made sure that the thread is in
perfect shape. Galling of threads can also contribute
to significantly large friction forces.
rt
Torque tension scatter varies largely with the size of
the bolt, coatings, interface friction, and joint
geometry. With large bolts (>M30) this scatter may
be as high as 300%.
Figure 10 shows the torque vs tension relationship
measured for M8, property class 8.8, Zinc electro
plated bolts and nuts (without any lubricant)
p μt rt μ r tightened on the same joint. Each bolt assembly is
K1 = , K2 = , K3 = b b used only once. The solid line shows the theoretical
2πD D cos α D
relationship between tension and torque assuming
Figure 9. Parameters associated with typical friction values for Zn coated interfacing
Torque--Tension Relationship. surfaces. The spread between the six samples are
quite significant. The recommended assembly
Term K 1D represent the contribution of the torque torque for the above bolts is 15.4Nm to achieve a
towards bolt elongation and joint compression, K 2D tension of 13.8kN which is 65% of the proof load. At

STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001 8

15.4 Nm torque the six samples achieved tension
values from 11 to 17 kN. The spread of 6kN is a 43%
variation on the desired tension value. If a 90% of the
proof load was desired (19.1kN) the torque values
from 17.5Nm to 33Nm were required to achieve the
desired tension on different bolts. If a 33Nm torque is
applied to each bolt, that would have failed several of
the above bolts!!!
Figure 11 shows the first and subsequent four
30.0
Tension (kN)
25.0
tightening of the Sample 2 bolt in the above
experiment. For a tightening torque of 15.4Nm,
tension values from 6 to 13kN were achieved
depending on how many times the bolt was tightened.
Again the spread on the desired tension is over 50%.
The above figures are typical for all bolt sizes,
however, the large bolts will have greater variations
in the torque tension relationship.

20.0
Theory
Sample 1
15.0 Sample 2
Sample 3
Sample 4
10.0 Sample 5
Sample 6
5.0

0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
Torque (Nm)

Figure 10. Torque Tension Relationship; M8, Class 8.8, Zinc plated bolts, first tightening six
samples. As plated, no lubrication. Proof Load 21.2kN, Breaking Load 29.2kN.
30.0

Tension (kN) 25.0

20.0
Theory
15.0 Series1
Series2
Series3
10.0
Series4
Series5
5.0

0.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0
Torque (Nm)

Figure 11. Repeated tightening of the above Sample 2 for five times.
6. CONVENTIONAL METHODS OF
TIGHTENING
Although, most of the practitioners understand the
importance of bolt tension in a bolted joint the
conventional tightening methods only provide a
vague indication of the bolt tension. Extensive
research carried out on torque--tension relationships
prove that under most uncontrolled situations using
torque as a measure of tension can lead to a error as
large as ±50%. Even under controlled conditions
torque on its own is not a reliable measure of tension.
On the other hand, the reliable tension measuring
systems are cumbersome and expensive. A
comparison of various methods available for
achieving pre--load in terms of their reliability and
relative cost are shown in Figure 12.

9 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001

 % ERROR /
0 5 10 15
Ordinary Spanner
Torque Wrench
Torque Control Bolt
Turn of the Nut Method
Torque Angle Signature
Load Indicating Washers
Heating the Bolt
Hydraulic Bolt Tensioning
Ultrasonic
Mechanical Tension Indication
Strain Guages
Smart Bolt System
Figure 12. Comparison of common tightening methods
Another commonly used method is the “turn of the
nut” method. This method, in fact, is the
recommended method by Australian Steel Codes
AS4100. In this method the nut is tightened to a “snug
tight” position and then tighten further fraction of a
turn depending on the joint geometry. However, the
standard does not specifically define the “snug tight”
position. According to the theory, the “snug tight”
position is where a step change in the gradient of the
torque vs angle curve occurs. In order to carry out this
process with any accuracy a torque sensor and an
angle encoder shall be used. Then by calibrating on
the desired joint with a direct tension measureing
device the required nut rotation after the snug tight
position can be determined. This will then provide a
method of tightening with some degree of accuracy.
However, for example, for tightening slew--ring bolts
this method may not be suitable. Snug tight position
for one bolt may change with the tightening of the
remaining bolts hence making this method not
reliable at all.
Load Indicating Washer (LIW) is another common
method of assuring desired tension. Again this
method will not give satisfactory results for
tightening slew--ring bolts as the firstly tightened
bolts will become loosen when tightening the
subsequent bolts. LIWs are capable of indicating the
tension only in their first tightening.
Hydraulic Bolt Tensioning is relatively popular in
heavy industries due to its simplicity. However, when
using this method, the bolt tension is known only
when the hydraulic pressure is applied (by measuring
the hydraulic pressure). Once the hydraulic pressure
is removed and the load is transferred from the jack to
the nut, bolt and the tightening flanges, the applied
tension on the bolt is relaxed. In one experiment
carried out by ATC on a large rock crusher showed
that at a 8bar hydraulic pressure applied on three M64
Flange bolts resulted in a 80 --90% of proof load
initially and then relaxed to 45 – 56% proof load once
STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001
RELATIVE COST
20 25 30 35 40
% Error
Relative Cost
the pressure is removed. It was found that the
relaxation is a function of bolt/nut size, number of
threads engaged, and the flange dimensions among
other parameters. However, it was not possible to
establish a firm relationship with the applied
hydraulic pressure and the final bolt tension.
Heating the bolts at tightening is another method of
obtaining a pre--load. The bolt is heated to a known
temperature with the help of a concentrated heat
source. This will elongate the bolt. The nut is
tightened at this point and let the joint to cool down.
The shrinkage of the bolt will impart a pre--tension on
the joint. This method is not very advisable for
tightening high tensile bolts as the tensile strength of
the bolt may be significantly affected by heating the
bolt. Control of the bolt temperature is extremely
difficult and large variations in temperature may be
observed over the bolt. This may result in variable
material properties over the bolt.
Another factor that affects the bolt tension is the joint
temperature. Especially if the joint is made of
dissimilar materials, the differential thermal
expansion of the bolt and joint materials will cause
variations in the bolt tension. Even if the bolts are
tightened accurately to the desired tension value, if
the joint is subject to temperature variations the
working tension on the bolt may change. It is not
possible to theoretically estimate these changes to a
sufficient accuracy. Only direct tension
measurement will provide the engineer with the real
time working tensions on such bolts.
7. LOADING BEYOND YIELD
In most situations the pre--tension load will be less
than the yield load, however, due to the applied loads
the bolt exceeds yield load. The work hardening
materials generally exhibit elastic –plastic behavior.
For simplicity, perfectly linear--elastic and perfectly
linear plastic behavior of the bolt material is assumed.
The typical stress strain relationship for such material
10
is shown in Figure 13. The gradient of the elastic zone load exceeds the previously applied maximum load
is the Young’s Modulus ( E) and the gradient of the the bolt act as an elastic bolt but with a reduced
plastic zone is the plastic modulus ( E p). For all pre--load ( F i2). However, if the load exceeds the
elastic plastic materials E p < E. Once the material previously applied maximum load, then, once the
reaches plastic stage when the applied load is load is removed the pre--load is further reduced.
removed the material will not return to its original The load displacement diagram for this case is shown
shape. The plastic deformation is as shown in the in Figure 15.
Figure 13.
k bp
Based on the previous analysis therefore, Fb
k bp < k b < k c.
Fi kb
σ
Ep F i2 k ap= k bp+k c
kc
Fa

Áp Fc
E δ
Á
Figure 15. Load displacement curve
Áp beyond yield
Figure 13. Perfect Elasto--Plastic behaviour
8. BOLTING TO YIELD
If the pre--load is close to the yield point once a cyclic
external load F a is applied, the relationship between In some occasions in order to get the highest
F i, F b, F c and F a is as shown in Figure 14. Note the clamping force the bolts are tightened beyond yield.
Especially, when relying on torque to tension the bolt,
variation in gradient once the yield point is passed.
a better certainty can be achieved when the bolts are
The gradient of the plastic zone of F b vs F a curve is;
tightened beyond the yield, as the effect of variation
kbp kb in torque on the tension is lower. This has to be done
Gbp = < Gb = carefully, especially with high tensile bolts where the
(kbp + kc ) (kb + kc ) separation between the yield and the failure load is
Similarly the negative gradient of the F c vs F a curve relatively narrow. Furthermore, if a bolt is yielded it
in the plastic zone is; shall not be reused, as this will alter the geometry and
kc kc the mechanical properties of the bolt.
Gcp = > Gc = In the authors opinion this method is suitable for
(kbp + kc ) (kb + kc )
friction grip joints where the joint is loaded in shear
Y2 and there is no possibility of extra tensile load applied
Y1 on the bolt.
F by G bp
Fb This method shall not be used to tighten the bolts if
there is any uncertainty on whether any additional
Fi Gb dynamic or static tensile load will be applied on the
F i2 joint during its life span.
Gc
Fc The relationship between F i, F b, F c and F a for a joint
tightened beyond yield is shown in Figure 16. This is
very similar to the Figure 14, without the elastic zone
in the bolt tension. When unloading the joint the bolt
G cp tension follows a line parallel to the elastic line and
Fa
end up with a residual plastic displacement which
leads to a reduction in pre--tension to F i2.
Figure 14. Bolt tension and clamping force G bp, G b, G cp and G c are the gradients as defined
variation crossing yield earlier.
As k bp is always smaller than k bG bp will always be
Once the applied load F a is removed the pre--tension
in the joint will not reach the original value F i as smaller than G b. Similarly, G cp will be larger in
shown in the Figure 12. Due to work hardening of the magnitude to G c.
material the new yield point is now moved from Y1 to As shown in Figures 16 and 17 if the applied load is
Y2. In a subsequent loading event unless the applied increased and then reduced in a joint where the bolts

11 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001

are tightened to yield the bolts will lose pre--load due
to plastic deformation. Therefore, tightening to yield
is not suitable for joints where the bolts may be
subject to additional tensile loads.
F Nuts, serrated washers, spring washers, Cottor pins,
Fb
Fi G bp Gb
G cp
F i2
Gc
Fc
Fa
Figure 16. Bolt tension and clamping force
variation beyond yield
F confirmed that if a typical bolted joint is tightened to a
k bp
Fi
kb F a=k aδ
F i2
k a=k b+k c δ
δ2
Figure 17. Load displacement curve
beyond yield
9. VIBRATION LOOSENING:
It is common experience that some bolts will be
loosened when subject to vibrations and dynamic
loading. There were several attempts to understand
the mechanism of vibration loosening. A large group
of researchers believe that the mechanism is
somewhat similar to that of vibratory bowl feeders
and vibratory conveyors. In general, it is a
combination of the inertial forces generated by
particular vibration, and friction forces.
Through proper design it is possible to develop
mechanical systems where the bolts have a tendency
to be tightened under applied loads (eg., some Lawn
Mower Bolts, Wheel Nuts). On the other hand,
certain situations promote vibration loosening either
due to lack of consideration at design level or due to
mere complexity of a particular joint. Especially
when the joint incorporates soft gasket materials
STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001
and/or different member materials the complexity
increases significantly.
There are several devices available in the market to
prevent vibration loosening. Lock nuts, Nylok
to name a few. All of these devices provide additional
friction force or interlock to the bolt/nut. Depending
on the nature of vibrations and other conditions there
will be a finite resultant loosening torque that the
fastener system has to resist. The said devices help
resisting this torque.
The fact that 90% of the applied torque is going to
overcome frictional forces, as discussed earlier, may
be of value to prevent vibration loosening. The
frictional torque is directly proportional to the
remaining tensile load on the bolt. Therefore, if the
bolts are tightened to a particular pre--load, in such a
way that the remaining tension on the bolt under
applied load is adequate to generate a friction torque
larger than the loosening torque vibration loosening
will not occur. The tests carried out by ATC
pre--load higher than 65% of the yield load of the
fastener vibration loosening will not occur even
under severe vibration conditions.
There may be some special occasions where it may
not be feasible to apply such pre--loads to the joint. A
suitable anti--loosening device may be used in such
situations.
10. CONCLUSIONS:
A simple approximate analytical approach is
presented. This may help engineers to better
understand the performance of bolted joints.
In general, the following specific conclusions are
made;
1. The contribution of applied load on the bolt
load in a pre--loaded tensile joint depends on
the stiffness ratio of the bolt and the joint.
2. Large number of slender bolts is better than a
small number of large bolts for a tensile joint
subject to dynamic loads.
3. If feasible, longer bolts provide better proper-
ties for a dynamic joint than shorter bolts.
4. Correct pre--load (pre--tension) is paramount
in achieving high--performance dynamic ten-
sile joints.
5. Calibrated torque wrench is not a reliable
method of achieving a desired bolt tension.
6. Common Hydraulic bolt tensioning methods
does not adequately account for the relaxation
of the joint.
12
7. LIWs are not suitable for most group--bolted F a02 Applied Force to Separate the Clamp after a
joints. Tightening sequence and process is plastic cycle
critical to optimize the use of LIW. F b Bolt Tension Force
8. Heating the high tensile bolts may result in F b′ Peak Fluctuating Bolt Tension Force
poor quality joints.
F b0 Bolt Tension at Separation of the Clamp
9. Tightening to yield is not suitable for tensile
F c Clamping Force
joints.
Fi Initial Bolt Tension – Pre--load
10. No anti--loosening devices are necessary if the
F imax Maximum Pre--Load
bolts are tightened to at least 65% of the yield
load. F i2 Pre--load remaining after a plastic load cycle
11. Most joint failures are due to insufficient pre-- F y Yield Load of the Bolt
load in the bolts. G Gradient of Graph F b, F cvs F a
K Nut Factor
11. BIBLIOGRAPHY K1 Nut Factor due to bolt stretch
Bickford, John H., An Introduction to the Design and K2 Nut Factor due to thread friction
Behaviour of Bolted Joints, Marcel Dekker, Inc, New K3 Nut Factor due to Under head/nut Friction
York 1990.
k Stiffness
Juvinall, Robert C., Fundamentals of Machine
ka Joint Stiffness
Components Design, John Wiley & Sons, New York,
1983. kb Bolt Stiffness
Ziada, H.H., Abd--El--Latif, A.K., “Areas of contact kc Clamp Stiffness
and pressure distribution in gasketed bolted joints”, L Length
1E(1) Journal--ME, Vol 62, November, 1981, pp Lb Effective Bolt Length
77--83.
Lc Effective Clamp Length
Finkelston, R.J., Wallace, P.W., “Advances in
High--performance mechanical fastening”, Paper Lg Joint Grip Length
800451, SAE, New York, N.Y., Feb 20, 1980. p Thread Pitch
Fisher, J.W., Struik, J.H.A., Guide to Design Criteria r Radius
for Bolted and Riveted Joints, John Wiley and Sons, rt Thread Effective Radius
New York, 1974. rb Under Head/Nut Effective Radius
Gorenc, B., Tinyou,R., Syam, A., Steel Designers’ S′ a Peak Fluctuating Stress
Handbook, 6th Edition, UNSW Press, 1996
Sm Mean Stress
Australian Standard AS4100--1998, Steel Structures. Su Ultimate Tensile Stress (UTS)
Sy Yield Stress (YS)
12. NOMENCLATURE
T Torque
Ab Effective Stress Area of Bolt Greek Symbols;
Ac Effective Stress Area of Clamp δθ Infinitesimal Bolt Rotation
D Diameter of Bolt Shank δ Elongation
Db Bearing Diameter of Bolt Head ∆ Variation
Dh Hole Diameter μ Friction Coefficient
Dj Effective Diameter of Joint μ b Under Head/Nut Friction Coefficient
E Young’s Modulus of Elasticity μt Thread Friction Coefficient
Eb Young’s Modulus of Elasticity Bolt α Flank Angle of the Thread
Ec Young’s Modulus of Elasticity Clamp Indices:
F Force p Plastic State
Fa Applied Force b Bolt
F a′ Applied Peak Fluctuating Force c Clamp
F a0 Applied Force to Separate the Clamp i Initial

13 STEEL CONSTRUCTION VOLUME 35 NUMBER 1, MARCH 2001