Bolted Joints and Threaded Fasteners

Page 1 of 14

SEED Guides Bolted Joints and Threaded Fasteners < >

3. Design Procedures for Statically Loaded Bolted

Joints
Design Flow Chart for Statically Loaded Bolted Joints
Figure 3.1 sets out the principal steps in selecting a threaded fastener to sustain a steady
loading. Like any design procedure, it cannot cover every case and eventuality; nor can it
be followed blindly, and the user must exercise thought and judgement during the
process. It may be necessary to iterate some of the steps before a satisfactory solution is
obtained.

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

http://www.co-design.co.uk/dpg/bol/bol3.html

Page 2 of 14

09/02/2009

Bolted Joints and Threaded Fasteners

Page 3 of 14

Table 3.1 Bolt Loads & Torques

THREAD Maximum Axial force (kN)
SIZE
(MM)

For Bolts/Screws of Grade *

8.8
10.9
12.9
See Note 1

Torque to Produce
this Force (Nm)
8.8
10.9
See Note 2

M3

2.3

3.25

3.9

1.3

1.8

2.1

M4

4.0

5.65

6.75

2.7

3.8

4.6

M5

6.55

9.2

11.1

5.5

8.0

9.5

M6

9.25

13.0

15.6

9.5

13.0

16.0

M8

17.0

23.9

28.7

23.0

32.0

39.0

M10

27.1

38.0

45.7

46.0

64.0

77.0

M12

39.5

55.5

66.7

80.0

110

135

http://www.co-design.co.uk/dpg/bol/bol3.html

12.9

09/02/2009

Bolted Joints and Threaded Fasteners

Page 4 of 14

M14

54.0

76.0

91.3

125

180

215

M16

75

105

126

195

275

330

M20

117

164

197

385

540

650

M24

169

237

284

60

930

1100

M27

221

311

374

980

1740

1650

M30

269

379

454

1350

1850

2250

M36

393

553

664

2280

3210

3850

Note 1
The Maximum Axial Force is based on not exceeding a combined stress of 90% of the
yield stress (or 0.2% proof stress for materials which do not exhibit a yield point) acting
on the previously defined stress area.
Note 2
The torque to create the maximum axial force is based on a coefficient of friction of
0.125 at the thread and the bolt head (or nut face). The value for friction can vary widely
in practice. (See table 5.1) Equation 5.1 or 5.2 enable torques for other frictional values
to be calculated.
* For an explanation of the material grading system, see Section 6

Figure 3.1 Flowchart for statistically loaded bolted joints

3.1 Having ascertained the maximum load that the joint is required to carry, it is
necessary to apply a factor of safety to this load to ensure that the constituent parts of the
joint remain in contact with each other when the load is applied. If the bolts are
considerably less stiff in the axial direction than the remainder of the joint (which is
normal) and if the bolts were to be tightened only to the value of the load to be carried,
then there will be little residual contact pressure when the joint is loaded. Section 3.6
explains this situation mathematically and in greater detail. Consequently, it is usual to
apply a factor of safety to the calculated load at the joint. A factor of safety of 2 is
recommended. This load is considered to be the "minimum preload at the joint".
3.2 The method of achieving the desired bolt load must be chosen next. Table 3.2 (page
15) describes various methods and their associated accuracies. By choosing a method and
considering the associated tolerances, a "maximum preload at the joint" can be
calculated.
3.3 Configuring of the joint in terms of the number of bolts, their style, their size and
their disposition is an iterative process which may require a number of attempts before a
satisfactory solution emerges. One approach is to develop a spreadsheet which considers
one or more styles and creates a combination of numbers of bolts versus their diameter. If
pressed, the user may include a revision of the tightening method into the iteration loop
as well in order to reduce the maximum bolt load. Reductions in the factor of safety

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 5 of 14

should be avoided unless of a trivial nature (say 5%).

In terms of bolt pattern/disposition, the user is advised to study the issues discussed in
Section 3.5.2 and shown in Figure 3.4. The main issue here is that the fasteners need to
be as close to the line of action of the load as possible and the more flexible the joint
components are the more important this issue becomes. Similarly, if the joint components
are comparatively flexible then the proximity of fasteners to one another needs to be
decreased. Section 3.5.2 and Figure 3.5 explain this issue in more detail.
3.4 For most initial design purposes, the decision can be made to conclude the bolt
selection at this intermediate "finish" point on the flowchart, where reference is made to
this section. For final design work the next steps of the flowchart should be completed
taking the user to the second '"finish" point. Experience with use of the method will
enable the designer to make better informed judgements concerning the need to proceed.
In many instances, where the joint is stiffer than the bolts by a considerable margin, the
remaining steps may serve to confirm the previous design and no revision will be
necessary. However, until the designer has acquired that level of experience, caution
should be exercised and the full procedure undertaken.
3.5 Estimation of Stiffness of the Bolt and Joint Members
The stiffness (k) of any axially loaded member is given by the equation:

where: A = loaded cross sectional area; E = Young's modulus; l = loaded length

Thus, in order to achieve the desired high ratio of joint stiffness to bolt stiffness the
following ratios should be maximised:

3.5.1 Length Ratio

In the case of through-drilled joints clamped by a nut and bolt, the clamped length of the
joint and the grip length of the bolt are clearly identical. However, since the tensile stress
in the bolt does not fall immediately from its full value to zero at the loaded faces of the
bolt head and nut, it is usual to consider the effective length of the bolt to be slightly
greater than the actual grip length. This has the beneficial effect of slightly reducing the
bolt stiffness. An addition of 0.4 times the local shank diameter at each end is suitable see Figure 3.2

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 6 of 14

Where the bolt is screwed into a tapped hole in one of the joint members, the latter is
clearly not wholly in compression. This tends to reduce the stiffness of the joint
compared to the bolts. The following effective lengths are suggested:

Here, the effective joint length, lje is greater than the effective bolt length, resulting in a
less favourable stiffness ratio than for the case of a bolt in a through-drilled hole.
3.5.2 Area Ratio
The designer often has considerable freedom in selecting the cross sectional areas of the
bolt and joint members.
The bolt area will be determined by strength considerations. With a standard bolt, the
critical section is usually the thread root diameter, which is less than the nominal bolt
diameter, and is also subject to stress concentrations. The bolt shank may therefore be
waisted down or drilled axially to give a cross sectional area somewhat below the thread
root area, without loss of strength. A small length of the full shank diameter may have to
be retained near the joint interfaces for lateral location purpose, but the resulting bolt will

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 7 of 14

be more resilient than a standard one, and will therefore attract a smaller proportion of
any load applied to the joint. With a waisted bolt, it is also possible to increase the
underhead radius and thus relieve the stress concentration in this region. Both measures
increase the effective fatigue strength of the bolt which is important when the joint load
has a dynamic component.
The joint area should always be as large as possible, provided the joint members can be
made stiff enough to ensure reasonable parity of pressure over the whole joint interface.
It is pointless to employ large, thin flanges which will distort under the clamping load
and result in a concentration of load around the bolts, but little elsewhere. (See Fig 3.4)

For design purposes, we must try to estimate an effective area for use in calculating the
joint stiffness. If the diameter of the joint faces Dj, does not exceed the bolt head or
washer diameter DH, the compressive load in the joint may be assumed uniform, so that:

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 8 of 14

For irregular shaped joints, the joint diameter Dj may be estimated by drawing lines, at
452 to the bolt axis from the collar diameter of the bolt or nut to the joint interface, (see
Fig 3.5)

In order to ensure effective clamping of the joint by the bolts in a high duty connection,
the size and number of bolts should be chosen so that the joint diameters of adjacent bolts
touch or overlap slightly.
3.5.3 Modulus Ratio

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 9 of 14

It is therefore desirable in the interests of overall joint stiffness to employ confined

gaskets or '0' rings for sealing and allow the jointed components to abut directly, in
preference to using interposed gaskets (see Figure 3.6). (a) Confined gasket, giving stiff
joint. b) Interposed gasket, giving low joint stiffness .

3.6 Preloading of Bolted Joints

3.6.1 Behaviour of a Bolted Joint under a Varying Axial Load
The three diagrams of Figure 3.7 show a bolted joint: firstly when the nut has been hand
tightened so that the components are snug but before significant load is applied, secondly
when the nut has been tightened to produce a pre-load or pre-tension P in the bolt, and
thirdly under the subsequent influence of an external load W applied to the joint.
When the pre-load is applied, the bolt stretches and the joint members compress under
the influence of forces applied at the nut and bolt faces. The compressive load applied to
the joint is clearly equal to the bolt tension P, so that the extension of the bolt relative to
the amount by which the joint members compress will be in inverse proportion to their
relative stiffnesses. Since the joint members are usually stiffer than the bolt, the amount
of joint compression, j, is usually less than bolt stretch, b.

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 10 of 14

The result can be plotted on a load/extension diagram as shown in Figure 3.8. It can be
seen from this diagram that the external load W is shared between an additional load on
the bolt and a reduction in compression of the joint, in proportion to their stiffnesses.

Thus by designing the joint members to be very stiff and the bolt to be resilient, and
applying a high preload to the bolt so that the joint faces do not separate when an external
load is applied, a large external load can be seen to cause a small increase in the bolt load
but a much larger decrease in the joint load. This has important implications later when
fatigue is considered on joints with fluctuating loads but, in the context of steady loads
applied and removed relatively infrequently, it is important to note the need for a preload
to prevent joint faces separating and also that the increase in the bolt load is likely to be
very small.

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 11 of 14

Figure 3.8 Load/extension diagram for a pre-loaded joint with an extension applied load
3.6.2 Choice of Preload
Having established that it is desirable to preload a bolted joint that is subsequently
subjected to a tensile load, we must now decide how much preload to apply. On the one
hand, we must ensure that the joint faces will not separate under the application of an
external load to the joint, and on the other we must not apply such a high preload that the
bolt will fail when we tighten it up. Let us define the minimum preload to avoid
separation of the joint members as Pmin and the maximum preload which can be applied
without causing failure of the bolt in service as Pmax
From Figure 3.8, it can be seen that the safety factor on separation of the joint faces is
equal to:

In order to ensure that the joint members do not separate, it is prudent to apply a high
safety factor to the condition of joint parting. A figure of at least 2 is suggested:

In practice, Wj is not usually known at the initial design stage, so a first assumption of
Pmin 2W can be used. We have now therefore specified a lower limit for the preload,
Pmin. The upper limit of preload must be such that the bolt will not fail, either during
tightening or under subsequent external load application. It is suggested that a maximum
value of preload, Pmax, is made equivalent to a nominal tensile stress at the minimum bolt
cross section of 85-90% of the bolt material yield strength is a good initial choice.
METHOD

ACCURACY REMARKS

Yield Control 5%

Yield is detected during tightening by continuous

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 12 of 14

monitoring of torque (T) and angle of turn of the nut (0)

and noting the reduction in dT/d0 which occurs at yield.
Complex equipment is required. A similar accuracy may
be achieved using a proprietary load-indicating bolt,
known as the 'Rotabolt'
Angle Control 10% at best The nut is tightened hand-tight and then turned through
(Turn-of-thea further specified angle, calculated to produce the
nut)
required strain in the bolt. The hand-tight condition is
difficult to control accurately unless the joint members
are stiff and flat.
Direct
10% at best
Measurement
of Bolt
Stretch

The extension of the bolt on tightening is measured and

compared to a specified value. Both ends of bolt must
be accessible for measurement, which must be
accurately carried out since extension is normally small.
(Typical strain 0. 1 %).

Torque
Wrench:

The nut is tightened to a specified torque. The resulting

tension produced in the bolt varies considerably with
thread and collar friction. A torque wrench setting is
therefore worthless unless Strain accompanied by an
indication of whether assembly should be dry or
lubricated. Additional errors will result from lack of
consistency of the wrench and any inaccuracy in
calibration.

Strain
Calibrated
Torque
Calibrated
Uncalibrated

10-15%
15-25%
25-50%

Air-stall
Power
Wrenches
Calibrated

25-40%

Uncalibrated

30-100%

Air operated power wrenches in which the spanner :

socket carrier is designed to stall at a pre-set torque.
Used for automated assembly in mass production.
Maintenance of accuracy depends upon frequent
recalibration against a known standard.

Table 3.2 Methods to achieve a specified preload in threaded fasteners

3.6.3 Control of Preload
A permissible range of preload, Pmin to Pmax has now been established, based on criteria of
joint parting and bolt strength. The tightening of the nut must be controlled and
monitored to an accuracy which will result in a preload somewhere within this safe
range. If a preload equal to the mean of Pmax and Pmin is specified, the permissible error in
preload control is (Pmax - Pmin)/(Pmax + Pmin). The choice of the method used to control the
tightening of the nut therefore influences the required strength of the bolt, and indeed the
design of the complete bolted assembly.

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

Page 13 of 14

Table 3.2 gives a guide to the accuracy of preload control offered by a range of
commercially available methods, together with comments on their limitations. In
selecting the method to be employed, the designer should be sure that his intentions are
within the capabilities of the service environment, as well as his assembly shop, if the
fasteners concerned will require periodic re-tightening in service.
3.6.4 Loss of Preload
The initial preload applied to the bolt tends to be degraded by subsequent application and
removal of the external load on the joint. This is due to embedding of the contacting
surfaces and flattening of surface roughness, and tends to occur rapidly with the first few
load applications, and more slowly thereafter. The effect is most pronounced with soft
materials and poorly finished surfaces.For design purposes, the amounts of embedding
shown in Table 3.3 may be taken as representative for pairs of steel surfaces.
Type of surface

Machined
Thread

Formed
Thread

Machined
Surface

Ground
Surface

Amount of embedding at
each interface (mm)

.005

-003

.004

.002

Table 3.3 Typical embedding at interfaces of a bolted joint (steel)

For aluminium surfaces, the figures quoted should be trebled. For soft gasket materials
(which may also bed down due to heat) the amount of embedding may be much larger.

http://www.co-design.co.uk/dpg/bol/bol3.html

09/02/2009

Bolted Joints and Threaded Fasteners

http://www.co-design.co.uk/dpg/bol/bol3.html

Page 14 of 14

09/02/2009