Joint Stiffness

A typical joint is composed of two components, the fastener and

the members. Each has a stiffness that contributes to the overall
stiffness of the joint, and are identified in the figure. Members

Fastener Stiffness

The fastener generally consists of two distinct sections, the

Fastener
threaded and the unthreaded. The overall stiffness of the fastener
is determined by modeling each segment as a spring and noting
that they are springs in series, so the overall stiffness is determined from the relation

1 1 1
= +
k k1 k2

Definitions of the fastener length (L), the grip length ( LG ), overall thread length ( LT ), threaded length within
the grip region ( lt ), fastener diameter (d), washer thickness (t or tw ), etc. is given in Table 8-7 (p 462) of the
text for through-the-thickness fasteners (side a) and non-through-the thickness fasteners (side b). For
fasteners with long threaded regions the stiffness of the threaded region is

At E
kT =
lt

The stiffness in the unthreaded region of a fastener is given by

Ad E
kd =
ld

In these two expressions:
At = tensile-stress area of the fastener (Tables 8-1 and 8.2 on pp 448 and 449)
lt = length of the threaded portion of the grip
Ad = major-diameter cross-sectional area of the fastener
ld = length of the unthreaded portion in the grip
E = elastic modulus of the bolt material

Therefore, the estimated effective stiffness of the fastener in the clamped region of the joint ( kb ) is given by

At E Ad E At ld E + Ad lt E
+
1 1 1 kT + kd lt ld lt ld A l E + Ad lt E At ld + Ad lt
= + = = = = td =
kb kT kd kT kd   
At E Ad E  At Ad E 2
At Ad E 2 At Ad E
  
 lt   ld  lt ld

As a result

At Ad E
kb =
At ld + Ad lt
Member Stiffness

The stiffness of the member is determined by considering the effective spring rate of each component of the
member. For joints with multiple members, this accomplished by considering a number of springs in series.
For a joint consisting of n members we have

1 1 1 1
= + +… +
km k1 k2 kn

The general expression for the stiffness of a frustrum is given by

π Ed tan α
k=
( 2t tan α + D − d )( D + d )
ln 2α
( 2t tan α + D + d )( D − d )
D
It is generally accepted that α = 30o should be used for
most cases. As a result

0.577π Ed t
k=
ln
( t + D − d )( D + d )
1.15
(1.15t + D + d )( D − d )
This equation is solved for each frustrum in a joint and d
the results are combined in order to define the member
stiffness km . For example, assume a joint with two members. The total thickness of the joint is h = t1 + t2 . In
order to identify the number of frustra in the joint;

1. Locate the mid-plane of the

joint at h / 2 .
o
2. Construct two equilateral 60 washer
frustra 1
triangles, one above the
mid-plane and one blow the
1
mid-plane.
3. The washer diameter is t1 h/2
generally taken as a
reference point of the top 60 o
and bottom surfaces of the h = t1 + t 2 60 o
joint from which
t2 frustra 2
construction of the triangles 2
is begun. A general rule of
thumb is that the washer frustra 3
diameter is 1.5 times the
bolt diameter d w = 1.5d .
o
4. Count the number of frustra 60
that exist. In the case
shown to the right, there are three. Frustra 1 is above the mid-plane, frustra 3 is below the mid-plane,
and frustra 2 is the region where the two equilateral triangles overlap. The dimensions from each of
these regions is used to define the stiffness of the individual frustra, and subsequently the stiffness of
the joint.
For the upper frustra (frustra 1) the stiffness is defined d w = 1.5 d
from the relation

0.577π Ed 60 o
k1 = washer
ln
(1.15t + D − d )( D + d )
(1.15t + D + d )( D − d ) 1
t1 h/2
where D is the diameter at the top of the frustra
d
( D = d w = 1.5d ), d is the bolt diameter and t is the
thickness of the upper frustra ( t = t1 ). Using these
diameters and thickness, and noting that the elastic
t1 tan 30 o
modulus for frustra 1 is E = E1
D1
0.577π E1d 0.577π E1d
k1 = =
ln
(1.15t1 + d w − d )( d w + d ) ln
(1.15t1 + 0.5d )( 2.5d )
(1.15t1 + d w + d )( d w − d ) (1.15t1 + 2.5d )(0.5d )

We note that the diameter D1 shown in the figure is D1 = 2t1 tan 30o + d w = 1.15t1 + 1.5d . This is used as D in
the computation of the stiffness for the middle frustra (frustra 2). The thickness of the middle frustra is
t = (t2 − t1 ) / 2 . For the middle frustra we use E = E2 , so the stiffness of the middle frustra is given by

D1
0.577π E2 d
k2 =
ln
(1.15 (t2 − t1 ) / 2 + D1 − d ) ( D1 + d ) 1
( t2 − t1 )
(1.15 (t2 − t1 ) / 2 + D1 + d ) ( D1 − d ) 2

Using D1 = 1.15t1 + 1.5d results in

0.577π E2 d 0.577π E2 d
k2 = =
ln
( 0.575 ( t 2 − t1 ) + 1.15 t1 + 1.5d − d ) (1.15t1 + 1.5d + d )
ln
( 0.575 ( t 2 t1 ) + 0.5d ) (1.15t1 + 2.5d )
+
(0.575 (t2 − t1 ) + 1.15t1 + 1.5d + d ) (1.15t1 + 1.5d − d ) (0.575 (t2 + t1 ) + 2.5d ) (1.15t1 + 0.5d )

The lower frustra (frustra 3) is handled in a similar manner,

so that

0.577π Ed
k3 =
ln
(1.15t + D − d )( D + d ) t2
2
1
( t1 + t2 )
(1.15t + D + d )( D − d ) 2

where D is the diameter at the bottom of the frustra

( D = d w = 1.5d ), d is the bolt diameter and t is the thickness
o
60
of the bottom frustra ( t = (t1 + t2 ) / 2 ). Using these diameters
and thickness, and noting that the elastic modulus for frustra
1 is E = E2 , the stiffness of the bottom frustra is d w = 1.5d
 0.577π E2 d 0.577π E2 d
k3 = =
ln
(1.15 (t1 + t2 ) / 2 + 1.5d − d ) (1.5d + d ) ln
(0.575 (t1 + t2 ) + 0.5d ) (2.5d )
(1.15 (t1 + t2 ) / 2 + 1.5d + d ) (1.5d − d ) (0.575 (t1 + t2 ) + 2.5d ) (0.5d )

The bolt stiffness must also be considered. Assuming a through bolt, Table 8-7 gives

Ad At E
kb =
Ad lt + At ld

The stiffness of the member (joint) is therefore

1 1 1 1 1
= + + +
km k1 k2 k3 kb

Assuming a 12-mm (0.012 m) diameter ISO class 8.8 bolt, with t1 = 20 mm ( =0.02 m ) ,
t2 = 25 mm ( =0.025 m ) , E1 = 207 GPa , E2 = 113 GPa , and Eb = 207 GPa we get the following.

Frustra 1:

k1 =
0.577π E1d
=
(
0.577π 207 × 109 ( 0.012 ) )
ln
(1.15t1 + 0.5d )( 2.5d )
ln
(1.15 (0.020 ) + 0.5 (0.012 )) (2.5 (0.012 ))
(1.15t1 + 2.5d )(0.5d ) (1.15 (0.020 ) + 2.5 (0.012 )) (0.5 (0.012 ))
4502.7 × 106
k1 = = 4473.9 × 106 N/m
0.000870
ln
0.000318

Sililarly

0.577π E2 d
k2 =
ln
( 0.575 ( 2 t1 ) + 0.5d ) (1.15t1 + 2.5d )
t +
(0.575 (t2 + t1 ) + 2.5d ) (1.15t1 + 0.5d )
=
(
0.577π 113 × 109 )(0.012) =
2458 × 106
= 58770 × 106 N/m
ln
(0.575 (0.025 + 0.020 ) + 0.5 (0.012 )) (1.15 (0.020) + 2.5 (0.012 )) ln
0.0016894
(0.575 (0.025 + 0.020 ) + 2.5 (0.012 )) (1.15 (0.020 ) + 0.5 (0.012 )) 0.0016202

For the third frustra we have

k3 =
0.577π E2 d
=
(
0.577π 113 × 109 )(0.012)
ln
(0.575 (t1 + t2 ) + 0.5d ) (2.5d ) ln
(0.575 (0.025 + 0.020 ) + 0.5 (0.012 )) (2.5 (0.012 ))
(0.575 (t1 + t2 ) + 2.5d ) (0.5d ) (0.575 (0.025 + 0.020 ) + 2.5 (0.012 )) (0.5 (0.012 ))
2458 × 106
= = 2344.8 × 106 N/m
0.0009563
ln
0.0003353
If we now consider the bolt, we note that for the specified type of bolt used, Table E-3 gives

H = 10.8 mm, LG = 45 mm, and H + LG = 55.8 mm

From Table E-17; L = 60 mm, ld = 60 − 30 = 30 mm, lt = 45 − 30 = 15 mm, Ad = π (122 / 4 ) = 113 mm 2

From Table 8-1; At = 84.3 mm 2 , and using EQ (8-17) with Ad = 113 × 10 −6 m 2 and At = 84.3 × 10 −6 m 2

113 × 10 −6 (84.3 × 10−6 )(207 × 109 )

kb = = 466.8 × 106 N/m
113 × 10 −6 (0.015) + 84.3 × 10 −6 (0.030)

For the entire joint we therefore have

1 1 1 1 1 1 1 1 1
= + + + = + + +
km k1 k2 k3 kb 4473.9 × 106 58770 × 106 2344.8 × 106 466.8 × 106

km = 358 × 106 N/m