CHAPTER 8

DESIGN OF SCREWS, FASTENERS AND CONNECTIONS

Thread Standards and Definitions

All threads have a standard form of profile or geometry and made according to the right hand
rule unless otherwise noted. The thread angle is 60o and the crest of the threads may be either
flat or rounded. Two widely used standards are the American National (Unified) thread and the
metric threads. Fig.8.2 shows the thread geometry of the metric M and MJ profiles. The M
profile replaces the inch class and is the basic ISO 68 profile with 60 o symmetric threads. The
MJ profile has rounded fillet at the root of the external thread and a large minor diameter of
both internal and external threads. This profile is especially used where a high fatigue strength
is required.

Terminology
The terminology of screw threads, illustrated in Fig.8-1, is explained as follows:
The major diameter d is the largest diameter of a screw thread.
The minor diameter dr or d1 is the smallest diameter of a screw thread.
The pitch is the distance between adjacent thread forms measured parallel to the thread axis.
In U.S units, number of thread forms per inch N is used instead of the pitch and pitch is the
reciprocal of N.
A multiple-threaded product is one having two or more threads cut beside each other (with
suitable phase difference).
The lead l is the distance the screw or nut moves parallel to the screw axis for one turn of it.
For a single thread, the lead is the same as the pitch; a double threaded screw has a lead equal
to twice the pitch, lead is three times the pitch for triple threaded and so on.

Thread specification
Metric threads are specified by writing the diameter and pitch in millimetres, in that order as
M12X1.75.Unified threads are specified by stating the nominal major diameter, the number of
threads per inch and the thread series for example 0.625”-18 UNRF.

Tensile-Stress area
Great many tensile tests of threaded rods have shown that an unthreaded rod having a diameter
equal to the mean of the pitch diameter and minor diameter will have the same tensile strength
as the threaded rod. The area of this rod is called the tensile-stress area A t of the threaded
rod.

Other forms
Square and Acme threads, shown in Fig. 8-3a and 3b, respectively, are used on screws when
power is to be transmitted. Modifications are frequently made to both square and Acme threads.
For instance an included thread angle of 10 to 15 o; the modifications retains most of the high
efficiency inherent in square threads and make the cutting simpler. Acme threads are some
times modified to a stub form by making the teeth shorter. This results in larger minor diameter
and a somewhat stronger screw.
THE MECHANICS OF POWER SCREW

A power screw is a device used in machinery to change angular motion into linear motion, and
usually, to transmit power. Familiar applications include the lead screws of lathes, and the screws
for vices, presses, and jacks.

Equations for Load and Torque.

A square-threaded power screw with single (start) thread having a mean diameter d m, a pitch p, a
lead angle , and a helix angle  is loaded by the axial compressive force F. We wish to find
expressions for the torque required to raise, or to lower this load. By representing the
summation of all the unit axial forces acting upon the normal thread area by F and equilibrium of
the system under the action of the various forces as shown in Fig 8-6 the final form of the
equations are:

F [l / d m   ]
P= to raise the load
1  ( l / d m )

F [   (l / d m )]
P= to lower the load
1  ( l / d m )

Using the relation tan  = l/dm and representing co-efficient of friction  by the friction angle 
(i.e.  = tan ) the equation can be simplified as

P = F. tan ( +) for raising the load and

P = F. tan (-) for lowering the load.

Noting that the torque is the product of the force P and the mean radius d m/2, we can write;
F .d m  l  d m  Fd m
T=   or tan ( +) for raising the load and
2  d m  l  2

F .d m  d m  l  Fd m
T=   or tan ( -) for lowering the load.
2  d m  l  2

Self locking.
It may turn out, in specific instances where the lead is larger or the friction is low, the torque T
will be negative or zero, this will result that the load will lower itself by causing the screw to
spin without any external effect. When positive torque is obtained from this equation, the screw
is said to be self-locking. Thus the condition for self-locking is
dm > l or  > tan .
This relation states that self-locking is obtained whenever the co-efficient of thread friction is
equal to or greater than the tangent of the thread lead angle.
Efficiency
An expression for efficiency is also useful in the evaluation of power screws and can be given by
the equation

To F .l
E = 
T 2 .T

The proceeding equations have been developed for square threads where the normal thread
loads are parallel to the thread axis of the screw. In case of Acme or other threads, the normal
load is inclined to the axis because of the thread angle 2 and the lead angle . Since the lead
angle is small, this inclination can be neglected and only the effect of the thread angle (Fig. 8-
7a) considered. The effect of the angle is to increase the frictional force by the wedging action
of the threads. This can be accounted by dividing the friction terms by cos . ( i.e. replacing 
by /cos  or  sec ).

Collar friction
Usually a third component of torque must be applied must be applied in power screw applications.
When the screw is loaded axially, a thrust or collar bearing must be employed between the
rotating and stationary members in order to carry the axial component. Fig 8-7b shows a typical
thrust collar in which the load is assumed to be acting at the mean collar diameter d c. If c is the
co-efficient of collar friction, the collar torque required is
F c d c
Tc = and the total torque is T = Tt +Tc.
2
THREADED FATENERS

Threaded fasteners are of basically three types. The bolt and the nut, the screw and the stud.
Drawing of a standard hexagon-head bolt is shown in Fig. 8-8. See Table A-26 for dimensions.
The ideal bolt length is one in which only one or two threads project from the nut after it is
tightened. The hexagon-head cap screw is similar to a bolt but the diameter of its head is much
smaller and slightly thinner than that of the bolt. Three common cap-screw head styles are
shown in Fig. 8-9 and its dimensions are listed in Table A-27. A variety of machine screw head
styles are shown in Fig. 8-10. A stud is a rod threaded on both ends. Several styles of hexagonal
nuts are illustrated in Fig. 8-11; their dimensions are given in Table A-28. The material of the nut
must be selected carefully to match that of the bolt.

The purpose of a fastener is to clamp two or more parts together. The clamping load stretches
or elongates the bolt (screw); the load is obtained by twisting the nut, until the bolt has
elongated almost to the elastic limit. If the nut does not loosen, this bolt tension remains as the
pre load or clamping force. This clamping force is called the pre-tension or bolt preloads. Of
course, since the members are being clamped together, the clamping force, which produces
tension in the bolt, induces compression in the members.

TENSION CONNECTIONS

When a connection is desired which can be disassembled without the destructive methods and
which is strong enough to resist external tensile loads, moment loads, and shear loads, or a
combination of these, then the simple bolted joint using hardened-steel washer is good solution.
Such a joint can also be dangerous unless it is properly designed and assembled by a trained
mechanic.

External load and the Joint design

Let us now consider what happens when an external tensile load P, as in Fig. 8-12 is applied to
bolted connection to which the preload Fi, has been correctly applied
Let Pb the portion of P taken by bolt and Pm the portion of P taken by the member
Fb the resultant bolt load = P b + Fi.
Fm the resultant load on the members = Pm - Fi.
P
Since P = Pb + Pm and b=m= and =
k
kb
Fb = Pb + Fi = P + Fi = C.P+ Fi. …………..(1)
kb  k m
km
Fm = Pm - Fi = P - Fi = (1- C).P+ Fi …………..(2)
kb  k m
Where C and (1-C) are co-efficients of P in the above equations, respectively, termed as spring
constant, stiffness constant or stiffness co-efficient.

Safe or Leak Proof Joint -Bolt Pre load-Static loading

One way of ensuring the joint safety is to limit the load on the bolt such that the stress
induced on it is less than the proof strength of the bolt material
 Fb
The tensile stress in the bolt =
At
CP Fi
i.e. b = 
At At
To ensure safety introducing a load factor n we can write
CnP Fi
 = SP
At At
Hence Fi = SP.At - CnP
Any value of n > 1 ensures that the bolt stress is less than the proof strength

Another means of ensuring a safe or leak proof joint is to require that the external load be
smaller than that needed to cause the joint to separate. If separation does occur, then the
entire external load will be imposed on the bolt, which is undesirable again.

Let Po be the value of the external load that would cause joint separation.
At separation Fm = 0, hence

(1-C) Po – Fi = 0
Assuming a factor of safety or overload factor n guarding against joint separation
Po
n = and by substituting Po = n.P
P
Fi
Fi = n.P (1-C) or n =
P (1  C )

The stress-strain diagram of a good quality bolt material is shown in Fig. 8-15. Notice that
there is no clearly defined yield point and that the stress progresses smoothly upto fracture,
which correspond to the tensile strength. This means that no matter how much pre-load is given
to the bolt, it will retain its load –carrying capacity. This what keeps the bolt tight and
determine the joint strength. The pre-tension is the muscle of the joint and its magnitude is
determined by the bolt strength. If the full bolt strength is not used in developing the tension,
then joint is weaker and money is wasted.

Good quality bolts can be preloaded to the plastic range to develop more strength. Some of the
bolt torque used in tightening produces torsion, which increases the principal tensile stress (on
the bolt). Thus as a rule, bolt will either fracture during tightening, or not at all.
Bowman recommends a preload of 75 percent of proof load, which is about the same as RB&W
recommendations for reused bolts. In view of these guidelines, it is recommended for both
static and fatigue loading the following be used for preload:
O.75 Fp for reused joints
Fi =
0.90 Fp for permanent connections.
Where Fp is the proof load, obtained from the equation F p = At. Sp
The proof strength Sp can be obtained from Tables 8-4 to 8-6. For other materials, an
approximate value is Sp = 0.85 Sy.
BOLT STRENGTH

The bolt strength is the key factor in the design and analysis of bolted connections as noted
above. In the specification standard for bolts, stating the minimum proof strength, or minimum
proof load, and the minimum tensile strength specifies the strength. The proof load is the
maximum load (force) that a bolt can withstand without acquiring a permanent set. The proof
strength is the quotient of the proof load and tensile-stress area. The proof strength thus
corresponds roughly to the yield-point strength and is about 90 percent of the 0.2 percent
offset yield strength.

The SAE and ASTM specifications for Steel bolts are given in Table 8-4 and 8-5. Specification
in terms of Mechanical-Property Classes, of metric fasteners is given in Table 8-6.

Spring constant or Stiffness constants

Fastener:
The stiffness of the portion of a bolt or screw within the clamped zone will generally consist of
two parts that of the unthreaded shank portion and that of the threaded portion. Thus the
stiffness constant of the bolt is equivalent stiffness of two springs in series and hence can be
given as
Ad AT E
kb =
Ad lT  AT l d
The members
There may be more than two members included in the grip of the fastener. All together these
act like compressive springs in series, and hence the total spring rate of the members is
1 1 1 1
  
k m k1 k 2 ki
If one of the member is a soft gasket, its stiffness relative to the other members is usually so
small that for all practical purposes the others can be neglected and only the gasket stiffness is
used. On the other hand if no gasket is used , the stiffness of the members is rather difficult
to obtain, because the compression spreads out between the bolt head and the nut and hence
area is not uniform. Assuming a general con geometry (Fig.8-14)having half-apex angle , the
spring rate or stiffness of the frustum is given by

P Ed tan 
k  
 ( 2t tan   D  d )( D  d )
ln
( 2t tan   D  d )( D  d )
The equation is to be solved separately for each frustum in the joint and assembled to obtain k m

Using the grip as l = 2t and d w as the diameter of the washer face, which is about 50% greater
than the fastener diameter for standard hexagonal head bolts and cap screws; If we use  = 30o
( Osgood reports a range of 25 o    33o)final equation for connected parts of the same
material(same E value) is given as
0.577Ed tan 
  0.577l  0.5d  
km = 2 ln 
 5 0.577l  2.5d  

  
Torque requirements

Having learnt that a high preload is very desirable in important bolted connections, we must next
consider means of ensuring that the preload is actually developed when the part are assembled.
For this the elongation of the screw, the torque wrenching or the turn of the nut method may
be used. However torque wrenching is more relied on and has a built-in dial which indicates the
proper torque.
A good estimate of the torque required to produce a given preload can be obtained by combining
the basic equations thread resistance torque and collar torque(developed earlier) and with the
mean collar diameter of washer face as 1.25 d, the final equation is
d m tan    sec 
T = Fi.d [ ( )( )  0.625 c ]
2d 1   tan  sec 
Or T= K.Fi.d where K is the torque co-efficient given above

Gasketed joints

If a full gasket is present in the joint, the gasket pressure p is found by dividing the force in
the member by the gasket area per bolt and for joint to be effective certain minimum pressure
is to be maintained on the gasket and hence the desired pre-load to be given is
Fi Ag po. ( where po is the minimum gasket/seal pressure)
In full gasketed joints the uniformity of the pressure is important. To maintain uniformity, bolts
should not be spaced more than six bolt diameters apart. But to maintain wrench clearance,
bolts should be spaced at least three diameters apart.
FATIGUE LAODING

Most of the time bolted joint encounters fatigue loading and the type of fatigue loading is
generally one in which the externally applied load fluctuates between zero and some maximum
force. This would be the situation in a pressure cylinder. Because of the preloading the load on
the bolt will be Fmax =Fb and Fmin = Fi.
Therefore the alternating component of bolt stress is

Fb  Fi kb P CP
a = = 
2. At k b  k m 2. At 2. At
Then, since the mean stress is equal to the alternating component plus the minimum stress, we
have

Fi C.P Fi
m = a +  
At 2. At At

Now applying the Goodman’s criterion the joint can be designed using the final form of the
equation

 Fm Fa 
At = N  K f  where N is the factor of safety desired.
 S ut Se 
(Note : The mean and variable loads F m and Fa respectively are per bolt and S e is the fully
corrected endurance limit value)

BOLTED AND RIVETED JOINTS LOADED IN SHEAR

Riveted and bolted joints loaded in shear are treated alike in design and analysis. In Fig.8-21 is
shown a riveted connection loaded in shear and various modes of failure in them. In structural
design it is customary to select in advance the number of rivets and their diameter and spacing.
The strength is then determined for each method of failure. If the calculated strength is not
satisfactory, a change is made in the diameter, spacing, or number of rivets used, to bring the
strength in line with the expected loading conditions. It is not usual, in structural practice, to
consider the combined effects of the various failure modes.

Centroids of bolt groups

In certain applications the diameter (or even spacing) of the bolts(rivets) need not be the same
In order to determine the forces which act upon each bolt it is necessary to know the location
of the centriod of the bolt group. In many instances the centroidal distance can be located by
symmetry. Otherwise this can be done easily applying the law of statics, as illustrated in Fig.8-
22.
Shear of bolts and rivets due to eccentric loading

In this case the total load taken by each bolt will be calculated in three steps.
In the first step the shear load V is divided equally among the bolts so that each bolt takes a
load of F’= V/n where n refers to the number of bolts in the group and the force F ’ is called the
direct load, or primary shear. The moment load or secondary shear is additional load on each bolt
due to the Moment M. It is evident that
M = F”.rA + F”.rB + F”.rC
where F” is the moment load and r A , rB , rC etc. are the radial distances from the centroid to
the centre of each bolt. The force taken by each bolt depends upon its radial distance and
hence the bolt farthest from the centriod takes the greatest load, while the nearest one the
least. We can therefore write
" " "
FA FB F
  C

rA rB rC

M .rn
Hence F”n =
r  rB2  rC2  ....
2
A
where the subscript n refers to that particular bolt whose load is maximum or to be found

In the third and final step the direct and moment loads are added vectorially to obtain the
resultant load on each bolt. Since all the bolts or rivets are of the same size, only the bolt having
the maximum load need be considered and analysed.
Now
R S sy
R= (F '
) 2  (F "
)2  2.F ' .F "
.Cos and max = 
A N

SET SCREWS

Unlike bolt and cap screws, which depend upon tension to develop a clamping force, the set screw
depends upon compression to develop the clamping force. The resistance to axial motion or
rotary motion of the collar or hub relative to the shaft is called the holding power. This holding
power, which is really a force resistance, is due to the frictional resistance of contacting
portions of the collar and shaft as well as any slight penetration of the set screw into the shaft.

Fig. 8-26 shows the point styles available with socket set screws. These are also manufactured
with screwdriver slots and with square heads. Table 8-13 lists the values of seating torque and
corresponding holding power for inch-series set screws. The value listed applies to both axial
holding power, for resisting thrust, and the tangential holding power, for resisting torsion.
Typical factors of safety are 1.5 to 2.0 for static loads and 4-8 for various dynamic loads. Set
screws should have a length of about half of the diameter. Note that this practice also provides
a rough rule for the radial thickness of a collar or hub.
KEYS AND PINS

Keys and pins are used on shafts to secure rotating elements, such as gears, pulleys, or other
wheels. Keys are used to enable transmission or torque from the shaft to the shaft-supported
element. Keys are square or rectangular in cross section, standard size of these, together with
the range of applicable shaft diameter, are listed in Table 8-15. The length of the key is based
on the hub length and the torsional load to be transferred.
4.T
i.e. l > 2.T(Sy)/w.d (shear) or ( S y ) (bearing stress)
h.d
The Woodruff key shown in Fig 8-28 basis of general usefulness, especially when a wheel is to be
positioned against a shaft shoulder, since the key slot need not be machined into the shoulder-
stress concentration region. The use of the Woodruff key also yields better concentricity after
assembly of the wheel and the shaft. This is especially important at high speeds, as for example,
with a turbine wheel and shaft. Dimensions
Pins are useful when the principal loading is shear and when both torsion and thrust are present.
Tapered pins are sized according to the diameter at the large end. Some of the most useful
sizes of these are listed in Table 8-14
 ARP Bolt Analysis

12
Diameter bolt d mm
 Young's modulus bolt E 210
GPa
 Thickness clamped material lm 21
mm
 Young's modulus clamped material E 210
GPa
 
 Bolt stiffness kb=AE/lm 109 N/m
 Diameter d2=1.5d mm
 Diameter d3=d2+lm·tan(φ), φ=30º mm
 Stiffness clamped material km 109 N/m
 Joint stiffness factor 1)
Cm=kb/(kb+km)
1)
For simplicity, the clamped materials are frequently assumed to have a
stiffness of three times the bolt stiffness, which results in a joint stiffness
factor of Cm=1/4. With Cm=1/4, it follows that only one fourth of the applied
load P is taken by the bolt.

When performing a bolt analysis, it is important to take a very methodical approach. The

methodical approach can be reduced to three steps. The first step is to perform a stress

analysis on the bolt. The second step is to perform a material analysis on the bolt. The third step

is to do a fatigue analysis on the bolt.

In the stress analysis, the first step is to find the stiffness of the bolt and the member

(km and kb). The stiffness of the member is dependent on the stiffness of the members. The

frustum begins from the head of the bolt and extends along the bolt at thirty degrees from the
horizontal. It is at its widest at half the length of the member being clamped together. Its

beginning width is 1.5*d. A new frustum must be started whenever the material being clamped

changes. For more explanations on calculating frustums, refer to Figure 1.

The equation for calculating the stiffness of each frustum is:

k  (( pi * E * d * tan( alpha ))) / ln((1.15 * t  D  d ) * ( D  d )) /((1.15 * t  D  d ) * ( D  d ))

Once each frustum has been found the total stiffness of the member can be calculated

in the following manner: 1 / km  (1 / k1)  (1 / k 2) .

To calculate the stiffness of the bolt, the following equation may be used.

kb  ( Ad * At * E ) /( Ad * lt  At * ld ) . The joint constant (C) can then be calculated with the

following equation: C  kb /( km  kb) For further instruction on calculating the stiffness of the

bolt, refer the bolts poster and figure 1.

Figure 1
 At this point, the material analysis can be completed for the bolt. Luckily, the material

analysis of bolts is very simple. Material properties can be found by examining tables. You will

need to find the proof strength (Sp) and the tensile stress area (At). By multiplying these values

together you are able to find the proof load (Fp).

Using the proof load value that was just found, you are able to calculate the optimal

preload (Fi). This equation changes based on whether the bolts will be reused or are final

connections. For permanent connections the equation is Fi  .9 * Fp . For reusable connections

the equation is Fi  .75 * Fp .

The static factor of safety and the joint factor of safety can now be calculated by using

the initial preload values. The equation for the static factor of safety can be calculated using

the following equation: n  ( Sp * At  Fi ) / C * P , where P is the external load applied to the

bolt. The joint factor of safety can be calculated using the following equation:

n  ( Fi ) /( P * (1  C )) .

Then you may proceed to the fatigue analysis. The first step in this process is to

calculate the alternating stress and the mean stress. The alternating stress can be calculated

using the equation  _ a  (C * P) /( 2 * At ) and the mean stress can be calculated using this

equation  _ m  ((C * P ) /( 2 * At ))  ( Fi / At ) . The fatigue factor of safety can be found by

n  Sa /  _ a . The value Sa can be found with the following equation

Sa  ( Sut  ( Fi / At )) /(1  ( Sut / Se)) . Sut is the ultimate strength of the material and Se is the

endurance limit. Note that the endurance limit is dependant on the size factor, the load factor,

the surface factor, and the temperature factor.