ENDURANCE LIMIT—

APPROXIMATE
ESTIMATION
• A number of tests are required to prepare one S–N curve and
each test takes considerable time.
• It is, therefore, not possible to get the experimental data of
each and every material.
• When the laboratory data regarding the endurance limit of the
materials is not available, the following procedure should be
adopted:
• The endurance limit of a component is different from the
endurance limit of a rotating beam specimen due to a number of
factors.
• The difference arises due to the fact that there are standard
specifications and working conditions for the rotating beam
specimen, while the actual components have different
specifications and work under different conditions.
• Different modifying factors are used in practice to account for this
difference. These factors are, sometimes, called derating factors.
• The purpose of derating factors is to ‘derate’ or reduce the
endurance limit of a rotating beam specimen to suit the actual
component.
• Only four factors that normally require attention are discussed.
• The relationship between (Se) and (S’e) is as follows:
(a) Surface finish Factor
• The surface of the rotating beam specimen is polished to mirror finish.
The final polishing is carried out in the axial direction to smooth out
any circumferential scratches.
• This makes the specimen almost free from surface scratches and
imperfections. It is impractical to provide such an expensive surface
finish for the actual component.
• The actual component may not even require such a surface finish.
When the surface finish is poor, there are scratches and geometric
irregularities on the surface.
• These surface scratches serve as stress raisers and result in stress
concentration. The endurance limit is reduced due to introduction of
stress concentration at these scratches.
• It should be noted that ultimate tensile strength is also a parameter
affecting the surface finish factor.
• High strength materials are more sensitive to stress concentration
introduced by surface irregularities.
• Therefore, as the ultimate tensile strength increases, the surface finish
factor decreases.

Ka can be estimated from

the plot.
• Shigley and Mischke have suggested an exponential equation for the
surface finish factor.
• This equation is based on experimental data points obtained by Noll
and Lipson. This equation is in the following form:

• Only for steels

(b) Size Factor:
The rotating beam specimen is small with 7.5 mm diameter. The larger
the machine part, the greater the probability that a flaw exists
somewhere in the component.
• The chances of fatigue failure originating at any one of these flaws are
more. The endurance limit, therefore, reduces with increasing the size
of the component.
• The size factor Kb takes into account the reduction in endurance limit
due to increase in the size of the component.
• For bending and torsion, the values of size factor (Kb) are given in
Table.
(c) Reliability Factor
• The laboratory values of endurance limit are usually mean values.
There is considerable dispersion of the data when a number of tests are
conducted even using the same material and same conditions.
• The standard deviation of endurance limit tests is 8% of the mean
value. The reliability factor Kc depends upon the reliability that is used
in the design of the component.
• The greater the likelihood that a part will survive, the more is the
reliability and lower is the reliability factor. The reliability factor is
one for 50% reliability.
• This means that 50% of the components will survive in the given set
of conditions.
• To ensure that more than 50% of the parts will survive, the stress
amplitude on the component should be lower than the tabulated value
of the endurance limit.
• The reliability factor is used to achieve this reduction. The reliability
factors based on a standard deviation of 8% are given in Table.
(d) Modifying Factor to Account for Stress Concentration
• The endurance limit is reduced due to stress concentration.
The stress concentration factor used for cyclic loading is less
than the theoretical stress concentration factor due to the
notch sensitivity of the material.
• To apply the effect of stress concentration, the designer can
either reduce the endurance limit by (Kd) or increase the
stress amplitude by (Kf).
• The modifying factor Kd to account for the effect of stress
concentration is defined as
q, Kf and Kt relationship

• q and kt can be obtained from graphs.

• The above mentioned four factors are used to find out the endurance
limit of the actual component.
• The endurance limit (Sse) of a component subjected to fluctuating
torsional shear stresses is obtained from the endurance limit in
reversed bending (Se) using theories of failures.

• The endurance limit in axial loading is lower than the rotating beam
test.
• For axial loading (Se)a = 0.8 Se
 REVERSED STRESSES—DESIGN FOR
FINITE AND INFINITE LIFE
There are two types of problems in fatigue design—
(i) components subjected to completely reversed stresses, and (ii)
components subjected to fluctuating stresses.
The mean stress is zero in case of completely reversed stresses.
The stress distribution consists of tensile stresses for the first half cycle and
compressive stresses for the remaining half cycle and the stress cycle passes
through zero.
In case of fluctuating stresses, there is always a mean stress, and the stresses
can be purely tensile, purely compressive or mixed depending upon the
magnitude of the mean stress.
• Such problems are solved with the help of the modified Goodman
diagram, which will be discussed later.
• The design problems for completely reversed stresses are further
divided into two groups—
• (i) design for infinite life, and (ii) design for finite life.

• Case I: When the component is to be designed for infinite life, the

endurance limit becomes the criterion of failure. The amplitude stress
induced in such components should be lower than the endurance limit
in order to withstand the infinite number of cycles.
• Such components are designed with the help of the following
equations:
• Case II: When the component is to be designed for finite life, the S–N
curve is to be prepared as shown below.
Procedure to draw S-N Curve
• It consists of a straight line AB drawn from (0.9 Sut) at 103 cycles to
(Se) at 106cycles on a log-log paper. The design procedure for such
problems is as follows:
 Assignment 1- (Machine design-I)
Q1. A cylindrical pressure vessel with a 1.5 m inside diameter is subjected to internal steam pressure of 1.5
MPa. It is made from steel plate by triple-riveted double-strap longitudinal butt joint with equal straps. The
pitch of the rivets in the outer row is twice of the pitch of the rivets in the inner rows. The rivets are arranged in
a zigzag pattern. The efficiency of the riveted joint should be at least 80%. The permissible stresses for the plate
and rivets in tension, shear and compression are 80, 60 and 120 N/mm 2 respectively. Assume that the rivet in
double shear is 1.875 times stronger than in single shear. Design the longitudinal joint.

Q2. A cantilever beam made of cold drawn steel 20C8 (Sut = 540 N/mm 2) is subjected to a completely reversed
load of 1000 N as shown in Fig. below. The notch sensitivity factor q at the fillet can be taken as 0.85 and the
expected reliability is 90%. Determine the diameter d of the beam for a life of 10000 cycles.
CUMULATIVE DAMAGE IN FATIGUE
• In certain applications, the mechanical component is subjected to
different stress levels for different parts of the work cycle. The life of
such a component is determined by Miner’s equation.
• Suppose that a component is subjected to completely reversed stresses
( ) for (n ) cycles, ( 2) for (n2) cycles, and so on. Let N be the number
1 1 1

of stress cycles before fatigue failure, if only the alternating stress ( )

1

is acting. One stress cycle will consume (1/N ) of the fatigue life and
1

since there are n such cycles at this stress level, the proportionate
1

damage of fatigue life will be [(1/N )n ] or (n /N ). Similarly, the

1 l l l

proportionate damage at stress level ( 2) will be (n /N ). Adding these

2 2

quantities, we get
SODERBERG AND GOODMAN LINES
• When a component is subjected to fluctuating stresses as shown in Fig.
there is mean stress (m) as well as stress amplitude ( a). It has been
observed that the mean stress component has an effect on fatigue failure
when it is present in combination with an alternating component.
• The fatigue diagram for this general case is shown in Fig. In this
diagram, the mean stress is plotted on the abscissa. The stress amplitude
is plotted on the ordinate. The magnitudes of (m) and ( a) depend upon
the magnitudes of maximum and minimum force acting on the
component.
• When stress amplitude (a) is zero, the load is purely static and the
criterion of failure is Sut or Syt. These limits are plotted on the abscissa.
When the mean stress (m) is zero, the stress is completely reversing and
the criterion of failure is the endurance limit Se that is plotted on the
ordinate.
• When the component is subjected to both components of stress, viz.,
(m) and (a), the actual failure occurs at different scattered points
shown in the figure.
• There exists a border, which divides safe region from unsafe region
for various combinations of (m) and (a). Different criterions are
proposed to construct the borderline dividing safe zone and failure
zone. They include Gerber line, Soderberg line and Goodman line.
(i) The Goodman line is safe from design considerations because it is
completely inside the failure points of test data.
(ii) The equation of a straight line is simple compared with the equation
of a parabolic curve.
(iii) It is not necessary to construct a scale diagram and a rough sketch is
enough to construct fatigue diagram.
 MODIFIED GOODMAN DIAGRAMS
• The region OABC is called
modified Goodman diagram.
All the points inside the modified
Goodman diagram should cause
Neither fatigue failure nor yielding.
• If the mean component of stress (m) is very large and the alternating
component (a) very small, their combination will define a point in the
region BCF that would be safely within the Goodman line but would
yield on the first cycle. This will result in failure, irrespective of safety
in fatigue failure.
• While solving a problem, a line OE with a slope of tan is constructed
in such a way that,
• The modified Goodman diagram for fluctuating torsional shear
stresses is shown in Fig. below. In this diagram, the torsional mean
stress is plotted on the abscissa while the torsional stress amplitude on
the ordinate. The torsional yield strength Ssy is plotted on the abscissa
and the yield line is constructed, which is inclined at 45° to the
abscissa.
• A line is drawn through Sse on the ordinate and parallel to the abscissa.
The point of intersection of this line and the yield line is B. The area
OABC represents the region of safety in this case.
• A fatigue failure is indicated if
Problem 1: