Scribd
Upload
57 views

Closed-Form Fatigue Equation: N N N N N N N N N N D

1) The document introduces a closed-form fatigue equation that predicts fatigue damage on steel structures based on the Palmgren-Miner cumulative damage hypothesis and stress distributions. 2) It derives an equation for expected fatigue damage over time by integrating contributions from all stress cycle components, using a Weibull probability density function to approximate the long-term wave distribution. 3) The equation contains parameters like scale and form that are functions of extreme stress ranges and structural configurations, and can be estimated based on stress data and the number of load cycles.

Uploaded by

Syaifur Rohman
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
57 views

Closed-Form Fatigue Equation: N N N N N N N N N N D

1) The document introduces a closed-form fatigue equation that predicts fatigue damage on steel structures based on the Palmgren-Miner cumulative damage hypothesis and stress distributions. 2) It derives an equation for expected fatigue damage over time by integrating contributions from all stress cycle components, using a Weibull probability density function to approximate the long-term wave distribution. 3) The equation contains parameters like scale and form that are functions of extreme stress ranges and structural configurations, and can be estimated based on stress data and the number of load cycles.

Uploaded by

Syaifur Rohman
Copyright
© © All Rights Reserved
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 5

5.

CLOSED-FORM FATIGUE EQUATION

Method of predicting the fatigue damage on steel structure has been


introduced by Palmgren-Miner through the accumulative damage
hypothesis:
m
ni n1 n2 n3 nm
D ......... (5.1)
i 1 N i N1 N 2 N 3 Nm
If p(S) is the stress pdf which can be defined in such a way hence p(Si)dS is
equivalent to the number of oscillation of stress component with the peak
value lies within an interval dS and with the mean value of Si. Further by
taking f and T as the mean frequency which vary randomly and the overall
operational time, respectively, hence the increase of the damage due to Si
which will take place during an interval T is:

T f p(S i )dS
dD (5.2)
N S i
5-1
N(Si) is number of cycle which would bring about damage at stress level Si.
From eq. (5.2) the expected damage that would take place in a certain
period T, could then be obtained by integrating contributions of all cycles of
the stress components, that is:


p L (S )
E ( D) T f ds (5.3)
0
N (S )

If T is the operational life of the structure as initially designed (ie. TL),


hence TL= nL/f. So eq. (5.3) could be rewritten by substituting N=A/Sm , as
follows:


n
E ( D) L S m p ( S )ds (5.4)
A 0

5-2
It is important to notice that p(S) is the continuous form (theoretical) of
pL(S) in eq. (4.3), which represents a discrete distribution. From various
investigation on the long-term wave distribution it has been concluded that
PL(S) could be closely approximated by Weibull pdf, namely:

x 1
x S S x
p L (S ) exp (5.5)
l l l

where:
l = scale parameter
x = form parameter
The value of l is a function of the extreme stress range.
The value of x is a function of the structural configuration and operational
sea site; for the general equation x may range between 0.75 up to 2.0; for
ocean structure x may range between 0.9 (mostly for large structures) up
to 1.1. (mostly for small structures).

5-3
If parameter Se is defined as the extreme stress occurs once during overall
cycles nL , thence l could be calculated as:

x x
S S
ln nL or (5.6)
l Se
l Se( ln nL ) 1/ x

Parameter x will be obtained from iteration on the basis of results from


pL(S) and nL .

By substituting eq. (5.5) into eq. (5.4) we can find:

x 1
n

x S S x
D L Sm exp ds (5.7)
A 0 l l l

5-4
Integral in eq. (5.7) could be simplified by implementing the gamma
function, G(n) and substituting into eq. (5.6) as follows:

x
S
Taking x will give
l

exp x dx
nL m (1 m / x ) 1
D l x (5.8)
A 0

If the gamma funtion is defined as :


G(n) e t t n1dt (5.9)
0

Approximation of gamma
function:
G( x) 0.0076 exp (1.6 x) 1.26

2x x / e
Stirling Formula :
G( x 1)
x

5-5

You might also like