See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/320837837

Modified Rainflow Counting Algorithm for Fatigue Life Calculation

Chapter · January 2018

DOI: 10.1007/978-981-10-6002-1_30

CITATIONS READS

5 3,529

4 authors, including:

Jyoti Jha
Indian Institute of Technology Bombay
18 PUBLICATIONS   28 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Friction characterization in hot working of Titanium alloy View project

We are working on Titanium alloy, magnesium alloy, Inconel View project

All content following this page was uploaded by Jyoti Jha on 29 April 2018.

The user has requested enhancement of the downloaded file.

 Modified Rainflow Counting Algorithm for Fatigue Life Calculation
Vaibhav Shinde*, Jyoti Jha, Asim Tewari, Sushil Mishra
National Centre For Aerospace Innovation and Research, Indian Institute of Technology Bombay Powai,Mumbai, 400076, India

Abstract: Fatigue life calculation of a varying load spectrum is a complex task. There is various approach suggested to
simplify the load spectrum into simple major and minor cycle. One such approach is Rain flow counting methodology.
In this paper, a new Microsoft Excel based algorithm for the Modified Graphical Rainflow counting method is
presented. Validity and Accuracy of algorithm has been demonstrated by various simple as well as complex examples.
This method allows the application of Miner's rule in order to assess the fatigue life of a structure subject to complex
loading.
Keywords: Cycle Counting, Rainflow counting Algorithm, Fatigue analysis, Hysteresis Loop

1. Introduction

It has been estimated that 90% of the structure failure is due to the Fatigue. Fatigue failure occurs due
to the varying load with respect to time. The fatigue life of the component which is subjected to constant
varying load can easily be calculated using S-N curve; however, when the varying load is random it is
difficult to estimate the fatigue life. There are several methodology has been suggested to convert the
random load spectrum to simpler cycles and one such approach is cycle counting. In the cycles counting
methodology, same cycles grouped together and miner's rule used to calculate the damage and fatigue life.
There are number of methods for the cycle counting, i.e. 1) Level crossing cycle counting, 2) Peak counting
method, 3) Simple-range counting method, 4) Rainflow counting method, 5) Range-pair counting method,
6) Two-parametric fatigue characteristics [6,9].
Rainflow counting method is widely practiced across the industry to convert the fatigue load spectrum
to simpler cycle. This method is the most accurate and also it gives the average value [12].

2. Concept of Rainflow Counting Method

There are numerous approaches for the load-cycle counting but Rainflow Counting methodology(RFC)
is based on the Stress-Strain response which is a memory characteristic of material for the various
amplitude loading. The path of stress-strain curve is influenced by the residual state of stress due to prior
strain history under variable cycle loading conditions. A stress -strain curves form a closed loop called
hysteresis loop, which is a pair of reversals. In RFC, one cycle is counted for one closed loop for a
corresponding constant amplitude [5, 7].
There are few approaches in Rainflow counting method to convert the complex stress-cycle to simple
major and minor cycle. In this paper, Modified Graphical Rainflow Counting Method (MGRM) has been
explained. This method is simple to understand but it is more complex to code into the computer program
[13]. Modified Graphical Rainflow Counting Method (MGRM) can be explained very well with the
following example shown in figure 1.

Figure 1 : Variable amplitude loading

To get a Cycle count from the variable amplitude loading we have to find out the sequences of the peak
and valleys. To get a peak and valleys, turn the sheet clockwise 90 0 as shown in figure. Drops of water
released from each maximum and minimum value of the loading. The drop will stop when the peak (or
valley) value is greater than or equal (or smaller than or equal) to the value of the peak (or valley) from
where the water drop starts. The water drop also stop when it merges with the path of the previous drops
path (point 2' as shown in figure).

a)

b)

Figure 2 : Rainflow counting Method a) Load Sequence and b) its Stress-Strain History

The Whole concept of the MGRM has been clearly illustrated in figure. In this figure one can see the
cycles count for the closed hysteresis loop. Whole cycles and Half cycles are shown in figure 2(b).

3. Computer programming of MGRM

Before giving data to program it is necessary to filter it or one can say to arrange the data in such a way
that there should be a successive extreme values. This can be done by adjusting equal value points with the
only one point of the corresponding value and polygonal line processing [1]. By using these techniques we
can obtain a data in such way that each point in the data is either local maximum or minimum. We can also
use a filter to eliminate the cycles which are less than the threshold value [8, page No. 10].
The algorithm developed for the modified Rainflow counting has been explained in the following
flowchart. Load the data and the time sequence in the form of array S[i] and T[i] respectively. At the
starting of the data if S[1] > S[2], then S[1] will be the peak or else it would be the valley. For the other
peaks and valleys go to step 1.

Figure 3: Flowchart for obtaining peak and valley

1) If S[i] < S[i+1] then go to the Valley or else go to the peak (i.e. step 2)
2) If S[i] = Peak (i.e. values of the peaks stores for the particular i in the step 5) then i+=1 and go back for
the next loop, or else go to the step 3.
3) Let j=i+1, and check for the condition S[j]<= S[i-1] with the j+=2, if condition satisfied at the j = b
then go to step 4 or else go to step 6.
4) Peak = Maximum value of the range (S[i] to S[b]) and store this peak for the particular corresponding i
value. In this, if (b-i) = 1 then S[i] itself a peak value, store this value and go to step 1 or else go to step
5.
5) In this step, effect of the step 4 will be implemented on the peaks which presents between the S[i] and
S[b]. Thus one can get the peaks for the values between S[i] to S[b] (which used to check in step 2).
For this let x = b with the x-=2, then search for the value S[x] = Peak and when this condition satisfies
then Peak at i = b is equal to the maximum value of range (x-2 to i). that has to be store so that when in
step 2, i comes equal to b it goes to next loop. Thus loop of x is repeated as shown in flowchart. After
reaching x at i+2 go to the step 1.
6) This is same as step 5 but instead of going up to b we have to go up to the last of the data i.e. S[n]. All
other procedure and conditions are same as step 5.

Flowchart of the Valley is also same like peak but the conditions are the reverse of the peak. After
getting peaks and valleys, the data have been processed and arranged in a Microsoft excel macro to get the
closed cycles.
 4. Illustrative Example

Whole algorithm can be explained well with the following simple example (Based on figure 1). Loads
Sequence : 0, 10, 5, 15, -4, 3, -9, 6, -1, 20. Peaks and valleys for the give load sequence can be obtained
with the help of computer program (Fig. 3) are as shown in table 1.

Table 1: Peaks and Valleys

Valley Peaks
0 15
5 10
-9 3
-4 20
-1 6

After getting a peaks and valleys we have to arrange the data as shown in Table 2. Thus half reversal
obtained as shown in table 2.
Table 2: Half Reversals

Stress from
Peak/Valley No. of
sequence (X),
(Y), MPa Cycles
MPa
0 15 0.5
10 5 0.5
5 10 0.5
15 -9 0.5
-4 3 0.5
3 -4 0.5
-9 20 0.5
6 -1 0.5
-1 6 0.5
20 --- ----

Table 3 : Number of complete cycles

From Sigma Sigma Sigma No. of

To (MPa) Range Sigma Alt
(MPa) Max Min Mean Cycles
10 5 10 5 5 2.5 7.5 1

-4 3 3 -4 7 3.5 -0.5 1

6 -1 6 -1 7 3.5 2.5 1
 In table 2, we have to write the peak and valleys alternatively in the column Y and stress sequence in the
column X with help of simple computer code. when first value (starting value i.e. 0 in our example) in
stress sequence is valley then in the Y column we have to start with the Peak and vice versa. when Xi = Yk
and at the corresponding Xk = Yi then closed cycle will form. After getting full cycles, cycles the same
Maximum and minimum values of closed cycles bunched together. Output from above table by using
simple code will be as shown in table 3.

5. Conclusion

Modified Graphical method using Microsoft excel macro has been explained in this paper. Since the
entire data can be processed in the excel sheet, it can easily used for fatigue life calculation.

REFERENCES

1. Qing Deng and Hongjie Yuan, “The Algorithm for Graphic Method of Rain-flow Counting in
Programming”, School of reliability and System Engineering, Beihang University, Beijing, China.
(2011)
2. Peter J. Massarelli, Thomas T. Baber, “Fatigue reliability of steel highway bridge details” ,
Virginia Transportation Research Council, Charlottesville, Virginia (2001)
3. Lakshmi GopiReddy, Leon M. Tolbert, Burak Ozpineci, "Lifetime Prediction of IGBT in a
STATCOM Using Modified-Graphical Rainflow Counting Algorithm",The University of
Tennessee, Oak Ridge National Laboratory, USA (2012)
4. C.H. McInnes , P.A. Meehan., “Equivalence of four-point and three-point rainflow cycle counting
algorithms”, CRC Mining, School of Engineering, International Journal of Fatigue 30 (2008) 547–
559.
5. K. L. singh and V. R. Ranganath, "Gyele counting using rainflow algorithm for fatigue analysis",
NAL, Bangalore, India.
6. Bogdan Ligaj, "An analysis of the influence of cycle counting methods on fatigue life calculations
of steel", SCIENTIFIC PROBLEMS OF MACHINES OPERATION AND MAINTENANCE , 4
(168) 2011, page No. 25-43
7. N. Hong, "A Modified Rainflow Counting Method," international Journal of Fatigue, Vol. 13,
Page No. 465-469 (1991)
8. L. L. Schluter, H. J. Sutherland, "User’s Guide for LIFE2’s Rainflow Counting Algorithm", Sandia
National Laboratories, New Mexico and Livermore, California (1991)
9. American Society for Testing and Materials (ASTM),Standard Practices for Cycle Counting in
Fatigue Analysis, ASTM E1049-85((Reapproved 2005), ASTM International, West
Conshohocken, Pennsylvania.(2005)
10. Fengjie Yin, Mark Cerkovnik, "Assessment of Fatigue Damage from Variable Amplitude Loads in
Risers", University of Windsor, Windsor, ON, Canada ()
11. C. Amzallag, J. P. Gerey, J. L. Robert and J. Bahuaud, "Standardization of the Rainflow counting
method for fatigue analysis", France (1993)
12. Krzysztof Mainka, Markus Thoben, Oliver Schilling, "Counting Methods for Lifetime Calculation
of Power Modules", Infineon Technologies AG, Warstein, Germany (2011).

View publication stats