Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Fatigue life calculations

I came across in the documentation that the life is calculated using the formula.

S_r = SRI1(N_f)^b

But, even after maintaining the co-efficients, TRTMENT, FINISH and [KF] as 1, the results are off (when compared with the manual calculation using the above mentioned formula) by a random factor (approximately 0.5), for each unique component I take for fatigue analysis (Refer image attached). Why is it so?

What is the calculation happening at the back end of the software?

<?xml version="1.0" encoding="UTF-8"?>

Why is the factor 0.5 coming into play?

Find more posts tagged with

English

OptiStruct

Accepted answers

All comments

Altair Forum User

Hi @SKGPrashanth

Is it possible to share the model file with us so that we can have a deeper look into the model?

Please share file using File Transfer link available in my signature.

Altair Forum User

Hi @SKGPrashanth

Is it possible to share the model file with us so that we can have a deeper look into the model?

Please share file using File Transfer link available in my signature.

Unable to find an attachment - read this blog

Altair Forum User

Hi Pranshanth,

Thank you, Can you let me know what is the maximum and minimum loads or Stress amplitude and mean stress?

Altair Forum User

The maximum and minimum loads are 2000 N and -2000 N respectively. (defined by sine curve)

Stress amplitude is 198 Mpa

Mean Stress is 0 MPa (Because of fully reversed loading)

Altair Forum User

Prashanth,

Thank you.

Since you mentioned as maximum stress, I wanted to confirm with you about the Sa and Sm.

I will check and update to you soon.

Altair Forum User

Prashanth,

Thank you.

Since you mentioned as maximum stress, I wanted to confirm with you about the Sa and Sm.

I will check and update to you soon.

In the meanwhile, I tried the exact same problem in Ansys Workbench; where I got exactly the same result as that from the formula.

What am I missing out in Radioss?

Altair Forum User

Hi Prashanth,

I am checking with experts. I will get back to you soon.

Rahul Rajan_21763

Hi,

This reply is on behalf of Prakash & Expert.

SN curve in OS is based on the stress range. Stress range = 2*amplitude.

Looks like Ansys is using amplitude instead of stress range.In future version we will have option to choose between amplitude and range.

Altair Forum User

Thank You, @Rahul R and @Prakash Pagadala for your efforts.

Based on your answer, I doubled the stress amplitude value and then substituted the resulting stress range in the formula S_r = SRI1(N_f)^b.

When I used this new SRI1 value(It is double the old value), I'm getting the results as expected.

One last question. Does the formula S_range = 2×S_amplitude hold good even for load curves that are not sinusoidal?

Rahul Rajan_21763

Yes the relation range = 2 amplitude is valid in any case, even on random loads. The rainflow count will be the same.

Altair Forum User

Thanks for the clarification.

Altair Forum User

@Rahul R Do we also have to multiply the fatigue strength value by 2 while defining FL in MATFAT like we are doing for SRI1 value?

e.g if the fatigue strength of a material is 100 MPA then do i have to put 200 for FL in MATFAT?

TIA

Altair Forum User

HI,

No, since Sr= 2*Sa we used this in the equation but for Sfl or FL it is not required to double the value.

Altair Forum User

@Prakash Pagadala What if i have Stress at 1000 cycles and not at 1 cycle. In that case I have to use Point, 1 segment method for SN curve in Fatigue Process Manager. Do i have to multiply this stress@ 1000 cycles by 2 for SR1000 input?

Kindly revert back at your earliest convenience.