fatigue analysis to obtain the damage of a given number of cycles

[LIAM_7]

Dear all,

First of all, thanks for taking the time to check my thread.

I am a beginner with fatigue analysis. I need to obtain damage due to the application of 40.000 cycles of pressure inside a pressure vessel. I am using ANSYS Mechnical for this.

The thing is that I can carry out a static structural simulation and then use the fatigue tool to estimate fatigue life of the stucture (number of cycles before failure), damage, safety factor...

But I have no idea how to impose a given number of cycles (for the static results obtained) and compute damage (should be of course D<1).

Is it possible to do this in ANSYS Mechanical?

Any tips will be highly welcome.

Mnay thanks!

 

[TGS4]

What pressure vessel code are you using?

 

[LIAM_7]

Hi TGS4,

Thanks for your answer.

I am following 13445-3 Direct Route.

Maybe I am wrong, nut the text mentions a procedure to obtain N (number of cycles)based on strcutural stress, correction factors (welded and unwelded regions...), but I didn't understand how to apply a certain number of cycles based on a linear elastic static stress analysis and obtain damage, safety factor...

I am using ANSYS mechanical.

Is this specified somewhere in the norm and I missed it? Or it is not possible to fix the number of cycles and get the fatigue response of the structure due to the imposition of this fixed number of cycles?

Thanks in advance for your support!

 

[hacksaw]

Your dynamic analysis provides the cycling rates and the dynamic stresses.

 

[LIAM_7]

Hi hacksaw,

Thanks for your answer, but I don't get the answer to my questions:

1. Is it possible to impose a given number of cycles after a static structural analysis in ANSYS Mechanical to obtain the damage due to fatigue phenomena produced by this given number of cycles? In other words, I want to know the fatigue response of my vessel after 50000 cycles of operation.

2. Is the code 13445-3 The direct route providing a method to assess fatigue for a given number of cycles instead of predicting number of cycles for a certain stress range?

Many thnaks and I apologise if I am not properly explaining my problem

 

[Morcuse]

I would highly recommend that you go through the ASME PTB-3 examples, and that you find a methor to go over the details with you.

 

[LIAM_7]

Thanks Morcuse for your answer.

Do you have an answer to the questions posted?

Thanks!

 

[TGS4]

LIAM_7 - you need to stop focusing on your calculation tool (ANSYS) and focus instead on the analysis methodology. Go through the EN-13445-3 method first, then figure out how to use your calculation tool.

 

[LIAM_7]

Hi TGS4,

Thanks for your advice and for telling me what I have to do.

Do you have an answer/recommendation related to the specific questions posted?

Thanks

 

[TGS4]

The answers to your direct questions:
1. Yes - follow the EN-13445 procedure.
2. The damage fraction is calculated using a Miner's rule approach: D=n_i/N_i. Again, this is covered in the procedure.

 

[LIAM_7]

Hi TSG4,

I was now expecting the answer I found you provide several times in the forum: pay for an expert

Have a nice day

BTW first question is related to ANSYS software...no need to read a single word from 13445 to answer it...

 

[centondollar]

BTW first question is related to ANSYS software...no need to read a single word from 13445 to answer it...

Common sense should answer it. If you run one linear static analysis, you get one result - no cycles involved. If you run several linear static analyses with all probable load combinations, you get stress ranges including maximum and minimum stresses.

The typical code approach is based on calculating stress ranges (max/min) and then using empirical rules and empirical factors to consider resistance to fatigue. This answer was already been given by TGS4.