Fatigue analysis of materials - when stress above yield

[morris9791]

Dear Experts,

We have certain design criteria that certain components must adhere to e.g. No structural failure and some small localised yielding acceptable, other components no yielding allowed etc.

When it comes to fatigue however, if one comes across some localised yielding, can this be acceptable in fatigue analysis or must all stresses be below yield before even considering such an analysis?
We use the strain life approach which I understand accounts for local plasticity (Neuber method).

Thanks
Ed

 

[rb1957]

your design criteria seems very specific. presumably your fatigue criteria is just as specific ... ie required service life, required safe-life factors. so if you achieve these with the best analysis there shouldn't be much problem. though if i was the client i'd've asked for testing to support the analysis.

 

[corus]

It's acceptable in fatigue analayis, depending on your allowable limits. For materials (away from welds) there are limits which will give an infinite life. Without looking I think infinite is defined as being above 1e7 cycles based roughly on a stress range of less than half UTS, and where the stress is factored to take into account surface finish, thickness etc. Above that stress the number of cycles to failure will be determined by the stress range and the SN curve for that material, whether it's above yield or not.

corus

 

[morris9791]

Thanks Guys,

That I understood, I just wanted some backup and ease of mind

Cheers

Steady Eddie

 

[gwolf2]

> When it comes to fatigue however, if one comes across some localised yielding, can this be acceptable in fatigue analysis or must all stresses be below yield before even considering such an analysis?
We use the strain life approach which I understand accounts for local plasticity (Neuber method).

What are you asking? In the first sentence you seem unsure about how to approach high stress yielding, and yet in the second you say you use strain-based fatigue based on Neuber corrections.

What's the question?

 

[Stringmaker]

What you're referring to is also known as Low Cycle Fatigue (strain life based). I work in the gas turbine engine industry and this is very common in most parts which are hot, spin fast, or a combination of the two. Depending on your parts and requirements you may also want to consider a Fatigue+Fracture based part life. By convention, fatigue life is typically the number of cycles until a crack initiates. Depending on your application and specific part characteristics it may also be acceptable to take some sort of benefit for the interval of part life between crack initiation and fracture. That method is widely used within aerospace.

 

[morris9791]

Gwolf2 and Stringmaker,

I was just a little unsure in the back of my mind despite what the strain life approach was telling me

Thanks for your responses guys,
Much appreciated.

best Regards
Steady Eddie