Fatigue behavior of engineering plastics

[klamka33]

Hello
I've recently started to deal with fatigue analysis of plastic components.
I found some strange relationship innthe fatigue curves - where the S-N curve coming from the 0-max (pulsating tensile) cycle is higher than the curve from the alternating one (compression-tension). Looking to the literature, I also found the following statement: "Specimens may be stressed in tension only, compression only, or both tension and compression, which is generally considered the most severe situation." (Introduction to Fatigue of Plastics and Elastomers, McKeen)
This seems to be in conflict with the theories of mean stress correction, which basically say that the allowable variable stresses decrease with increasing mean stresses.
Do you have an idea what's the reason for that dicrepancy?
Found similar behavior also for other plastic like polioxymethylen - suppllied by Celanese to r DuPont.

Best Regards
B.K

 

[SWComposites]

The theories are likely based on metal fatigue data. Plastics and composites behave differently.

 

[klamka33]

I agree that plastics will behave differently, however I would not expect totally inverse relation - where tensile mean stress is going to improve the lifetime
To support my feeling I'm attaching an article where several polymers (PC, PP, ABS) have been tested at different R ratios and each of them have shown a detrimental effect of tensile mean stress
Even looking at PC polymer which should be similar to Iupilon that I brought at first post I see that cycle with R=0 is clearly worse than symmetric cycle (R=-1)

 

[rb1957]

maybe the data is saying that plastics are susceptible to fatigue in compression ?
or that alternating stress sign is more critical than pure tension ?

maybe poisson effects ??

another day in paradise, or is paradise one day closer ?

 

[klamka33]

Hi rb1957
maybe the data is saying that plastics are susceptible to fatigue in compression ?
According to curve from my initial post, material is suspectible to fatigue in compression, for higher stresses than for tensile or alternating load

or that alternating stress sign is more critical than pure tension ?
That's what is actually present in data from supplier, trying to understand if it's correct and why is that...

maybe poisson effects ??
Hmm, that's interesting point. Poisson ratio is 0.44 for POM and 0.37 for PC, how it may affect the fatigue?

I've submitted a question also for DuPont, hope they will answer to that

Thanks,
B.K

 

[rb1957]

You may get a better answer from a university

another day in paradise, or is paradise one day closer ?

 

[SWComposites]

If you have reason to question the data, then run your own tests.

 

[SWComposites]

There is lots of polymer composite fatigue data showing R=-1 is way lower than R=0.1 .

 

[3DDave]

It makes a bit of sense - under tension the molecules that can shift will align with the load while under compression they will tend to fold out of the way. Compare to a piece of rope but at the molecular scale. Add in fibers and the behavior is even more different from metals.

 

[klamka33]

You're right, there could be something happening in microscale that will cause such behaviour.
Of course in composites the story can be completely different than in other materials

For now, I'll stick to use the reversed loading curve, with "Goodman like" mean stress correction suggested by literature, that do not take any benefits of this lifetime increase in tensile-tensile region

Thanks,
B.K

 

[goutam_freelance]

Test results show composite fiber-metal laminates (Glare) may be stronger in fatigue than monolithic 2024-T3 .

Engineers, think what we have done to the environment !

https://www.linkedin.com/in/goutam-das-59743b30/