Plastic behavior under relaxation followed by cyclic loading 1

[Tribomo]

Hi all, I have been browsing around this forum for quite a while now and decided to make my first post as I couldn't find the answer to my question anywhere.

I am currently designing a plastic part that will be subject to stress-relaxation, creep and fatigue, and I am struggling to make a good design/choose the right material as I am not certain of how a plastic will react to this situation.

First of all, the plastic will be compressed by 7% of its length.

The part, still subject to a compression of 7%, will have a compression stress of roughly 10-12 MPa added to it. This stress will sometimes be constant, sometimes cyclic.

With a strain of 7%, I am already passed the yield point on most plastics, so if I add an extra 10-12 MPa I will have an increase of strain until failure ???

If followed by the strain of 7%, I let my part relax until I have a reduction of 20 MPa, how will my part react to an increase of 12MPa, will it be an (mostly) elastic deformation ?

Hope this all makes sense. Thank you very much for your help

 

[Demon3]

Hello, your question is very relevant. Fatigue is one of the leading causes of plastics failure and yet one of the least tested properties because it takes too long. You can find some data but mostly for cyclic loads. Your application seems like a combination of creep (from the constant load) and fatigue from the cyclic load.

I would look for data for different polymers to see whether your application seems to fall within what plastics can do and contact suppliers who have data. I will look look for books that contain fatigue data. From memory I know that Polypropylene Definitive User's Guide and Databook does. Maybe I can find others.

Having said all of that none of it will tell you for sure whether your application will work or not. You will need to do testing at a lab set up for such testing. The only one I know of is

http://www.norner.no

I toured their site and saw such equipment.

Dr. Chris DeArmitt

Plastics consultant to the Fortune 500:

http://www.phantomplastics.com

Webinars on plastics, fillers & impact modification:

http://www.plastictraining.com

 

[Tribomo]

Dr. DeArmitt, thank you very much for your useful feedback.

Would you think in this situation fatigue may be an issue ? As we are far into a compression-compression cyclic ?

Is there any reason why compressive data (stength, yield, stress-strain curve, etc) are hard to find ? I would think that a stress-strain curve for plastics in compression and tension are very different, why giving only the tensile data then ?

I could reduce the 10-12 MPa load to a 4-5MPa if required. If we simplify the situation to a strain of 7% followed by a relaxation of 15 MPa and the followed by a short term stress of 5 MPa (no creep considered for the moment), how will my stress-strain curve looks like ? Is there any reference that could be point out to me in this regard ?

Thanks again for your help.

 

[Demon3]

Tensile testing is far more common and one hardly ever sees equipment for compressive testing although that is important as well. I have read guidelines for estimating compressive properties from tensile properties but the details are lost on me. I think it was in this book: Mechanical Properties of Polymers and Composites, Second Edition

I had a quick look for fatigue books and this one looks promising.

Fatigue and Tribological Properties of Plastics and Elastomers 2nd Edition - PDL

PDL books are very good and focus on practical aspects.

Dr. Chris DeArmitt

Plastics consultant to the Fortune 500:

http://www.phantomplastics.com

Webinars on plastics, fillers & impact modification:

http://www.plastictraining.com

 

[MikeHalloran]

I can think of only a few materials that could retain some useful mechanical properties after the specified initial strain of -7 pct, including most grades of brass, and the softer alloys of aluminum and steel, and neat PTFE.

I'd expect most other plastics to fail by cracking or crumbling. ... perhaps even developing flaws that will not reveal themselves except under tension or chemical attack.

Yours does not sound like a winning application for plastics.




Mike Halloran
Pembroke Pines, FL, USA

 

[Pud]

Have a look at the harder grades of polyurethane - TPU, or the polyester elastomers (e.g. Hytrel 8238 by DuPont - there are others)




www.tynevalleyplastics.co.uk

It's ok to soar like an eagle, but weasels don't get sucked into jet engines.

 

[Tribomo]

Dr DeArmitt

I will most definitely take a look at these 2 books. I am newish to plastics and I am trying to gather as much information as I can on this domain.

Mike

Yes, you are right. I realised that yesterday after hours of looking at different plastics. Some reinforced Nylon, PC with a strain @ yield could take the 7% compression, but then they couldn't take the 10-12 MPa stress without big strain. I also discovered that with relaxation, the modulus of elasticity decreases over time ? The plastic will have get to an operational temperature of 70ish degrees C, so that won't work. I could go with glass filled to get good creep/relaxation/(fatigue?) resistance, but then I couldn't have a strain of 7%. I'll have to go back to my design and change the geometry so the initial strain will be reduced.

Pud

Softer material would be a good idea, but I forgot to mention that the stress of 10-12MPa can't generate a strain more than the original strain (7%), so a moderate stiffness is required


Thanks all for your feedbacks

If it is of any interest for anyone, I have found answer to my questions on the behaviour of plastic with complex loading (relaxation + creep, etc). There is a nice paper called Experimental study of cyclic visco-elasto-plastic behaviour of a polyamide fibre strap.

Cheers

 

[Pud]

Some of the hard grades of TPE do not feel like elastomers. They don't bounce (much) either.

www.tynevalleyplastics.co.uk

It's ok to soar like an eagle, but weasels don't get sucked into jet engines.

 

[desertfox]

Hi

Whilst its not good to over stress the material I don't see where fatigue fits in, to my knowledge fatigue cracks only grow under tensile stresses.
That said I am no plastic expert I only normally work with metals.

 

[Demon3]

Interesting comment. Would you agree that for an unconstrained sample under compression it will bulge leaving the outer skin at the sides under tension? Maybe that's the explanation.

Chris DeArmitt PhD
President - Phantom Plastics LLC

Consulting, ideas and training on plastic materials

http://www.phantomplastics.com

 

[desertfox]

Hi Demon3

I would agree that a rod under compression would expand in the transverse direction but I think that any associated stress would be very small indeed, usually in metals if you stay roughly less than 25% of the tensile strength whatever the application your component won't fail in fatigue.

Looking at some examples of rods in compression or tension it appears they use poissions ratio only to obtain the transverse strain but the stress is not calculated, although I believe that the transverse strain is more to do with the shear stresses acting transverse to the applied load.

If I am wrong anybody please shout up

 

[Compositepro]

Desertfox, by your logic a polymer rod in a compression test would never crack. They do. The tensile hoop stresses and shear stresses that occur during a compression test are quite signeficant.

 

[desertfox]

Compositepro

My reference to cracks growing was in relation to fatigue, fatigue normally occurs under tensile cyclic loading well below the materials yield.
I searched to try and find fatigue failures under compressive cyclic loading but had no luck.
If you can provide a reference for fatigue failure under compressive stresses or the compression test you mention in your post I would be grateful.

Desertfox

 

[desertfox]

Compositepro

See this thread where you actually say the same thing as I have.

Thread2-204767

 

[Compositepro]

In that thread I say the same as I do in this one. In the real world, when there is a compressive strain in a material in one direction, there will be tensile strains in the other directions. These tensile strains, caused by compressive loading, can cause cracks and fatigue.

I only take issue with one comment you made:
"a rod under compression would expand in the transverse direction but I think that any associated stress would be very small indeed"

Associated stresses are not "very small". They are proportional to Poisson's ratio. Which is significant.

 

[desertfox]

Well I've just made an example up, take a solid steel bar 25mm diameter and 100mm long, place 5000N force on each end, the compressive stress is around 10 N/mm^2, now the stress in the other direction based on the axial strain and a poisson's ratio of 0.3 calculates out at 3N /mm^2 so just less than a third of the compressive stress.
So the load required to achieve a high tensile stress in the transverse direction would have to be very high, so high that the compressive stress would probably exceed the yield stress of the material before the tensile stresses where high enough to cause failure by cyclic fatigue, this is also stated in the post I provided the link too previously.

The OP asked about a fatigue test and not just about a compression test which you mentioned in your post.
I would agree that in a compression test a force could be applied so high that cracks could appear but then we wouldn't be looking at fatigue.
Cracks can also appear in the material due to the shear stresses acting at 45 degrees to the axial stress again this as nothing to do with fatigue.

 

[Compositepro]

The OP was talking about plastic, not steel.

 

[desertfox]

Indeed the OP was talking about plastics and I stated in my first post I was no plastic expert.

I looked on this site

http://www.toray.jp/plastics/en/products/amilan/technical/tec_001.html

if you look at the stress v strain curve (B) it has a linear portion up to a yield point and a modulus, given at poisson's ratio of about 0.4 for Nylon66 ,it would be reasonable to get a similar result for this plastic material in accordance with my earlier example.

If you have a reference to material that counters my standpoint then please post it.