Fatigue strength of thermoplastics

[tgmcg]

I'm designing some machine elements that are subject to fully-reversed alternating loads. The parts were originally milled from various grades of aluminum, eg. 6061-T6, 7075. Peak stresses per FEA are below 5ksi. We have design flexibility to reduce stresses even further.

I'm considering injection molding several of the parts using 30-40% glass or carbon fiber-reinforced high-performance thermoplastics, such as Lexan, Ultem, RTPU, LCP or TPI. (Nylon 6/6 is out due to moisture issues). However, I have been unable to find fatigue properties (eg. s/n curves) for any of these materials.

I'd be grateful for any guidance you may provide on the suitability and use of such thermoplastics for structural elements subject to alternating loads.

Many thanks,

Tom

 

[plasgears]

1. Contact the mfr for fatigue properties.

2. Don't be locked into reinforced polymers; experience shows that neat, tough selections have better fatigue properties. I had a plastic gearbox that operated into stall repeatedly, and Nylon 6 did the trick, by recommendation of the molder. The previous selection was acetal.

As a guide, evaluate candidates with high:

elong X yield pt.

This is a figure of merit for toughness and fatigue strength. It's analogous to the area under the stress/strain curve, which is energy absorbed before failure.

 

[tgmcg]

Plasgears,

Thank you. I'm not locked into any material selection until I can get a handle on the fatigue issue. Gotta be 100% certain before investing in tooling.

The tensile strength, modulus and elongation for neat RTPU are 6.5 ksi, 0.22e6 psi, and 160%.

For 30% GF RTPU, they are 20.5 ksi, 1.1e6 psi and 5.5%

For 6061-T6, they are 40 ksi (yield), 10e6 psi, and 17% elong.

So, in view of the +/- 5 ksi working stress, it would appear the 30% GF provides a significantly better safety factor on tensile strength with elongation comparable to the original aluminum. Stiffness/defection is also much better.

On the surface of it, neat RTPU doesn't appear to be up to the task.

Are their any specific neat polymers you might suggest?

Best regards,

Tom

 

[tgmcg]

Correction....the elongation of 30% GF is not exactly comparable to 6061-T6, but a whole lot closer than for net RTPU.

 

[strokersix]

McGuinness,

Why not machine some samples from neat material and test them? You will get a good idea of the material performance without investing in tooling.

I second plasgear's comments regarding neat versus reinforced materials. Try neat first. Reinforced materials have more processing issues compared to neat materials.

 

[tgmcg]

Strokersix,

Thank you.

I want to engineer the solution first via FEA analysis, to be followed by testing and fine-tuning if necessary....just as one would for a metallic component.

If the fatigue behavior of thermoplastic materials and components cannot be reliably predicted via normal engineering analysis, then thermoplastics may be the wrong approach entitrely. There are too many parts involved for a trial-and-error approach.

Best regards,

Tom

 

[TVP]

McGuinness,

It would beneficial to have some technical discussions with the various plastic resin suppliers, as they will have data on fatigue, FEA practices, etc. I have some good fatigue data that GE Plastics developed for their products. If you want a reference handbook, try this one:

Fatigue and Tribological Properties of Plastics and Elastomers

http://www.williamandrew.com/books.asp?id=88420715

or

http://www.knovel.com/knovel2/Toc.jsp?BookID=379

 

[tgmcg]

TVP,

Thank you for the references.

It's not clear from the Knovel reference whether fiber-reinforced compositions are covered. Do you know if they are?

Yes, it looks like I'll need to discuss more fully with the plastic resin suppliers.

The lack of fatigue data on the web is curious....possibly suggesting this might not be a common application. Then again, perhaps I'm reading too much into it.

Best regards,

Tom

 

[EdStainless]

Using polymers, you will either require a lot less load to get the same deflection, or you will deflect a lot further. Will the reduced stiffness hurt you? Could vibration become an issue?
Based on my experience I will echo what has been said, try to use a neat grade. If you need a filled grade for strength try to use non-fiber reinforcement. That will minimize the issues with directional properties.

= = = = = = = = = = = = = = = = = = = =
Corrosion never sleeps, but it can be managed.

http://www.trenttube.com/Trent/tech_form.htm

 

[tgmcg]

Ed,

The reduced stiffness and tensile properties for neat grades may be a show-stopper.

It all boils down to the availability of engineering data. If fatigue properties are not available, or cannot be expected to yield a predictable result (in which case they cannot be called "engineering" data), then I will have to use another material or process.

Best regards,

Tom

 

[TVP]

McGuinness ,

Yes, reinforced grades are covered in the Knovel version. It allows you to select various grades and plot them on a common graph, which is a nice feature. I just made some comparisons of some neat vs. filled Nylong 6,6 grades a few minutes ago. Maybe they have a demo version that you could review.

 

[BillPSU]

Before going to any plastic verify the environmental conditions of usage has no impact on the material. Temperature, sunlight, oil, water, humidity, shock loading, electrical condition are some of the considerations. Continuous use temperature is another concern and material creep and flow are other considerations. Mounting the plastic part may require modification due to low dynamic stress levels of some plastics. They may have high static yield points yet are susceptible to working stresses.
Talk to your resin supplier and get a full understanding of the material being specified. Plastics are not like metals in that the properties can vary not only by the resin supplier but also by the processor.

Bill

 

[tgmcg]

Bill,

Thank you. Yup, it's becoming clearer every day that plastics are a whole 'nuther world for us old-time metal-heads.

The environmental conditions for this application do not appear to be terribly harsh. It's for recreational equipment. Ambient temp between 32-120F, direct sunlight exposure, high humidity with intermittent splash and immersion, salt . It may or may not be stored under cover when not in use...I suppose we can make some assumptions about that and advise the customers accordingly.

The most highly stressed items will probably be cast aluminum for strength and stiffness purposes. We'll see...I need to develop a higher level of confidence and skill in the use of reinforced thermoplastics before deciding to use them for the most critical items. However, about 75% of the parts can probably be injection molded with little risk.

Developing this product has been a very interesting and educational experience, indeed.

I'll be heading down to the Pacific Design & Manufacturing show in Anaheim next week where I hope to plug many of the remaning holes in my knowledge base....which resembles swiss-cheese at this moment.

Best regards,

Tom

 

[unclesyd]

Check this website out and look at the box on the right hand side.

http://www.designnews.com/article/CA454623.html?nid=2334&rid=1942813189

 

[tgmcg]

Unclesyd,

Thank you.

Wow! IDES Prospector seems to do it all. Unfortunately it's
price is a bit steep and is overkill for my needs. However, a few days access to the package might fix me up.

I'll check out their demo options....

Best regards,

Tom

 

[strokersix]

Check this company for creating molded parts from most any engineering resin directly from your CAD model:

http://www.protomold.com

 

[plasgears]

It is worth every effort to try transforming parts to plastic, if at all possible. But be aware that plastic parts doing the same job as metal will look different:

- bigger,
- ribbed,
- constant thickness design,
- undercuts, etc.

I was involved in an ice crusher actuator that found its way into modern two door fridges. The incumbent article was expensive, heavy, and even larger. The new design had a ribbed plastic gearbox, plastic gears in the first few stages, and helical gears. The DC motor was parallel to the flat gearbox and allowed installation in the door instead of in the freezer space.

It created an industry wide sensation. Weight savings were multiplied over carload quantities, and it pleased everybody in the design, mfr, and distr system. It can be exciting working in what appears to be low tech areas.