ABS characteristics in low temp service

[nigelt]

We have an issue with failure of inj mld'd ABS parts due to stresses from a co-eff of thermal expansion event - actually not an event, more of a sustained load. The failure is occuring at low temps (-10, -20 deg C : still awaiting info). The assembly is an ABS holder for a disc made of ceramic. So the ABS wants to shrink, but the ceramic restrains it from doing so.
A quick calc suggests that the stress is below 1000psi (650 - 800psi) and as I have understood it, that should be safe enough - except that I suspect that the safe working stress drops off quickly below a certain low temp value.
All of which is a long winded way of asking, if this is true, where I'd get curves of this effect for generic matls. Sure, I can go to resin suppliers, but are there generally accepted low temp thresholds for say, ABS, PC, PBT etc?
Thanks folks!

"The ideal client is one possessed of great good sense and perfect judgement;
that is to say, one who agrees entirely with the designer at all times and in every respect."
Fenwick Williams, Naval Architect.

 

[craigj]

There are some really good books out there regarding temp related material properties for engineering resins.

Try searching:

http://knovel.com/knovel2/default.jsp

This is a online reference book subscription sevice (download PDF pages), but I think that you can get a free trial membership. Anyway I cant think of any particular titles right now...standby

CJ

 

[craigj]

1. Effect of Temperature and Other Factors on Plastics
By: Plastics Design Library Staff © 1990 William Andrew Publishing/Plastics Design Library

2.Effect of Creep and Other Time Related Factors on Plastics and Elastomers
By: Plastics Design Library Staff © 1991 William Andrew Publishing/Plastics Design Library |

Cheers,
CJ

 

[nigelt]

Thanks CJ,
I checked out the first reference, and whaddya know, the curves are between +56 C and 200 C.
Of course that won't help at all.
Investigation continues....

However, I found a useful perspective in the book "Structural Analysis of Thermoplastic Components", Trantina & Nimmer Ch.5 Table 5.3 (P.213) and Ch.6 P.285 where reference is made to the brittle failure of ductile plastics. These under the general heading 'temperature effects upon failure mode'.
Thx., Nigel

 

[craigj]

Nigel,

The reason their curves start at 53C is probably because that is the transition temperature.

I did find a reference for Cycolac T (ABS)

As you would expect your yield actually goes up by 50%
Flextural yeild @73F = 10,500psi
Flextural yeild @-40F = 15,600psi

Probably a bigger concern is the impact properties
Izod impact @ 73F = 6.5 ft-lb/in
Izod impact @ -40F = 1.5 ft-lb/in

Of course different blends will have different properties but this at least gives you a good understanding.

cheers
CJ

 

[nigelt]

Thanks Craig for your continued interest.
The real lesson, or sharp reminder here is, I think, that below particular temperatures we must stop thinking of polymers as ductile behaving materials and be diligent in considering them as brittle.
So (for injection moulded parts) generous fillet rads, minimal knit lines, minimal residual stress creating features, minimising or eliminating regrind %'age, avoiding solvent cleaners that might result in enviromental stress cracking, are doubly critical.

The considerations seem - according to lit - to be:
- temperature.
- notch sensitivity.
- rate of loading.

Incidently, PC/ABS blends seem to be among the best performers for impact at low temp. Seems to go ( for these common matls) PC/ABS better than PC better than ABS, but this needs confirming.

Regards,
Nigel


"The ideal client is one possessed of great good sense and perfect judgement;
that is to say, one who agrees entirely with the designer at all times and in every respect."
Fenwick Williams, Naval Architect.

 

[craigj]

Actually my understanding is that polycarbonate is the most impact resistant material by itself (most energy absorbtion, even at low temp). The reason that ABS is added to PC is for molding purposes (it flows nice). Different companies use different amounts of each for their alloys.

Fun stuff!
cheers
Craig

 

[RichGeoffroy]

Nigelt:

Stress can be a difficult concept with which to design. Stress is both time- and temperature-dependent. Not only is it difficult to assess the actual stress conditions of the part during service, but what does one use for failure stress --- deviation from the proportional limit, engineering yield stress, or the stress at break? Further, one must consider under what conditions should the reference data be developed, at a high strain rate, or low strain rate, at what temperature? All of these constraints make it difficult to design a part using data sheet properties. (Review Plastics Engineering FAQ712-985, “The Fallacy of Tensile Strength”.)

It’s often better to design a part using strain limits. The normal rule of thumb is to stay under 2% strain for a part which might experience a periodic short-term load. (A word of caution here, too many short-term periodic loads can lead to cyclic loading failure. Therefore a lower initial strain rate would be required under cyclic loading conditions.) When a part is expected to be exposed to a constant stress or strain for long periods like weeks at a time, it’s better that the part not experience a strain in excess of 1%. For parts that are expected to last many years under constant load, the part should be designed to limit the initial strain to between 0.5 and 0.75%.



Rich Geoffroy
Polymer Services Group
polyserv@cox.net

 

[RichGeoffroy]

Nigelt:

Please note that the parenthetical in the second paragraph in the previous post should read "lower initial strain", not "strain rate".



Rich Geoffroy
Polymer Services Group
polyserv@cox.net