Design to meet drop spec 2

[bvanhiel]

Are there any methodologies for estimating the survivability of plastics during drop?

I'm currently taking two approaches. The first is using FEA to estimate the amount of strain energy stored in the part at yield and relating that to the energy of the drop. I trust FEA to tell me where failure is likely, but I don't really trust it to tell me the survivability. I only have access to linear FEA

The second approach is to compare material properties to a prototype that we broke during a drop. My first intuition is to compare the resilience (Sy^2/2E), toughness (total area under the stress-strain curve) and charpy impact of cadidate materials to the material of the prototype. I'm more than a little surprised that the resilience and/or toughness does not correlate more directly to measured impact.

I'm inclined to treat the charpy impact as the best predictor. My intuition is that a material with 2X the impact strength should survive 2X the drop height.

Straight PC (no glass) looks like the best choice of reasonably common engineering materials.

Thanks,

-b

 

[patprimmer]

Dropped dart is by far the best indicator of real world impact strength as in impacts caused by the article being dropped. Charpy and Izod have the part restrained by a clamp and this has a profound effect on strength vs unrestrained.

Also many materials are very notch sensitive which has an effect real parts vs test plaques.

High elongation at break also improves dropped part impact strength. As an example, Acrylic has reasonably high Izod, but really low elongation. In the real world it breaks easy.

Most material suppliers quote Izod and Charpy because they reproduce well from day to day or lab to lab, but they do not translate well into real world performance.

There are a lot of factors apart from actual lab results on test plaques that influence dropped part impact test.

PC does indeed have very good impact strength for single impact in a clean solvent free environment but often fails where even simple everyday chemicals are involved or where fatigue is a factor.



Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[bvanhiel]

Pat,

I'm guessing then that even with a benchmark I could probably only predicte a trend rather than the actual performance using datasheets.

Would you rank candidate materials by impact first and then elongation? I don't think falling dart data is available for too many materials.

It's a consumer product, so exposure to household cleaners is possible but not likely. Presumably this is why many power tools are glass filled nylon? The gf nylons we looked at had poorer impact properties on paper...

-b

 

[patprimmer]

The impact strength ranking of the parts from different materials will depend a lot on part design.

For instance, some materials have poor weld strength, some have more susceptibility to reduced strength due to moulded in stress, some have higher notch sensitivity, some have higher elongation, some are very anisotropic.

Therefore if you have a stressed point, a material with higher elongation might fix it, however if you have a sharp notch, glass fibre might help so long as it flows accross the notch.

PC, ABS, PVC, supertough nylons, polyurethane, LDPE and many others can all be tough depending on circumstances and other properties required.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[cpretty]

From my experience, impact is a bit of a funny one to predict. Sometimes things are counter intuitive.

Izod impact is usefull, but doesn't take into account natural damping etc and as previously mentioned is very specific to a certain geometry. It is usefull when looking at different grades of the same material ie PC, but not so good across materials.

In terms of design you can look into FEA, but the devil is in the details that FEA does not accurately predict.

The only real answer is to try it and see. Drop it from every angle, because everything has a weakness. Unfortunately sometimes that means tooling and crossing your fingers. Having backup plans is always good.

The challenge is the balance between absorbing energy and transferring energy. Stiff areas transfer the energy well (but to where), but can also are good at transferring energy directly to weaknesses in the part (sometimes a long way away from impact location). Springy areas are good at soaking up some of the energy, but care must be taken to tolerate the deflection. It is where the stiff and springy areas meet where the failures seem to occur.

The other big issue is sharp edges causing stress concentrations. Watch the radii at the bottom of ribs and posts. Especially watch for toolmakers taking shortcuts on breaking edges at the base of ribs.

To illustrate the issues with design for drop test of a product:

We were trying to improve the impact resistance of PC parts at -35°C. We had to do both a ball impact and drop test.

A certain area would continually break. We added ribs, and it broke more easily. We machined off all the ribs and it still broke. The answer was to taper the ribs to spread the load.

The moral of the story for us was to create gradual transitions from stiff areas (esp screw bosses) to springy areas, and to really watch for sharp edges on ribs.

Another usefull trick is to go back to the 'old school' methodology of modelling parts. With PC or ABS it is quite easy to fabricate ribs, bosses and posts, then solvent them in using Methylene Chloride. If done well you can test theories without modifying tooling.

Cheers,
Craig

 

[patprimmer]

I agree about different grades. With homopolymers, molecular weight can have a big influence on impact. For instance a CD grade PC will have a lot lower impact than an extrusion grade, however in a thin section, the extrusion grade will suffer from moulded in stress.

Without seeing the part and knowing the application, it is impossible to say.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[bvanhiel]

The part has a "u" shaped section with a hole at the base of the "u" that needs to support a large bending load at impact. I know the hole will be the high stress area and have reinforced it as much as I can get away with. This is the area that cracked in the prototype part. Knowing that this is the location of highest stress I can locate the gate to move the knit line away from the high stress area. The part is not joined to any other components in the "u", so I don't really have to worry about deflection of the part trading loads to other components. The part similar to the handle on a hand tool in terms of size.

 

[patprimmer]

Why could you not use Pp or PE. They are both cheap and easy to mould and will survive an unspecified drop test under unspecified conditions with unspecified other requirements

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[bvanhiel]

Pat,

I don't think I can give you enough detail to solve this problem on the forum. At any rate, it sounds like the real answer is experimentation. What I'm hearing is that impact data is really only good for ranking materials and can't be a used to predict relative performance.

The answer that I was hoping for is a way to relate the drop test performance of two materials with a fixed geometry. I know that a cast urethane part survived the drop from a fraction of our spec height. I'd like to have some confidence that going with ABS (the intended final material) or PC with 3X better impact(or resilience, or toughness) will get us a 3X higher drop when the failure is at a stress riser.

Craig,

I think I've reached the same kinds of conclusions about part geometry for energy absorption. We're using the FEA to optimize the amount of strain energy absorbed while minimizing local stress.

-b

 

[Pud]

Bvanheil: I'm not a great fan of Du Pont, but their impact modified, 30% GF PA6 is a very, very tough material. (I do not think it's on their "standard" product listing, and I cannot recall the grade at the moment. When the wife gets back will post the grade...

We use it for inserts in track laying vehicle track parts. Serious impacts.

If you need stiff as well as impact, have a look at EMS Chemie "Grivory" range. Pricey but excellent stuff.

Others probably make similar. Just my experiences.

As far as ABS is concerned, there are as thousands of grades to choose from - high impact, high heat, high gloss, blah blah blah...

We have a customer who we make hollow ABS balls for (made in two halves, glued together, centreless ground to give less than .0005" roundness. They have what they call the "Bal Cony" test. They drop them from a second floor balcony onto their car park!! If they survive they are ok!!

H

http://www.tynevalleyplastics.co.uk

 

[patprimmer]

Harry

We have a similar deal here. It is for street lights and is the Pelican test. A weight deemed to be as heavy as any Pelican is likely to be is applied to the light for a prescribed time.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[cpretty]

The SAM test is a similar one in use here. Standard American Male - 150kg.

 

[Pud]

Zytel grade I was trying to remember is 73G30T.

Others make similar.

Harry

http://www.tynevalleyplastics.co.uk