brass insert pull-out in Lexan

[cambria]

Recieved a large delivery of Lexan parts that fail in torque test. There is a small threaded brass insert molded into the housing. The insert acts as a captive nut in the housing, and the housing is clamped onto a mounting bar. This normally takes foot pounds, this shipment didnt even take 10 inch-pounds before the inserts started sliding out.

I assume no dry-out of resin prior to injection. Sure, perhaps a hopper drier, but nothing before. But this is a guess.

Two questions:
1) cause?
2)is there any salvation? Can I dry out the housings (like anneal them) or are do I start from scratch? That puts me behind 3 months if I am lucky..

 

[CoryPad]

Poor insert retention is due to low bonding (chemical and mechanical) strength. Possible causes include porosity, contamination (on insert or from a release agent), and thermal expansion mismatch/stresses. There are possible salvations, such as using heat or ultrasonic staking of inserts.

 

[cambria]

We have an ultrasonic welder in house and ready to go, as well as a CO2 laser that I am considering aiming in the direction of the vendor. Actually we have a couple of heat stakers as well somewhere. Would you think zapping the insert with the Ultrasonic welder would do? I am thinking of a small rod-like tool impinging upon the insert and giving it a burst.

 

[cambria]

Just ran over and tried heat staking, in the sense of a small soldering iron heating up the insert. I kept it there until I saw an ever-so slight liquid ring around it. Let it cool. No dice, it held a little bit longer then gave way and slid out. Tried it a couple of times, willing to try more.

 

[CoryPad]

What is the hole condition? You may need a larger diameter insert if the hole has has changed. Perhaps you have some poor PC material - try testing the bulk mechanical properties of one.

 

[Demon3]

If remelting the plastic didn't work then your problem is probably that of release agent or oil on the metal. If that is the case then remelting by heat or ultrasonics won't help because the oil will still be there ruining the adhesion at the interface.

Chris DeArmitt PhD FRSC CChem

http://www.phantomplastics.com

Consultant to the plastics industry

 

[cambria]

Something smells in Denmark. Its just that its not from Denmark. Both comments ring true. It is very easy to think they simply opened a box of inserts and started throwing them into the mold without any cleaning. Perhaps inserts come pristine andthat is not necessary to do. But it is looking like a manufacturing process. I will check hardness but have to do that tomorrow. I am still in the lab, will run down, push a couple of inserts out, dry them off with acetone, and re-insert them then heat. I know its not the best but if by a miracle the inserts hold, it would seem that the cause has been discovered. Do you think drying them out say in 100F overnight will do anything?

back in 15-30.

 

[cambria]

Pushed one inseret through, cleaned with toluene (its what I had) then re-inserted. Ok, not good experimentally on at leaset three points, I know. I then heated the insert with a soldering iron as before. I believe I had more resistance than before - but it still went at 8-10 inch pounds. In an analysis of multiple hypotheses, this gives perhaps a suggestion that cleaning the insert surface improved the torque.

Thoughts: weak Lexan due to humid resin, weak Lexan due to poor process control, weak insert adhesion due to cold inserts or oily inserts.Any others?

 

[Demon3]

Environmental stress cracking on the Lexan due to oils or solvent is another possibility but then I'd expect you to see crazing.

Too low injection temperature so the PC doesn't wet out the metal surface, thereby giving poor adhesion.

Design tests for each hypothesis and you'll soon know the problem.

Chris DeArmitt PhD FRSC CChem

http://www.phantomplastics.com

Consultant to the plastics industry

 

[cambria]

Thanks. The circle is getting smaller. The lead time will be a killer now.

Thanks to all.

 

[patprimmer]

I agree with the above.

Presuming good original design (has this worked OK in the past) possible causes are:-
1) PC degraded due to inadequate drying of pellets.
2) PC resin inferior to previous batches due to low molecular weight.
3) PC grade has mould release chemical added.
4) Mould not fully packed.
5) Inserts not pre-warmed.
6) Inserts contaminated with oil or aromatics.
7) Inserts cleaned with aromatics or chlorinated hydrocarbons and residual solvents causing ESC.
8) Inserts have insufficient undercut or surface texture.
9) The mouldings are a different colour to the original batch or even if the same colour, have a different colour or masterbatch formulation.
10) The moulding machine operator started using a mould release spray or a different spray or a different amount of spray.

Except for insufficient pack all the other causes cannot be corrected in existing parts.

If under packed, it just might be possible to weld under pressure at increase localised pack and better grip the surface, but I very much doubt that can work.

Another costly fix might be to make up a jig to pull the existing inserts then use inserts designed to be applied with ultrasonics. They will have a larger undercut and possibly be barb shaped so they go in easier than they come out. This will depend on the wall thickness where the inserts are located.

I am at a bit of a loss to know why it will take 3 months to get replacement mouldings. Maybe you need another moulder.

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

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

for site rules

 

[cambria]

Thanks, thats a good summary. Parts have worked extremely well in the past. We supply the resin (GE Lexan, now Sabic)and that end of it I feel is controlled. Three months - one month to get the Lexan here (special colorant) plus two months turn including shipping. But if we can resolve the issue this will get expedited of course. The question is do I risk this or go with a different molder?

 

[patprimmer]

It is sounding like you have a moulder a long way from home. These issues are more easily solved if the moulder is close by. The greater the distance the greater the need for trust of integrity and competence.

If your moulder is in China, maybe you need to visit him for a week to ensure he is doing what is required, when required to fix this, otherwise move the mould.

If you choose to move the mould, the next problem might be actually collecting it on time and in good condition. That might also involve personal supervision and transport organised and paid for by you.

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

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

for site rules

 

[cambria]

Yes, a long way from home. I dont think there will be a problem bringing in the tooling if we bring it here. Our operation is rather old fashioned in that we are vertically integrated, except for the plastic injection. We invent, design and build it all. I am doing my best to resist doing the injection work in house. But next time its local for sure, if not closer.

 

[cambria]

Update:
Strong focus on the wrong insert for the application. High level of stress concentration at insert interface due to wrong insert surface (barbed) and thermal differential. Possible use by molder of mold-release compound. May very well go to post-mold ultrasonic welding. One more try then the tool comes home.

Thank you all for the good insight and experienced opinion.

 

[cambria]

More: Tests show the Lexan is in good shape and has not been disadvantage by excess humidity nor by release agents. The breass insert used was a straight cylinder with barbs, probably unheated and of unknown level of cleanliness. Thermal stresses are the probable culprit, aided by the additional stress concentration supplied by the little barbs on the insert. We are flying a 55 lb bag in for a final effort of heated inserts, cleaned, with a smooth finish, preferrably in a barbell like shape. Will come back in a week or so with results to provide closure.