Problem with leakage of welded Nylon components

[PLCRookie]

I am currently trying to establish the root cause of some leakage problems that we are experiencing with welded plastic components. I have been working on this part for the last year, and am really at a stand still now. The part is a small canister (approx. 3" diameter)that must not leak after welding, and be able to withstand pressure up to 300 psi. When I initially was assigned to this project, the burst pressure was the major defect that we saw. I had wild variation in the burst pressure and leakage was manageable (1% max). All the parts are 100%leak tested (bubble emission @ 80 psi), so we were o.k. there. I have changed the material to a prime Dupont Zytel 70G33LBK031, to try to correct the burst strength issue, and have had excellent results. Burst pressures since the change have not fallen below 450 psi, and are much, much more consistent. Part appearance is greatly improved, but I am now seeing huge leakage rates at times (up to 60%). It seems to be lot specific, as one lot that we just finished was .25% failure due to leak, and the previous was 45%. I have made multiple visits to the molder, but still cannot find the root cause of the problem. Now for the question....I would like to have some recommendations for testing (laboratory)that could give me some better information. We have tried some labs in the past, but we really didn't get anything we could use (other than "you need to do more testing...that will be $$$$$"). The parts are ultrasonically welded, and I am quite confident it is a material issue. What specific test could be performed on these parts?, and who can do it?
Any ideas or reccomendations would be greatly appreciated..

 

[mcgyvr]

How fast after molding are the parts welded?
Are the parts dried prior to welding?
Due to the glass fill additional weld area may be required versus the previous material.

 

[unclesyd]

Have you contacted DuPont's Technical Service?

 

[PLCRookie]

Moisture issue has been tested ad nausem. I actually have a commercial dryer in-house on loan from the maker. We weld the parts as soon as they arrive from the molder (usually within 2-3 days of molding). the parts are shipped in sealed bags with dessicant. I have tested the leaking batches after drying for 2 and 4 hours (165F -40 dew point). No change.
Thanks for the help.

 

[mcgyvr]

http://plastics.dupont.com/plastics/pdflit/americas/general/H76838.pdf

See section 11

 

[PLCRookie]

Weld area does not appear to affect the leaking. I have welded parts at different depths to check this theory, but the results were unchanged. With the previous material the weld depth had to be maximized to get the strength, but with the Zytel, minimum welds result in acceptable strength. The part has overall lenght tolerances, so the weld depth can only go so far. Excess depth of weld also can generated excess flash and part marking, which is undesirable.

Again, thanks to all.

 

[Compositepro]

In composite processing I often use glass fibers as breathing paths to allow air flow but not resin flow. It may only take one unwetted or fractured filament to create a gas leak. The ultrasonics that welds your polymer will not have good effects on glass fibers.

Fiberglass and carbon fibers are often used for making pressure vessels but almost always require a liner to prevent leaks.

I suggest you locate a leak and look at it under a microscope. 100 to 200 power would be adequte for this. It is often amazing what you can see when you actually look.

 

[patprimmer]

What was the previous material. Was it 33% glass filled nylon 6.6. Was it black. Have you tested the melt viscosity of both materials before and AFTER moulding. If the previous material was wet when moulded, it would loose molecular weight and therefore melt viscosity and strength, but would weld better. same deal if it were nylon 6. It would flow better and weld better but be weaker, same deal for glass content. more glass equals more strength but poor weld strength.

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

 

[PLCRookie]

Pat,

The previous material was Entec Hylon (basically carpet scraps). It was 33% glass filled nylon 6/6.I went thru the whole process of drying, etc.. but the variation of the materials was just too great to get any consistent results. The change to Dupont was extermely benificial, as the burst strength is solved as far a we are concerned. I just cant get the leakage figured out. I have samples of good and bad parts (welded and unwelded), but I cannot cut/section magnify, etc. I really need a resource for this, and am trying to get to Dupont to get some help.
We recently went to an outside lab, but it feels like we were taken, and management has lost its taste for more testing form those guys.

Do you (or anyone have a number or contact at Dupont? When I tried to contact them, I got shifted around quite a bit.

 

[unclesyd]

Have you tried going through your supplier?
It sounds like you purchase enough material to get some quick help.

Have you tried any Vydyne materials?
I'm very close to them, physically, and maybe able to get some help.

What does the welder supplier say?

My numbers for the old Technical Support Center are no longer valid. I believe that they have broken this group up by divisions.

 

[Pud]

I have mentioned before in posts the general lack of support from DuPont - unless you use a zillion tonnes per annum - others have found different.

The best support we have had is from EMS Chemie. The materials are much better for processing and nothing is too much trouble. (We are a very small user too!)

I would suggest that you try and contact your local EMS agent - you may be surprised.

(I have no commercial interest in EMS - just a user)

We are in the UK, but have heard the same from other sources.

Q: Why 6.6? Your comment about carpets would suggest PA 6 as (afaik), PA 6.6 is not used for yarn, but I see that it is PA6.6?

Cheers

Harry

http://www.tynevalleyplastics.co.uk

 

[patprimmer]

On the very limited data provided, I think 25% Gf 6 might do a better job re balance of properties. Like Pud, I would expect caret croppings to be mainly type 6.

If the supplier is using good clean carpet croppings and using a good quality glass, it should be fairly consistent.

Big variations in nylon mouldings are normally to do with moulding wet granules or NOT CONDITIONING before testing.

I also agree with Pud re DuPont service, if they identify you as a key or strategic account the service will be absolutely first grade, however if you are not on that list, they won't even return calls. Your best bet is to arrange throug your moulder to approach the rep for the DuPont material. If they can't help, find a supplier who can. There are lots of suppliers of nylons around ranging from the big multinationals like DuPont and BASF to local specialist compounders.

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

 

[Pud]

Oops! Forgot BASF - they are quite good too!

H

http://www.tynevalleyplastics.co.uk

 

[unclesyd]

The majority of the nylon carpet in the US is nylon 66. Until recently the our site made in the neighborhood of 800,000,000 lbs/yr of nylon 66 carpet yarn and 200,000 lbs/yr of nylon 66 molding resin.

 

[patprimmer]

unclesyd

I actually used Vydene as a feedstock when I was running the compounding plant a year or so ago. It was good resin. They had no agent in Aus, so we got no service, but then again we did not need it.

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

 

[unclesyd]

It is interesting you mention no agent in Aus as I mentioned this several times over the years and evidently it went nowhere. The numbers I posted above have turned around and it is 800,000,000 Vydyne and 200,000,000 carpet yarn. Most of the Vydyne is/was going to China.

 

[patprimmer]

unclesyd

It sounds like you have a good connection with Solutia.

If you don't mind I will contact you via a private forum.

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

 

[PLCRookie]

As you can tell, I am not a polymer guy, and I really appreciate the help form you experts.

The welding supplier was not much help. I attended one of their training sessions where I met many other people with the exact same problem. The welding suppliers response to polymer questions was "try plastics web".

Is there any testing that can be done on the leaking parts that would give me any insight as to the cause of the leak?
I know that it is a void in the weld, but other than that, i am stuck for the cause.

I will try to contact other suppliers to get opinions for other materials.

Again, much thanks for the help.

 

[mcgyvr]

Microscopic/microtome inspection will yield the best insight into the problems at the joint.

 

[Demon3]

As well as the excellent advices from my asteemed colleagues on the materials side. I wanted to mention that of course you must make sure there are no oils or silicones anywhere near where you do the welding. Those will ruin your joint completely.

I don't just mean obvious sources. For example, if the cleaning lady sprays furniture polish you'll get silicone everywhere. Deodordants and hair conditioner are other sources.


Chris DeArmitt