ABS - injection moulded part - internal crack - part failure - wall thickness difference.

[addenroy]

Imagine a 100mm long part which is wider at one end and tapers down from 30mm to 15mm dia over a length of 30mm and then is effectively straight for the remainder of the length - see attached sketch. The wide end is hollow and has a constant wall thickness (3mm) following the outside taper for 30mm and then it stops and the wall thickness jumps to around 9mm (i.e. a short cone shaped hollow). Beyond this, the wall thickness is fairly constant for around 70mm (which is in fact the handle of the part).
The part is gated at the open cone end and flows towards the thicker handle end, so the melt does flow around the core forming the cone shaped hollow and forms a small weld line as the flow fronts meet at the tip of the cone.
The wall thickness differences and gating into a narrower wall cannot be changed unfortunately - we know this design goes against guidelines, but we have to go this way.
We are finding that a small amount of parts are cracking near the tip of the cone shaped hollow, where the wall thickness goes from 3mm to 9mm. The break is cup shaped and shiny, apart from outer surface, approx. 1mm of skin. The break is actually through the 3mm wall thickness, approx 1mm from the end of the inner cone shaped hollow - see attached sketch.
These breakages are not occurring after moulding, also other items are assembled onto these parts and then they are packaged, all without breakages. It is only when they arrive at our distributors that some are found to be either broken, or else a white line is visible on the surface (most of these are moulded in dark colours or black) and it just takes a light knock to break the part.
We acknowledge differential shrinkage is occurring at the junction of the thin to thick wall and probably creating built in stresses. The cracking however does not happen in the majority of the parts (we've cut a few and they look fine). The parts are packaged and boxed adequately.
Sorry for the ramble, hope the above isn't too exhausting. The parts are made from Novodur P2HAT ABS. Does anyone know what could be causing this type of crack and if there are any ways to resolve (e.g. are there any other ABS grades less likely to crack)?
Thanks in advance for any assistance.
Andrew.

 

[Demon3]

My best guess is that there is a lot of stress locked into that section due to the way the flow occurs and then the part cools locking in that stress. You may be able to get better results by increasing the mold temperature for better flow and holding it longer in the mold to cool to give less locked in stress. Also try injecting slower if possible. What you want is a low viscosity and to allow the part to relax somewhat before it is cooled and the stresses become locked in. Along the same lines you could look for a higher MFI ABS.

If you wanted to actually see the stress pattern you could mold some parts out of transparent ABS (Terlux) and then look at the part under polarized light. The stress will show up as colored bands.

Chris DeArmitt PhD FRSC
President

Plastic materials consultant to the Fortune 100

http://www.phantomplastics.com

 

[addenroy]

Chris,

Thanks for the reply and advice - I've passed on your suggestions to our team. Aesthetics is a big thing, so anything that impacts a high gloss finish is an issue. We've used the clear ABS with polarised light for other projects - thanks for reminding me about it.

FYI our process settings to get nice shiny parts are:-
Mold temp:- 70C
Hot feed temp:- 240C
Injection time:- 3.3sec (4 cavity, total shot weight approx 130g)
Cooling time 15s.

Thanks again,

Andrew.

 

[Compositepro]

Your quest for good surface finish might be contributing to the problem. The stress is not created during injection but during cooling due to shrinkage. Freezing the part surface quickly will give a good surface finish but it prevents the surface from moving as the molten material in the center of the thick areas is cooling and shrinking, so stress is created. You could try annealing the parts after injection, although the stress relief might show in the part surface smoothness. Injecting some gas bubbles at the tip of your core, at the end of injection, might also work, or just keeping the core at a higher temperature so shrinkage occurs on the inside of your part rather than the outside surface.

 

[addenroy]

Compositepro,

Thanks for the feedback. I think you've made a good point regarding keeping the core at a higher temperature. This core is cooler at the start of the run, but then heats up over a number of cycles - this could explain why only a few parts are failing. We maybe need to reject more cycles at startup than we are doing at present. We'll investigate.

Thanks and regards,
Andrew.