These are 3 ideas that may lead to the root cause. 3% is not an acceptable scrap rate as some parts may have only a 5% profit margin. I wish you well.
- I believe the dye is added and I am finding out exactly how. No regrind is permitted. The carrier material and colorant is specifically intended for LDPE, but I am confirming this.
- LDPE was chosen for its flexibility because it is bent/stressed in its final use. There is no way around this since we cannot mold it in a curved shape. It is molded flat before being slid over a mating part that is arced/bent.
- Wouldn't it have less internal stress if cooled slowly and more if it is cooled fast (quenched)?
You are doing the best thing by talking to the material supplier and the dye supplier. However, beware of 'salesmen'. Get to know the technical specialists. They have a lot of resources to do tests and investigations.
Another processing parameter I thought of is packing time/pressure. Are you controlling it? This may affect knit-line strength.
We tried to heat form some LDPE tubing as a secondary operation and it was very difficult to make it hold a bend. The bend relaxes at room temp over time. It also 'creeps' to hold the shape it is shipped in. You may have 'creep' in your favor after installation in a curve. Eventually it will take that set.
A bit of micro analysis would go a long way to checking some of the excellent ideas posted. I would strongly recommend that a failed part be analyzed with FTIR microscopy across the knit line. A good analyst can tell you what chemical differences might be occurring. X-ray analysis [XPS] at the failed knit line can show silicone at extremely minute levels that other analysis cannot find.
Also, if your melt dispersion of the black concentrate is poor, you may very well have extremely dissimilar molecular weights across the knit line. Typically lower MW resins are used for the carrier resin, which would have far lower crack resistance than the main LDPE material.
We recently had parts molded from LLDPE instead of LDPE. I know the strength of LLDPE is supposed to be higher, but what other differences should I be concerned with? The parts fit, flex, and feel acceptable, but a material change (although it is a similar material) is something that must be thoroughly investigated before approving. Find more on our webiste: http://www.acomold.com