An injection-molded part with a pressed-in brass insert was found to crack during testing. There were multiple cracks, all originating from the insert. The cracked part was brought into my lab late one afternoon. I took some photographs of the cracks, then went home for the day. The next morning, I looked at the part again. Comparison with the photos I had taken the previous day confirmed that the cracks had grown overnight.
The presence of multiple cracks, and the fact that the cracks continued to grow even after the part was taken out of service, pointed toward environmental stress cracking. This is a common failure mode for plastics, in which cracks form at relatively low levels of stress as a result of chemical exposure. The stress may be much lower than the strength of the material. Most plastic parts have some level of molded-in residual stress, so cracking can occur even when there is no external stress -- if the part is exposed to the wrong chemical.
Some of the chemicals which can cause environmental stress cracking include fuel, oil, grease, solvents, adhesives, and cleaning products. It can often be difficult to find environmental stress cracking data for a given plastic with a given fluid. This is particularly true these days, since many resin manufacturers have closed or scaled back their laboratory facilities. You may need to do your own testing to make sure that the plastic you plan to use will not crack when exposed to any of the fluids it is likely to come into contact with.
I found some oil residue on the cracked part. Using Fourier transform infrared spectroscopy (FTIR), I was able to identify the specific oil type.
In order to test the environmental stress cracking resistance of the plastic to this oil, I drilled three holes in a sample of plastic, and inserted pins into the holes. The pins were oversized with respect to the holes. The first pin was oversized by 1 percent, the second pin was oversized by 5 percent, and the third pin was oversized by 10 percent. The interference fits between the pins, and the holes gave me three different stress levels. I then immersed the sample in oil and checked for cracks every 24 hours.
My testing showed that cracks formed in a short period of time, even at the lowest stress level. This indicated that the plastic was extremely sensitive to this oil, which explained why the failed part had continued to crack overnight in my lab.