Statistical Analysis Help



Hello Forum,

I am currently in an upper level engineering course at a university and am seeking some advice regarding my quality project. For the project my group has worked with a local company which produces and sells woodworking jigs and drill bits.

Our project is focused on their most popular drill bit which has received a large amount of customer complaints, primarily due to the pilot tip of the bit breaking. We are pretty sure that the driving factor of failure is due to the tips being too long. We conducted a gauge r&r (2 operators, 10 parts, 2 measurements per operator per part) initially to prove that we could accurately measure our bits. We then measured and tested a batch of 100 drill bits. 50 cycles per bit, unless a bit broke before that. Our tests were highly repeatable (same force, same torque and rpm's, and same wood to drill into). We broke 7 out of the 100.

Our company contact is interested in fallout rates and general confidence intervals. We have plans to test a batch of bits with shorter tips to hopefully help prove our hypothesis.

My question is there anything, any statistical analysis, I could do with the 7 out of 100 data? I'm feel like there is something interesting I could show or model, because 7% failure rate is very high. I just am not sure what to do. I am not a statistical wizard and any advice would be greatly appreciated.

Thanks in advance