I was working on a product with a gearbox. A two-inch-long bolt in the gearbox broke and jammed things up when the bolt head engaged with a lightening hole in one of the gears. The bolt broke with a very flat fracture at the first loaded thread, no shear lip, no beach waves, so I thought it was due to high-cycle fatigue.
I asked around and discovered that the same bolt had failed years before a few times. They solved the problem by increasing the installation torque. That kept the problem from happening again for a number of years. I gathered the reports written by the engineers who worked on the problem previously.
I calculated the first and second bending mode frequencies for the bolt and found the second mode frequency was pretty close to that of an adjacent gear mesh frequency. I was able to identify fretting on some bolts where the second bending mode had the iron-based bolt hitting the light alloy case at the antinodes. The previous study of the problem had included running the bolt with strain gages, and I was able to retrieve the magnetic tape data.
The previous study had reported on the normal operating stresses, which were deemed acceptable and matched that of a main power transmission gear mesh frequency. The gear I suspected was an accessory gear adjacent to the bolt failure locations. I took the magnetic tape data, and instead of looking at the steady state data, asked that the data be reduced into waterfall plots (amplitude vs. frequency vs. time) starting right from start-up.
Those plots revealed that as the alloy gearbox case material warmed up, it brought the second bending mode frequency of the bolt into a perfect match with the accessory gear mesh frequency, so there was a relatively brief period during warm-up where the strain was very high. Fatigue cycles add up pretty fast when the frequency is 20+ kHz.
I recommended removing the Iron-based full ground shank diameter bolt with machined threads, and replacing it with a Nickel Superalloy bolt with a pitch diameter shank and rolled threads. This gave a huge shift in the bolt frequency away from the excitation, a huge increase in strength and toughness, a huge increase in the clamping strain energy, and a weight decrease. Problem solved in fairly short order. Sometimes less is better.
This entry was submitted by Oscar Carlson and edited by Rob Spiegel.
Oscar Carlson is a retired mechanical engineer.
Tell us your experience in solving a knotty engineering problem. Send stories to Rob Spiegel for Sherlock Ohms.