Tension bolt failure is a fairly well known issue in racing. We also
encountered it on the large engines at Deere / RPI (5.8 liter per rotor). The
bolts have a vibratory mode that is excited at certain RPMs. The failure is
by fatigue. The solution is simply to use some kind of a snubber on the bolt
to reduce the amplitude of the vibration, and hence the maximum stress. The
RTV on the shank is a simple way of creating a snubber right at assembly. It
really hasn't anything to do with damping. At RPI, we used to put small
cross-section o rings on the shanks and use heat shrink tubing to hold them
in place. It is less messy. The reason that the failure mode doesn't show up
in most cars is that they do not spend sufficient time at the resonant RPM
(which I would "guess" in a Mazda 13B is close to 6000 RPM).

I think that better bolts are a good thing, but they won't by themselves
affect the root cause. It would be prudent to use snubbers even on the better
bolts that Paul is working on.

Furthermore, I don't believe the excitation is a function of engine balance.
I believe (with no facts to back me up!) that the excitation is from the
combustion event. No amount of balancing can eliminate the structure borne
vibration from the firing impulses. These combustion forces are of course
fully reacted within the engine structure and are not transmitted externally,
but there may be significant high frequency energy transmitted within the
engine components due to the combustion event.

Jon Lauter
Rotary Power Marine Corporation

 
The AirCraft Rotary Engine NewsLetter.              Powered by Linux.

http://home.earthlink.net/~rotaryeng/            http://www.linux.org