Rolls-Royce Archives
         « Prev  Box Series  Next »        

From the Rolls-Royce experimental archive: a quarter of a million communications from Rolls-Royce, 1906 to 1960's. Documents from the Sir Henry Royce Memorial Foundation (SHRMF).
Crankshaft design, torsional vibration analysis, and testing methods for different car engines.

Identifier  ExFiles\Box 63\4\  scan0074
Date  3rd September 1929 guessed
  
-2-

The shaft E.78456 for instance weighs 73.75 lbs. - 2 lbs. less than E.75216 - yet is 14% stiffer and has larger journal bearings (more area); it has wider and shorter webs. The energy stored per lb.wt. is increased for a given twist. The Chrysler shaft is somewhat similar in shape, but has solid pins and journals. It may be that we shall want to modify the constants in you formula slightly when we have more experience on these car shafts.

On E.78456 we were at first inclined to blame the large chamfers on the webs for the lossof stiffness in comparison with you formula. We find however, that the stiffness when chamfered, only drops from 52,700 to 52,000 lbs. ft./radian.

The shaft E.79351 weighs 90.875 lbs. and is an attempt to cure engine roughness generally at high speeds by a heavier shaft. We know that E.78456 already puts our critical speed up high enough.

I note your objection to the "simple" method by which we calculate car torsional periods. I think the method is justified theoretically by the fact that the equivalent inertia of a of a plain parallel rod is 1/3 of its total inertia concentrated at the nose.

There cannot be much difference mathematically between a system of flywheels along a rod and the same amount of inertia evenly distributed (stiffness the same in both cases). In the cases of two-node vibration, as occur in aero engines, there is perhaps less theoretical justification for this method.

With regard to bench tests to determine critical speeds, we always find these are borneout on the road. We think that our car flywheels are so heavy that the interference due to the dynamometer and its shaft is negligible. There is always at least one large fabric coupling in the shaft to the dynamometer. We prefer to determine the critical speeds on the bench, since the r.p.m. can be determined more precisely. We usually go by the 1/2 speed period (6 per rev.) as it is so sharp.

We find a marked difference between the severity of the periods on different cars. A well-worn engine is always worse than one with tight bearings, but different makes of cars differ still more. We tried a Chrysler car recently with its crankshaft damper removed, when we could barely pick out the 1/2 speed torsional period at all (this period is usually the worst in the normal running range). Whereas on our cars all the lower periods are very noisy. Infact we have quite a lot of trouble to get our dampers to work adequately and consistently, to suppress them. We should be interested to know if you have observed this, and know of an explanation.
  
  


Copyright Sustain 2025, All Rights Reserved.    whatever is rightly done, however humble, is noble
An unhandled error has occurred. Reload 🗙