[TheForge] Off-topic: ...dynamic balancing, vibration...

[xlch58 at swbell.net]

Peter Fels & Phoebe Palmer wrote:
> The shade tree approach would be to suspend the shafted assembly between 
> 2 sharp, hardened points in the centering holes on the shaft ends, in a 
> lathe. Then give it  some whirls and mark the bottom points when it 
> stops moving. Shave off metal opposite the marks and repeat till crazy 
> and the marks distribute randomly.
> But i doubt that'll get you to 10,000 RPM.
>
>   
That is statically balanced.  Dynamic is a different story.   Bruce 
hasn't really described the part, so it is hard to determine any 
solution.   All I have heard is it is a rotor on a shaft.    What is the 
size of the shaft, length of the shaft, how is it constrained 
(bearings).   What is the mass moment of inertia of the rotor?   Will 
the assembly be fixed in place, or will it move(gyroscopic forces)  What 
sort of loads will it be subjected to?    When balancing a crankshaft, 
the connecting rods are simulated with bob weights, and the balancer is 
set up with sensors that capture the info germain to a crankshaft.  With 
high speed rotors on shafting, whip is a big issue, so the shaft length 
and bearing arrangement are key.  At any rate, mount it in bearings that 
are pivoted with a strain gauge opposite the pivot.   Spin it up to near 
10k and take readings and decide if they are outside of acceptable 
range, then you need to look at the cause.  If it is imbalance, then you 
need a position sensor to correlate with the strain gauge.  You can 
figure out the response time required with math, think nyquist.    The 
position sensor can be integrated into the drive unit possibly.    I 
drew up some plans for a IC engine crank balancer, but I had the 
advantage of being able to use some off the shelf parts like a CPS.   I 
think you will find this endeavor is more expensive that having it 
done.    I would suggest that you spin it up and use a tach/strobe to 
verify the rpm.  Use the strobe to observe vibration (look at it at 
various angles and sweep the frequencies on the strobe), record what you 
see.    Keep in mind that it can shake like a bastard at 500 or 5000 rpm 
and it doesn't mean squat if you are going to run at a constant speed 
of  10k rpm.   There are many aircraft out there that the engine/prop 
shudders as it comes to a stop because the engine mounts are tuned to a 
frequency below idle - and many have a proscribed RPM range -- same is 
true with big ships. 

Charles