There is actually a number of products that do just this type of vibration
analysis. It's used alot in manufacturing plants in various "predictive
maintenence" packages to try to determine how close a
pump/motor/turbine/etc. is to component failure and thereby cause production
downtime. But prices run in the high hundreds to the thousands as you not
only need the pick-up/microphone but also an interface card that can measure
and store the audio data in high speed/high resolution (the standard I/O
sample rate of Windows 9x/NT/2k is only 20 ms) and a software graphing
package to display the results in a meaningfull manner. Then, of course, you
need to go though enough engines to be able to understand what you're
looking at/for in the data for that type of engine. For instance, does that
new bulge 1/3 down the chord mean the water pump is wearing normally or the
crank's main bearing is starting to go? And how bad is it? The more
expensive ones can use fuzzy logic or some form of AI to help interpret the
data.
Richard English
68 Cadillac C de V conv.
70 GT6 conv.
70 Intermeccanica Italia
71 Spitfire
74 Spitfire
78 Mercedes 450 SL
94 Toyota Land Cruiser
-----Original Message-----
From: owner-triumphs@autox.team.net
[mailto:owner-triumphs@autox.team.net]On Behalf Of derek evans
Sent: Friday, October 13, 2000 3:26 AM
To: triumphs@autox.team.net
Subject: crank resonance
>You can test
>the natural frequency of the shaft by hanging from a chain or such and
>"pinging" it with a light hammer, and measure the noise frequencies.
in my opinion this test would provide no useful information about what's
going on inside the engine.
for a start the first and last main bearings 'terminate' the ends of the
crank and stop it from moving freely and so create a different standing wave
pattern when the crank is 'pinged'. the other main bearings in effect
divide the crank into a number of shorter cranks each with the ends
terminated, their own standing wave pattern and presumably a higher resonant
frequency.
the dampening effect of the big ends (imagine putting a finger on a
vibrating guitar string) would also vary the free air resonance of the
crank.
and don't forget all this is dynamic; subject to twisting, and bending.
a better way to see what frequencies are produced would be to run the engine
with some sort of microphonic device attached (piezo sensor perhaps) and
look at the output with an audio spectrum analyser.
might sound far fetched; but not beyond a basic home computer and some
software to do this.
derek [1960 herald, fan on water pump :) ]
_________________________________________________________
|