Engine Health - Relative Cylinder Performance - using a 3-axis accelerometer

Ask any questions about using PicoScope 6 Automotive software here.
Post Reply
martinr
TwoWaves
TwoWaves
Posts: 53
Joined: Tue Mar 06, 2018 4:06 pm

Engine Health - Relative Cylinder Performance - using a 3-axis accelerometer

Post by martinr »

I can't decidewhether to post here or in the diagnostics sub-section or even in the NVH section, but I want to invite comments and feedback on a test I played around with last summer.

The test centres around using the 3-axis accelerometer (TA143) (and interface TA259) from the NVH kit to display the accelerations and decelerations of the crankshaft throughout the cycle, and therefore to show if any cylinder is underperforming. Actually, rather than measure the accelerationn of the crankshaft, which would not be practicable, the acceleration of the engine block is monitored instead. "Action and reaction are equal and opposite." The crankshaft rotates clockwise looking at the pulley, so the reaction of the block is to rotate anti-clockwise but is resisted by the engine mountings. So, the accelerometer was placed on a cylinder-head bolt pointing fore and aft with a positive acceleration displaying when the acceleration of the block was towards the bulkhead ie anticlockwise. The screenshot below is taken from a Rover 100 (K Series) with a transverse engine.

Looking for additional relevant information to display, Channel A shows the crank position sensor; Channel C displays the intake-manifold depression (using an intake-manifold vacuum sensor from a Renault), and Channel D shows exhaust pressure as detected by a GM fuel-tank pressure sensor (an idea taken from one of ScannerDanner's videos). And Channel B is, of course, the accelerometer output, which has been "cleaned" a little using a 1kHz low-pass filter.

Knowing the position of TDC on No 1 cylinder on the crank-position-sensor waveform allows rulers to be used to designate the 720 degree cycle. Alternatively, a simple secondary-ignition-pickup waveform from No 1 cylinder can be used as a sync.

engine health - relative cylinder health.jpg

As with the relative compression test, the absolute value of the accelerations measured are not important: the test aims to see if there are any consistent anomalies, and, if there are, whether they correspond to any anomalies in the exhaust pressure waveform or the intake-manifold-vacuum waveform.

No way could this setup compete with the elegant simplicity of the relative compression test; however, the relative compression test is done whilst cranking, and this test is done under load whilst driving. If there was, for example, intermittent or permanent reduced power that scan data failed to log, or an occaisionally-sticking valve or hydraulic lifter, perhaps this test using the accelerometer might be of use. I'm not suggesting that such a test makes it worthwhile to buy the NVK kit; however, if you already have the kit, here's a use for the accelerometer that wasn't envisaged when the kit was bought.

Anyway, I'm keen to know people's thoughts; is this just a theoretical exercise with little practical use or might there be some merit in it?

Post Reply