Go Down

Topic: Driveshaft vibration measurement (Read 116 times) previous topic - next topic

dtbingle

I'm looking to measure driveshaft vibration.  Typically vibration measurements consist of some frontend datalogger with a cable to an accelerometer, however a rotating driveshaft poses the problem of needing a wireless communication system or internal logging system (or a really, REALLY long cable - only kidding).

To maximize sampling rate and avoid latency of WiFi/Bluetooth, my thought is to use an arduino mini pro, sd card logger, RF receiver/transmitter, an accelerometer, and some sort of small battery.  The arduino mini pro would collect data from the accelerometer via SPI or I2C and log time + acceleration data to the sd card.  The RF receiver/transmitter would be used for a start/stop button for data collection. 

Each data collection "run" would be roughly 20-30 seconds long and the arduino + misc equipment would be spinning at 200 RPM and slowly ramping up to 4400 RPM.

How feasible is this setup?  Any suggestions or other products to better accomplish this?  Would the arduino and sd card boards survive long enough to collect 3 runs?

Paul_KD7HB

And what would you learn? That attaching all the stuff you mention will make the shaft out of balance! If you are serious, the measurements ALL have to be made indirectly. And remember the shaft vibrates in ALL planes.

Paul

wvmarle

Indeed - driveshaft balance and forces on the electronics are both going to be an issue.

How about attaching a light weight bearing to the shaft that is not attached firmly to a chassis or so, but in a way that it can not rotate itself but otherwise move freely in one or more axis. Then this bearing barely affects the actual vibration of the shaft, while picking it up, and then you can attach an accelerometer or whatever to this bearing to measure the vibrations.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

liuzengqiang

I was concerned with the acceleration that a rotating shaft produces at a small radius. Assume the shaft radius is 5cm, you need an accelerometer that can measure about 1000g of acceleration. Besides the balancing issue, you might find it hard to keep the logger in place instead of being spun off. Each gram of mass would require 10N of force to hold.

Can you elaborate what you intend to learn? Any optical methods instead?
Serial LCD keypad panel,phi_prompt user interface library,SDI-12 USB Adapter

dtbingle

Specifically looking to measure peak vibration levels for the driveshaft's 2nd bending mode.  If you visualize sticking an accelerometer onto the driveshaft, it would need to measure the Z axis acceleration component, not rotational acceleration or front-to-back acceleration (for lack of a technical term).  This g level generated along the z axis shouldn't be anywhere near 1000g (predicted under 20g).

The bulk of the equipment (arduino pro mini, rf receiver, sd card, battery = 30-40 grams max) could be attached at the driveshaft yoke with a wire running to the accelerometer attached at roughly 1/4 length position of the driveshaft, which CAE shows to be the most active for 2nd mode bending.  With the 30-40 grams attached to the yoke and maybe 1-2 grams for the accelerometer actually attached ON the driveshaft, it shouldn't be too difficult to rebalance and then we would be able to measure the peak of the 2nd bending mode.

The problem with optical methods is where do you mount it on a running vehicle to measure this rotating z-axis?  For example, the 2nd bending mode might be vibrating up and down in relation to the ground, but as it rotates, the 2nd bending mode might be vibrating side to side in relation to the floor (or any diagonal in between).  The only place would be on something rotating along with the driveshaft, which really only leaves the yoke.  This essentially puts us back at a setup similar to the arduino/accelerometer.

liuzengqiang

OK, help me here. So you don't want to measure the centripetal acceleration that is around 1000g but you want to measure acceleration along the drive shaft, which is 20g max?

You will need to align your accelerometer very well to avoid measuring an acceleration across from the axis you want. If you miss-align by 1 degree, 1000g*sin(1)~= 17g, your measurement may just max out with the miss-alignment. Plus, each accelerometer may have a max overall g rating.
Serial LCD keypad panel,phi_prompt user interface library,SDI-12 USB Adapter

wvmarle

Use multiple such optical instruments.

For lateral movement one above and one to the side, measuring the distance to the axle.

For longitudinal movement attach a flange to the shaft (so it's still perfectly balanced), then point the third sensor to face that flange and you have the third axis.
Quality of answers is related to the quality of questions. Good questions will get good answers. Useless answers are a sign of a poor question.

Go Up