We recently had a rash of shafts developing cracks in them which result in shaft failures. These shafts have pillow block bearings, and sprockets on them. Question any predictive tools to help identify these cracks?

In additional to your routine vibration measurement, you can also the Phase measurement for each bearing locations. Any change in the Phase from 90 to 180 degree, it is indication of the crack.

I understand about phase measurments but...what i was looking for is some way to put a probe to end of shaft take reading to indicate fractures? Maybe kinda in the future..huh

IF I'm not mistaken some folks on here have had success using airborne ultrasonics to detect shaft cracks.

If you are having many cracks, there is probably something wrong with the design. Most common is a step in diameter without a fillet to reduce the stress at that junction.

Ultrasonics is an interesting option. We just got a new ultrasonics gun..would like to know how to apply this to this problem.

Hunter,

I have measured ultrasound emissions from chain drives. They are relatively noisy (ultrasound) depending on torque transients, chain slack, chain wear, sprocket alignment and lubrication. If you were to try your "gun", you would have to sheild the microphone or use a directional micropone to reduce background. You would have to accurately measure the background. If your measured shaft levels increased, then you would have to evaluate whether caused was by a cracked shaft or simply a change in background. I don't know whether it will be easy or even possible for your situation, but it is worth a try. For special situations, I have made my own focusing sensor that has excellent directivity and improved amplification. Most likily you will have to live through one or more failure events to test the method, or inspect shaft by other means once you suspect a crack is present. You might simply conclude that ultrasound monitoring helps keep the chain drive in better conditon (alignment, tension, and lubrication) and thereby reduces shaft cracking ... priceless!

Walt

Deer Hunter,
Which instrument did you purchase? I am with a class at this moment and would like to know (I am vendor-neutral) so I can pass on the application to the technicians I am training.
If you can share any "sound files" and/or any waveforms, spectral analysis, etc.., not only I but the readers of this blog would appreciate it.
If its not too much trouble also post it on the "ultrasound" Maint Forum Site.
Please keep up informed.
Regards,
Jim / Ultra-Sound Technologies

Crack initiation can be diagnosed through vibration probes. Increase in 1X and 2X vibration is one symptom. Decrease in natural frequency is another symptom. A run up or coast down readings trending may be useful. UT will not be useful when the equipment is in operation. In standstill condition, magnetic particle testing can detect cracks if the shaft is magnetic. But as pointed by others, frequent shaft failures require RCA.
Regards
Irshad

quote:

Originally posted by AKHTAR:
UT will not be useful when the equipment is in operation.
Irshad

You must be meaning airborne UT, huh.
I used to pay people to find cracks in shafts in the mining industry while the equipment was running. They used UT (Kraut-Kramer) and were very good at it. When they said a shaft was cracked, we changed it.
It didn't have to be a perpendicular crack either. They found longitudinal cracks as well in a few shafts that had that type failure predominate. Depends on the crystal you use, the equipment, and the dude doing the testing, but it can be done very well.

D

Dave,
Ough, you cut me short. I have still to come across an UT measurement to find crack in running shafts. Did you really find cracks in shafts pointed by those dudes? If possible, will you guide through the procedure ?
Regards
Irshad