We had a cardan shaft locking up recently and I wanted to share the data. This was the first cardan shaft that I have experienced problems with.
While taking route data I noticed a slight vibration in our press section, the rolls in this area run very slow (40-80 rpm)and any vibration is a concern. One of the press rolls had a noticable axial movement. After taking data and reviewing it we used IR to verify problem.
The data had been showing a slight 4x peak and then a dominant 3x at the time of locking up.
The correlated waveforms were very interesting.
I feel like I have a better idea of what to look for when checking for cardan problems now.

Cardan_shaft.pps (143 Kb, 97 downloads)

Scott,

The correlated waveform is really interesting. Great post!

Not familar with a "cardan" shaft. Interesting data, good use of multiple technologies. My experience has shown that when 3x is present it typically indicates bearing inner race loose to shaft, 4x is bearing loose in housing KEEP IN MIND (no love joy/3x style coupling/blade pass, etc.) Can be quickly verified with dial indicator for shaft run out and vertical shaft play

Scott,

Very interesting. Now I think I have a better idea of what to look for, too. Thanks for sharing.

Why do you think there was a shift from 4 x to 3 x?

Dave,

I think we are talking about a log shaft with 2 universal joints like on a rear wheel drive car or truck. There are some fairly big ones on paper machines, bigger ones on calendars, and even bigger on steel rolling mills, I think.

The biggest I cover are on a plastic calendar and are about 24"-30" in max diameter, probably 20" tube dia. and turn at about 20-40 rpm.

quote:

Cardan shaft data

Scott
Thanks for sharing.
In axial, 4x is 2ª harmonic of 2x, that normal vibration for cardans.
Question: Excuse me. Where occorred the locking?
In spline? Or universal joint?
What hot point in thermal image?
Thanks

Ricardo Góz from Brazil

Danny,

I think that all four of the bearing in the universal were binding slightly as they passed one point, say 12:00 causing the 4 times harmonic. When the one bearing finally started locking up it was going through 3 tight positions each revolution.
Just my opinion

Ricardo,

I went back through my data trying to see if there was ever a dominant 2X and could not find one, the 4X although very low was still the dominant peak. Everything that I had read stated that the cardans would show 2X.
The failure occured in one of the universal bearings.
Have a great day,
Scott

Cardan_data.doc (115 Kb, 19 downloads)