Half Critical Excitation of Deadshaft Idler

[kamenges]

I recently had a customer contact me about a tension control issue on his machine, a web processing machine wiith loadcell based tension control. He said the machine ran fine except for a small speed range during accel and decel. I had seen this before and told him the machine was passing through a speed that caused the loadcell roll to rotate at the roll shaft natural frequency.

I went to his site and performed a test to excite the shaft and measure the natural frequency. I found the shaft natural frequency was exactly twice what I expected it to be, which caught me by surprise.

I found multiple references to excitation at half critical speed. However, the explanation provided by Roisum and Hartog seem to require a liveshaft roll or rotor of some type. They require either variable shaft springrate or variable shaft tangential loading due to imbalance, neither of which apply to deadshaft idlers.

i looked at this thread which was related (thread384-103687 ). But I can't find any explanation that directly applies to my case. Can I just assume there is a fairly imperceptable impact load that is causing the vibration or is there an explanation for half critical excitation that applies to deadshaft idlers?

Thanks,
Keith

 

[GregLocock]

Half speed excitation can also be due to bending of the shaft I think. I thought it only applies to assymetric shafts.

Cheers

Greg Locock

SIGlease see FAQ731-376 for tips on how to make the best use of Eng-Tips.

 

[kamenges]

Thanks for the reply, Greg.

I have seen descriptions of how shaft assymetry can lead to half critical excitation. But all those explanations were for live shaft applications. In my application the shaft is stationary and the roll body rotates around it. If the assymerty isn't rotating I'm not sure it can induce oscillations.

Keith

 

[Tmoose]

"If the assymerty isn't rotating I'm not sure it can induce oscillations."

Once upon a time I got to dynamic balancing various commercial process rolls for a living. There were many struggles with long rolls with uneven wall thickness. The importance of consistent wall thickness for balance-ability and good dynamic behaviour was lost on many roll manufacturers, especially shops just "getting into rolls."

Usually the wall thickness variations I saw were simple eccentricity, but I think a particulary evil roll would take delight delivering a whopping dose of ovality. A typical balance machine measuring in the horizontal direction would miss all the fun. A shudeer just went down my spine.

 

[kamenges]

Thanks for the reply, Tmoose.

After reading your post I realized I was being myopic. If the roll body itself had a defect causing the spring rate to change it would create forces at the required frequency. The roll body in question is steel so there is significant enough weight there that a change in spring rate may create enough energy to cause a problem.

Keith