Crane Motor Vibration

[compuagent]

Would appreciate some advice and help with the following situation: A vertical lift drive motor was changed on an old 3-axis automatic part storage and retrieval crane from a single speed motor to a variable frequency motor to ramp loads to travel speed. The crane raises a platform between 2 masts with a 1.75-inch pitch 140-2 roller chain. The platform is gravity lowered against the drive. The 40 hp 3 phase 1800 rpm drive motor is configured vertical drive end down mounted to a 75:1 ratio 2-stage worm reducer input shaft by a bell housing and an elastomer jaw coupling. The VF ramps the platform to travel speed of 68 fpm by roller chain over an upper and lower sprocket set, a tensioning sprocket, and finally the horizontal 24-tooth driving sprocket turning at full speed of 23.3 rpm. The motor weighs 520 lb. The rigid mounted reducer at the bottom weighs 1920 lb. WK^2 on the load side referred to the reducer high-speed input is 20 lb-in^2. WK^2 of the reducer is 31 lb-in^2 at the input, plus motor rotor inertia of 287 lb-in^2 is 318 lb-in^2. When the platform is raised the chain and motor visibly vibrate to a maximum at the middle of travel. The motor natural frequency was found to be 1125 cpm, and a harmonic of 560 cpm appears to be shaft rpm 23.3 times 24 teeth on the sprocket. Will speed change with a different sprocket change the vibration, and if so what can be expected?

 

[otaio]

Compuagent

First let me complement you for providing such a full technical description of your problem. A refreshing change which others using this and other forums should aspire to copy.

Real life problems such as these are challenging. While it is unlikely that anyone can dictate a solution to your problem it may be possible to give you some pointers for you to explore.

But first a number of questions.

1 What was the speed of the original motor?

2 Was this motor originally connected to the system through the piece of plant you refer to as ‘rigid mounted reducer’ (ie is it in essence a speed reduction gear box and if so what was the overall ratio)?

3 Did the orignal system suffer similar problems?

4 What is the centre line distance between the upper and lower drive socket sets?

5 Does any part of the vertical lift and related assembly and drives have vibration isolators fitted and if so where?

6 Regarding your reference to ‘natural motor frequency’. How was this measured?

7 Regarding your reference to ‘harmonic’. How was this measured?

8 Is the current travel speed of 68 fpm faster, slower or equal to the original travel speed?

Look forward to hearing from you.
Regards

Bob B

 

[compuagent]

Hi, Otaio. In reply to your questions. 1)the original motor was 1800 rpm. 2)the reducer is a 2-stage cone drive worm reducer with double output shaft, 75:1 ratio. The reducer is mounted to the lower frame of the crane without isolation. 3) I don't know the history of this crane, except it was in service for 12 years with a synchronous motor, before we put a vector drive motor on. 4) upper and lower sprockets are 57 feet apart. Using cartesian location, put a circle representing the drive sprocket at point (0,0); the tensioning sprocket will be at (1,0); the bottom sprocket will be at (2,-2); the top sprocket will be at (2,57). Draw lines tangent to the circles representing chain. The platform is hooked to the chain between the top and bottom sprockets. 5)Vibration isolation is not used at any component or subsystem. 6) Natural motor frequency of 1125 cpm was found using a frequence response function from an impact test with the motor assembled on the input adapter. 7)frequencies of 1125 cpm and 560 cpm were detected with an IRD Entek Datapack 1500 spectrum analyzer. 8)at 72 Hz we are running a little faster than the old design. Drive components were not changed. Also, the old motor frame was heaver than the lighter vector motor frame. The vector motor frame is a square fabrication. The motor has a shaft encoder and a brake on the back end. I would like to add weight to the motor by extending the frame bolts and adding weight in the form of plates above the brake to alter the motor natural frequency.

 

[butelja]

If the vibration problem is centered on 1 particular frequency (560 cpm), have you considered adding a dumbell type tuned vibration absorber?

Also, since the motor's natural frequency is higher than the excitation frequency, it seems that adding mass would be forcing the two frequencies closer together which would not be a good thing. It seems that you should be stiffening rather than adding mass.

 

[electricpete]

I agree that more details are better. Unfortunately, I got lost in these details. I guess it means I probably couldn't have helped anyway.

 

[compuagent]

Butelgia, thanks for the reply. Can you tell me more about the dumbell type tuned vibration absorber. i.e., description and how it is tuned. regards.

 

[otaio]

Compuagent

Thank you for supplying the additional information which I found very informative. The use of cartesian co-ordinates was inspired. (In respect of the idler does the chain run to the left of a vertical line through the idler axis or to the right)?

I note that you report that the ‘ old motor frame was much heavier than the lighter vector motor frame’. From this I concur that the old motor frame was also probably much stiffer and it is probably in this area that you should concentrate. (In other words it is the flexure of the drive train which is causing the problem).

I agree with Butelja that adding mass to the motor and hence lowering the natural frequency of the motor is probably not the approach to adopt. Likewise I would suggest neither is the tuned vibration absorber. However as a point of information I would suggest you view.

http://www.gmi.edu/~drussell/Demos/absorber/DynamicAbsorber.html

The natural frequency of the chain will be influenced by the chain’s tension – initially set by the idler wheel – the self weight of the chain, the load that carries , and the distance between the top and bottom sprockets. The displacement of the chain from the vertical will also be influenced by the interaction between the sprocket and chain. It is not possible to calculate to calculate the natural frequency of the chain in this real life application. It would need to be measured and given that information you could then look at the various exciting frequencies in context.

From what has been written I would do the following:

1) measure the vibration of the chain probably using a stroboscope. (So I know where I started)

2) stiffen the motor and drive mount.

If that solves the problem – then great.

If not I would then see if I could fit dampers between the platform and each tower with the view of limiting lateral movement.

Bob B