Vibration transmissiblity through V-belts 1

[olympic]

I want to determine the critical frequencies of a driveshaft, to check the design for shortcomings with regards to shaft whir etc. I have several sheaves along the length of the driveshaft that are taking power ( more accurate torque) from the system. I want to simplify the problem and analyze each component separately. Is this being a little optimistic? The key issue is whether potential vibrations in a drivehshaft can be transmitted through V-belts at a significant level to warrant analyzing the whole system, instead of simpler parts of the system. I have never studied the transmissibility of vibrations of v-belts, but have noticed that they can resonate at critical rotational speeds.

Thanks

 

[ivymike]

are the belts significantly more flexible than the rest of the system to be studied (do they isolate the system)?
If not,
- are the driven components of significant inertia relative to the rest of the system?
- are the belt torques time-varying to a significant degree?

 

[zekeman]

You have quite a problem in nonlinear vibration analysis, the stuff of doctoral theses (or would any candidate be crazy enough to tackle this one).
I would check the journals and transactions of the ASME and equivalent societies for this.

If this is a practical problem, then is testing an option?

 

[ivymike]

you make it sound so difficult - like you're expecting him to solve by hand for eigenvectors or something - there are analytical tools out there to make quick work of non-linear systems (especially for torsional vibrations).

I'd call up the belt supplier and see if you can get a range of approximate belt stiffness figures for the range of tensions that you're dealing with. Then I'd relate the belt stiffnesses to torsional stiffnesses at each pulley, and compare to the rest of the system. If they're much stiffer, then apply excitations & inertias directly to the pulleys. If they're much more flexible, then exclude the driven components and belts. If they're in the same neighborhood, then you're probably stuck analysing them.

 

[CESSNA1]

OLYMPIC: Interesting problem. As I see it you have a long shaft supported by bearings and driven by one prime mover. Along the shaft, and presumably between the bearings are pulleys with a belt to drive some other device. The belts and pulleys can be represented by masses on the shaft. The issues here are torsional vibration and shaft whirl. Tosional oscillations will be transmitted by the belts to the driven machinery as an alternating load along the belt. Shaft whirl will be more distinct and will result in alternalte loosening and tightening of the belts. Neither situation is desirable and can result in rapid failure particularly if it occurs at resonance. Do not forget the forcing function of the prime mover and proper shaft alignment.

Good Luck
Regards
Dave

 

[plawrence]

Olympic,

I don't know how big a system you're talking about, but for the most part, yes, vibration can be transferred from the belts and by the belts. I've seen many belts that are overtight and end up transmitting running speeds from driven to driver, or vice versa. Then you have belt frequency and its harmonics (depending on the number of belts in the system) whose amplitudes depend on sheave misalignment and belt/sheave wear.

But I applaud your efforts--my current nemesis is a belt-driven fan on a VFD that seems to have a critical speed that happens to correspond with 2X belt frequency. When the speed is right and belt conditions just so, it likes to tear itself apart.

Patrick

 

[amorrison]

Some sites obtained from googling “ belt vibration resonance” and similar words.

http://www.vibanalysis.co.uk/vibanalysis/belts/belts.html

http://www.me.cmu.edu/faculty1/wickert/DynamicSystemsLaboratory/

References_files/Nonlinear%20belts%20JSV.pdf


Interesting reads.

http://www.prime-mover.org/Engines/Torsional/contact1/contact1.html

http://www.prime-mover.org/Engines/Torsional/sport_av92/sport_av92.html

 

[Tmoose]

Every V-belt i've ever met has had thick and thin spots. Wrapped belts are worse than belts with cut faces. That translates into greater tension for the instant the fat spot is within the sheave groove, for a 2X/belt-revolution taffy tug. As unstiff as belts are, they are usually far stiffer than the support structure of many types of machines. The result is highly directional radial vibration at the driving and driven supports.

The "best" belt drive uses tiny belts at a large diameter.
The most economical belt drive is small sheaves and moose belts.

As far as transmitting vibration from other sources, their relative flexibility, and ability to slip a little bit under peaks of high load make me believe torsional stuff woul have a hard timebeing passed along.

 

[barbless]

Olympic,

Is this purely an academic problem, or is it real? if it is the latter, then I would probably fire the designer. No load transmitting sheave should be located so far from a bearing or restraint that shaft whirl could even be a matter for consideration. If it the former, then you must remember that vibrations can be transmitted to the shaft as well as from the shaft so a complete system evaluation is required. I suggest a computer program for nonlinear vibration analysis or finite element analysis such as those recommended in the text book "Vibration Problems in Engineering" by Weaver, Timoshenko and Young. As far as the belts ability to transmit vibrations is concerned, believe me; they can (with some allowances made for damping). I have Personally seen the flat surfaces of ductwork disintegrate due to resonance transmitted thru airflow.

Good Luck

 

[barbless]

Olympic,

Re: the above reply. The last sentence should state: I have personally seen the flat surfaces of ductwork disintegrate due to resonance caused by vibration transmitted thru airflow.

 

[thorngulch]

olympic:

I have seen a case where it appears that when a pump was operated near shut-off head, the load variance at the impeller created vibrations that were transmitted to the drive end which was coupled by a long v-belt drive to the motor. Under normal operation, the pump operated with no problem but under shut-off it destroyed the drive end of the pump. This was my analysis of the problem as the destruction occurred and no one was considering the severe load variations at shut-off head, but that was when the pump was destroyed. Belt tension etc. is hard to factor into the solution of the problem and prevention is better than analysis. I recommended a bypass valve to keep the pump away from shut-off head. I would recommend you do an analysis of how to protect the system rather than worry about analysis since v-velt drives are hard to model.

 

[APH]

How about belt tensioner to act as a dampener?