Oil filled vibration damper

[EnglishMuffin]

What is the best way of designing an oil filled vibration damper ? The frequency range of concern is perhaps 30-80Hz, at low amplitude (0-.001"). Should such a damper utilize a relatively long viscous restriction or is a short orifice good enough ? According to my edition of the Vibration Handbook (Harris & Crede), there is a rapid increase in damping as a function of frequency in this type of damper. Also, does anyone know of any textbooks/papers covering this subject? I haven't found much on the web.

 

[GregLocock]

Are you talking about the dampers that just have an inertial ring on a bearing, with no spring element? They usually just rely on the clearance between the flat face of the inertia ring and the pulley to create drag.

I don't thonk you'll find much on the web, we use hydraulic engine mounts which have tuned channels for the damping, and designing them seems to be a black art.



Cheers

Greg Locock

 

[EnglishMuffin]

I guess I am designing what amounts to a dashpot - ie a piston inside a cylinder with clearance. The piston will be connected to a friction pad pressing against a sliding surface, which is independantly supported by rolling element bearings - which provide most of the stiffness. There will actually be multiple units - maybe as many as 20. The pistons are about 3" diameter. The component I am damping weighs several tons. I am wondering whether the best way of doing it is to rely on the viscous losses in the gap (classic dashpot), or whether it would be better to utilize a sealed piston with a separate orifice. In the former case, the damping should be directly proportional to velocity, whereas in the second case it would increase in proportion to the velocity (I think). On the other hand, as I mentioned above, there appears to be a marked additional non-linear frequency behavior due to the "mass effect" so perhaps it's all highly non-linear anyway. One consideratin in the case of a piston with a large seal might be that at high frequency and low amplitude, the seal itself might tend to behave as a spring and swamp the oil shear effect in the gap.
I am obviously going to have to make one and test it, but I was just wondering what direction to go in.

 

[strokersix]

Englishmuffin:

I anxiously await further discussion in this thread.

My feeling (and it's just that, nothing more) is seal friction effects may be detrimental to what you are trying to accomplish and you'll have better luck with the gap viscous loss approach.

 

[GregLocock]

Yes it's all non linear, and I agree with stroker six, seals will make things far more complex. It sounds like a mighty machine.

Our hydromounts use rubber as the spring and seal (ie no stiction problem), and then an inertia tuned spiral channel to get the damping. This is coupled to the main fluid chamber via another spring, in effect (actually a diaphragm).

The system model has upwards of 50 elements... and correlation is good enough across the frequency range 5-300 Hz.

I don't like the sound of your friction pad, because of stiction and wear.

The most foolproof solution would be a dashpot, maybe with some ancillary ports if you want to get fancy. How critical is this thing? How will you cope with variations in viscousity with temperature?





Cheers

Greg Locock

 

[EnglishMuffin]

Well, the (anti) friction pads (Rulon) are necessary (I think) because the component being damped has to move transversely relative to the pads. They would be lubricated of course. I can't see any easy way to do it without some sort of seals - I had envisaged two small stub shafts with seals on them, coming out of each side of the main (seal free) damping piston. They would have slightly different diameters, thus providing a small bias force between one of them and low friction pad against the component. Since the oil supply to the whole thing will be dead headed, the oil will be at room temperature (or at least average machine temperature), so viscosity variation should not be too extreme, I hope. Does this make sense ?

 

[GregLocock]

Sorry, I got it into my head that you were talking about torsionals. I'll just zap the memory bank and start again.

OK, first problem is that your vibrations are small, and frequencies are low, so the velocities are small. Stiction will be your biggest problem in this scenario, I suspect. I think you'll be in the 'roll back' portion of any sliding seal - ie you might as well go with a bonded elastomer joint. With such small deflections I think you'll struggle with any fancy valving unless you can get the fluid velocity up by using a piston and a small orifice.

I've never designed a dashpot - where does this frequency dependent behaviour come from? just omega*x?









Cheers

Greg Locock

 

[sreid]

Have you considered Hydrostatic oil bearings? Stiction free movement with both shear damping in the direction of movement and squish damping normal to the movement. These are generally less expensive than rolling bearings for large loads. Universally used in grinding machines.

 

[tomaspin]

Hello Gents,

Having read this thread, I am wondering whether a viscous damper is required at all.
To summarise, you have a large machine weighing several tonnes, producing low amplitude vibration in the range 30-80Hz, and you wish to damp or isolate this vibration. Freedom to move in a lateral plane is also required.

Hydraulically damped mountings tend to be tuned to produce maximum damping in the 10-30 hz range, but at higher frequencies the damping is usually little different from an undamped elastomer mounting.

If you want hydraulically damped mountings, then ContiTech, Lord, and Trelleborg have a range of machine mountings, but I'm not sure that they will allow much lateral movement. A damped elastomer bush will damp in the radial (loaded) direction, with low axial stiffness, but the only 'industrial' bush design I am aware of is made by Freudenberg (simrit). These are limted to about 400kg static load, and again the peak damping is in the 18-20 Hz region, so I'm not sure how effective these would be in your application.

Depending upon the movement required, it sounds like like a good application for a laminated rubber bearing / compression pad. If you aim for a vertical natural frequency of about 10Hz (not difficult), then the transmissibility should be no more than 15%, and the natural hysteresis of the material will provide some damping. Laminated rubber bearings are relatively soft in shear, and can easily accomodate shear strains of 50%.

Does this sound like an option?

Regards

Tom

 

[EnglishMuffin]

Actually, this is a large machine tool, with a 500mm by 630mm ram, about 12 feet long, containing the machining spindle. The problem is to optimize the dynamic stiffness of the ram (ie - it is not an isolation problem). The previous version of this did indeed utilize hydrostatic bearings, but before I got involved with the design it was decided to eliminate them and go with linear roller slideways, together with hydraulic damping pads, so I have sort of inherited it and I am trying to make it work as well as possible. About the only advantage I can see to this arrangement is that the stiffness and damping functions are now to some extent separated, and perhaps the damping can be optimized independently of the stiffness, which is not the case with hydrostatics. There must be an optimum amount of damping in this case - since if the damping is infinite no energy will be dissipated - but determining the optimum value is something else again. It may well be the case that elastomer pads could do the job better. However, I am faced with a very tight time frame, political aspects etc, so if I can make oil damping work, that is the way I will have to go.
Greg Locock: I read about the frequency dependent behavior in Harris & Crede - it is apparently a very marked effect and they give a graph (experimentally derived). It seems to be related to the inertia of the oil in the viscous gap. Presumably it can be compensated for (up to a point) by using a bigger gap.