Help designing a simple vibration-isolator?

[duanebw]

Hi all,

I am trying to design a vibration isolator, to protect a fairly small object from vibrations caused due to travelling over rough ground.

I have used an accelerometer to record the G's in X Y and Z directions experienced due to the surfaces. I have done Fourier Transform on this time-series to get the frequency sprecta, and from this have selected the dominant frequencies to target.

I would like a simple passive isolating device comprising two shock absorbers from an RC car (these allow spring rate and damping oil viscosity to be adjusted), with the object mounted them, to allow the object to translate up and down (the plane experiencing the most motion). I therefore need to specify the requirements of the shock absorbers, and this is where I'm a bit stuck - not sure how to get there from what I know.

From here -

http://www.farrat.com/detailed-vibration-isolation-theory-68.html

I can see that I want the natural frequency of the isolator to be less than half of the excitation frequency (so if the dominant f_e is 10Hz, f_n of the isolator must be 5Hz or less).

The above link has equations to work out f_n given the spring rate, damping coefficient etc, which means I can iterate to get an acceptable spring rate and damping coefficient.

I've been told that I can calculate vibration velocities by multiplying the accelerations by 1/(2pi*f), and same again to get displacements, but I don't think is is very accurate at all?

I have been reading through Rivin's 'Passive Vibration Isolation', but am finding it very confusing. I have access to MATLAB and Simulink, but have never used Simulink before so would have to learn this from scratch.


Any guidance or pointers would be much appreciated!


Many thanks,
Duane.

 

[IRstuff]

I would start off by using a heavy structure to mount your object, and then isolate that system.

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[duanebw]

Unfortunately the mass of the structure the object is attached to is not adjustable.

 

[boyze]

you mention that the RC devices have an oil damper - what is the spring element? a road surface is apt to be a random vib input environment unless it has some "regular" features and the device speed is constant. were the accel's on a "rigid" device traveling over the road surface?

The Farrat info is the usual stuff for elastomeric isolators which are usually limited down to 4-5 Hz in standard configurations. The physics will be similar for other types. If you intend to use an elastomeric isolator(s) then you have the option of asking one of several standard isolator suppliers to size the device for you. They have lumped mass calculators, using the Farrat type equations but in a 6 degree-of-freedom system (suspended rigid body), that will select an off-the-self isolator, if available.


Have Fun!

James A. Pike

http://www.xl4sim.com

http://www.erieztechnologies.com

 

[FeX32]

"I can see that I want the natural frequency of the isolator to be less than half of the excitation frequency". That depends.
Are you mounting a camera onto an R/C vehicle? Or are you just using the shock absorbers from an R/C car solely as isolators?
More information on your project will allow us to provide better feedback for you.

Keep in mind that due to the transmissibility of the shock absorbers, they usually only isolate vibration at higher frequencies.
Also, if you are using 2 shocks, the way they are setup would make a considerable difference.


"Simplicity is the ultimate sophistication." L. da Vinci
- Gian

 

[GregLocock]

Is your data top secret, if not, post it? what is the target for vibration on the object?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[duanebw]

Thanks for the replies everyone.

I'm afraid I can't say too much about the system. However - the object is a small vibration sensitive device, around 300grams. The structure it is attached to is travelling over rough ground at a constant speed - the surface roughness is approximately consistent, so not entirely random. The vibrations were recorded by mounting the accelerometer to the structure, just as the object is.

The RC car shock absorbbers I am thinking of using look like so;

My current idea has the object mounted between the two shock absorbers, each compressed halfway when in the equilibrium position, to allow upwards and downwards movement, with damping and sprung forces in both directions.


Thanks,

Duane.

 

[FeX32]

With a pinned connection on each end (top and bottom) the system you mount will be subject to lateral forces that will cause significant vibration. This is because the pinned connection will be in a singularity to sideways forces.

Notwithstanding that, your mass is rather small. If the frequency (generally) of excitation is not too high you will likely select a spring rate that is quite low. This will case large "static" deflections and maybe a bouncing effect.
I would look into a High-static-low-dynamic absorber system (if you can get or develop one). They are not really available to purchase (from what I know), but relatively well developed in the lab. One of a many refs is

http://www.sciencedirect.com/science/article/pii/S0022460X08009437.

Cheers,


"Simplicity is the ultimate sophistication." L. da Vinci
- Gian

 

[boyze]

who knew the RC world had such sophisticated strut systems can you get spring dampers at the stiffness levels you're looking for? using the target 5 Hz and 300 grams the system stiffness is quite low.

Have Fun!

James A. Pike

http://www.xl4sim.com

http://www.erieztechnologies.com

 

[duanebw]

Thanks again for the replies.

Fex32, what do you mean by "This is because the pinned connection will be in a singularity to sideways forces."? The object should not be able to move side-to-side, relative to the structure, (apart from due to manufacturing tolerances between the shock absorber body and shaft), due to the design of the isolator. (My description above of the design is much simplified).

Regarding your second point, yes, this will be a trade-off I have to make - I am trying to keep the isolator and object total weight as low as possible, yet realise that the heavier the total, the more effective the system will be. I will check out the link now - thanks.

boyze, impressive right! You can also adjust the piston hole sizes, get ultra-slippery titanium-nitrided shafts, dualrate srings etc etc! From quick calculations, I should be able to get springs with a low enough spring-rate.


Thanks,
Duane.

 

[sreid]

Take a look at wire rope isolators.

http://www.enidine.com/Industrial/CWRMain.html

 

[duanebw]

Thanks, have looked at them, but they would allow the object to rock back and forth which is not desirable.

 

[IRstuff]

It's extremely unlikely that you will find anything that results in zero delta motion. What is your specification?

TTFN
faq731-376

Need help writing a question or understanding a reply? forum1529

 

[FeX32]

I just simply meant that you have to be careful how you setup the 2 shock absorbers.

I am trying to keep the isolator and object total weight as low as possible, yet realise that the heavier the total, the more effective the system will be.

If you keep the weight very small, the spring constant will have to be extremely small. You might end up having to use a slinky

If you use a high static low dynamic absorber you won't have to worry about weight.
Another very good solution to this problem is to use an active isolator. I have had much success with active isolators. Of course the disadvantage is complexity in development and power consumption.


"Simplicity is the ultimate sophistication." L. da Vinci
- Gian