Vibration damping and understanding. / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Vibration damping and understanding.

1 2 next

Move Thread

LAN_403

Patrick

Sat Sept 08 2018, 11:53PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

ive googled and googled but only become more confused.

I want to produce a broad spectrum vibration damping table of "medium effectiveness" kinda setting the bar low here. it would isolate a 2 lbs mechanical vibration generator damping down to minimal vibration to what it is attached to, then back up another 2 lb or so delicate object (camera or laser).

Im looking at a small (tiny) piston-electric generator for a 20 to 40 lbs rover. Even if there's a better solution to my power requirement, im wanting to understand the math and physics here. Im only interested in passive solutions.

For example i see this all the time:

But have no idea what it means. Does this show a spring bearing most of the weight? then a damper to act as a lossy component ? how do these two components act to cause decay ?

And then i see this:

1536451140 2431 FT1630 Vib 135 5 051003 F001

it looks like two more sophisticated cases of the upper diagram.

ive been studying these short desriptions :

Its 3 brief paragraphs on each of 3 cases.

the_anomaly

Sun Sept 09 2018, 02:05AM

Registered Member #19 Joined: Thu Feb 02 2006, 03:19PM
Location: Jacksonville, FL
Posts: 168

I think the spring is there to return the mass to its original resting position while the damper tries to stop its oscillation during return. In the first diagram, I don't think gravity is considered; the diagram is to represent motion in free space.

Now that I read the description, they say 'restoring force'. I think this is a poor way to state return to initial position.

MRMILSTAR

Sun Sept 09 2018, 03:36AM

Registered Member #62119 Joined: Sun Feb 04 2018, 04:59AM
Location: Cedar Rapids, Iowa
Posts: 136

Your diagram is a very good model of how the suspension system in an automobile works. For an electrical analogy, think about it this way. The damper can be thought of as a resistor. The mass can be thought of as a capacitor. The spring can be thought of as an inductor. With this analogy, what you have is a damped resonant system (RLC network) with a resonant frequency dictated by the values of the mass (capacitor) and the spring (inductor). If you remove the damper (resistor) and excite the mass (capacitor) with an impulse (bump), theoretically the system will resonate sinusoidally forever. However due to friction, heating, and other non-ideal component characteristics, the resulting oscillation will look like an under-damped oscillation gradually decaying to zero (the original mass position). By adding the damper (resistor), the oscillation of the system can be made to decay even faster. For an automobile suspension system, you would probably want close to a critically damped system.

I hope this mechanical-electrical analogy helps.

Dr. Slack

Sun Sept 09 2018, 06:08AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

You are at a huge advantage, starting from the position of being an electronic engineer.

All the concepts you have learnt about filters, spectrums, energy, signals, all apply 1:1 to the mechanical world if you make the correct transformations. You can use SPICE to show you what's going on. With those that support computed parameters and scaled displays (LTSPICE does, many others will), you could even have your inputs and outputs in mechanical units

Electrically, you have resistors that dissipate energy, and Ls and Cs that store energy, which are complementary and when connected together resonate.

Mechanically you have dashpots (series resistors) and internal heat-generating flexion (parallel resistors), and masses and springs that store energy, which are complementary and when connected together resonate.

Although in theory you could assign mass or springiness to capacitance, and then make all the other transformations to keep dimensions consistent, in practice it messes with your mind less if mass is capacitance and springiness is inductance. That way force is voltage (two pressure-type things) and displacement is charge (two quantity-type things). I'm not sure if I've got that the right way round, or the right powers of time in the right places, but you'll know and be able to correct me once you've done the sums. And energy of course is good old Joules.

Once that connection is made, you can use ALL the existing filter synthesis and analysis tools, ALL the filter types. Do you want stopband zeroes? (1) You got it! Limit your overshoot to x% following a step input? (2) You got it!

Of course the downside of network filter synthesis is needing to pay your specification dues. It only 'works' if you know what your excitation spectrum is, and what spectrum you can tolerate on your object, at the output of the filter. Or at least, you have to pretend that you have a spec and just assign passband and stopband frequencies, and a stopband attenuation depth.

You may not want a network-theory designed filter, maybe just an integrator, and a few stopband zeroes to cope with particular frequencies. You'll still be able to use SPICE to figure out what's going on.

(1) Those mass on the end of a spring things that are attached to cable structures like suspension bridges and ferris wheels.
(2) Arthur B Williams has graphs for overshoot versus order for all types, and versus passband flatness for Cheby types

Patrick

Sun Sept 09 2018, 03:32PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

MRMILSTAR wrote ...

I hope this mechanical-electrical analogy helps.

Dr. Slack wrote ...

You are at a huge advantage, starting from the position of being an electronic engineer.

I was hoping this would be the case.

im looking at some special elasto-viscous urethane compounds that are cheap and non-Newtonian, And supposedly demonstrate all three components (RLC) if properly arranged.

of course i would rather have a 3D MEMS sensor on the source side, then a MEMS on the "quiet" side and make my measurements that way as theres no way to calculate or simulate the noise pattern of a sputtering piston nor the acceptable "quieted" output of the mechanical system for optics.

enormously helpful please keep suggestions coming.

Sulaiman

Sun Sept 09 2018, 04:34PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Piston motor driven generators usually operate at 'constant' speed to regulate voltage
(and frequency if ac output),
with engine power (torque, fuel consumption rate) being indirectly determined by the electrical loading.
So the fundamental frequency is quite well defined, and you can model the 'sputtering' as harmonic.

In electronics we can cascade many filter elements easily and cheaply,
adding additional mechanical filtering stages becomes complex, and heavier.

In general, make all masses connected to the main body as light as practicable,
this produces less oscillatory energy (like VAR) in the resonant mass/spring and allows lighter damping.

Avoid having resonances at the exciting/forcing frequencies and their harmonics
and similarly, avoid having different resonant elements with similar resonant frequencies.

I know that for trains the suspension is designed,
then manufactured, then measured, then significantly 'adjusted'
before full production begins,
i.e. do not over-analyse the system,
do a rough implementation and 'tweak' it.

Patrick

Sun Sept 09 2018, 04:44PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Sulaiman wrote ...
i.e. do not over-analyse the system,
do a rough implementation and 'tweak' it.

Excellent.

Patrick

Sun Sept 09 2018, 04:52PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Sulaiman wrote ...

Piston motor driven generators usually operate at 'constant' speed to regulate voltage

But i cant use a normal alternator with its variable intensity rotor. i plan to use a properly sized three phase motor that will just be forced into a generator role. So voltage goes up with rpm and current withdrawn make the rpm go down. ill then use a SMPS to make the system appear to be a battery to all other circuits on my rover.

Sulaiman

Sun Sept 09 2018, 05:01PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Would it not be better to control the engine rpm to produce a constant voltage (dc or ac)
to make life easier for any smps(s)

Patrick

Sun Sept 09 2018, 05:30PM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

Sulaiman wrote ...

Would it not be better to control the engine rpm to produce a constant voltage (dc or ac)
to make life easier for any smps(s)

Yes a tiny dedicated Arduino will run a PID loop to the throttle servo, so the slow mechanics will respond to load (slowly) while the SMPS will smooth out the output with electrical 'speed'. Again to cause a battery like appearance.

1 2 next

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob