Help coming up with vibration control for machine tool 1

[fastline12]

There are apparent balance issues with a machine tool gear box that is causing vibration to be translated through the spindle and into the part. Rework of the assembly will take huge man hours and expense so will need to be planned. I am trying to come up with a quick solution to at least reduce the issues until we can go inside the unit.

I have already added approx 50lbs of lead to the assy but that does not seem to be fixing the issue. The trans sits on rubber isolators but is apparently not enough to repair this issue. I am curious if anyone has any "modular" type ideas where I can at least dial in a certain speed or frequency where they can make finish passes with an acceptable finish. I was thinking along the lines of a fluid damper but have no experience in designing them so would be purely a guessing game to some extent.

Again, I realize fixing the cause is what needs to happen but the machine cannot be down for a month right now.

thanks

 

[drawoh]

fastline12,

Do you understand what is causing the vibration? For example, your machine could be translating in X, Y and/or Z. Properly specified anti-vibration mounts and a splined U-joint shaft will isolate most of this. If the torque coming out of your machine is pulsed, your vibration is being transmitted through your drive shaft.

Have you correctly specified the isolators for the mass and frequency? Are you sure you are able mechanically de-couple your gearbox from the rest of the assembly?

I have seen someone isolate vibrations by sitting the module on rubber stand-offs, and running screws through the module and the stand-offs into the base. There was no possibility of this mechanism working, even if the rubber vendor had provided specifications for their parts.

JHG

 

[fastline12]

The power is transferred to the spindle via cog belts. The isolators are the basic T shaped units with a bore through them. A sleeve is installed through a pair of these. I was unable to get a durometer reading on them yet. They cannot be altered much because of belt alignment and belt tension concerns.

The vibration only translates in one plane (Y) because the head is a cantilever design so one plane is extremely rigid but another is less so and that is where our problems exist.

 

[GregLocock]

Measure the vibration at the toolpost bed and elsewhwer, in all 3 directions.

That'll tell you the frequncy of interest.

I must admit 500 lb of lead would be expected to solve many vibration problems, how big is this machine?

Cheers

Greg Locock


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

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

 

[fastline12]

Machine is only 9000lbs. VMC. We are at 50lbs, not 500lbs of lead. I supposed we can add more though. Maybe we are not near close enough.

The drive motor is face mounted into the trans standing vertical. I added my mass on top of the motor because that seems the place of the highest amplitude.

 

[Tmoose]

un- "Balance" usually creates vibration at 1X the rotating frequency of the component with excessive unbalance.

Identifying frequency of the problem vibration is key. Since the problem is reportedly making it to the part I envision there is a visual and maybe even measureable pattern on the part. Part rpm X pattern variations is THE frequency I'd be looking at. It is entirely possible to measure vibration on the machine that is not related to the problem. Whole body motion won't effect the part if the spindle and workpiece are moving together.

Got pictures of the machine, and the part problem?

Is the transmission on a separate frame from the "machine tool?"

When you say cogged belt do you mean timing type belts?

http://www.gatesmectrol.com/common/images/mectrol/ezlit/flange.gif

Or V belts with notches ?

http://www.reliant.com/en_US/Platts/art/HVAC_01_fig1.gif

Is this a grinding machine? Many times flat belts or possibly micro-V are the only types "smooth" enough for grinding applications.

 

[drawoh]

The power is transferred to the spindle via cog belts. The isolators are the basic T shaped units with a bore through them. A sleeve is installed through a pair of these. I was unable to get a durometer reading on them yet. They cannot be altered much because of belt alignment and belt tension concerns.

Why would you measure durometer hardness of your rubber mount?

Read the catalogue and find out the natural frequency of your system sitting on the anti-vibration mounts. If you have no catalogue with specifications, get another catalogue that does have them, and order stuff from that catalogue.

Anti-vibration mounts will increase vibration if they are not selected carefully. In some configurations, anti-vibration mounts provide a degree of freedom of motion that enables the vibration problem. A lot of systems do not vibrate because they are bolted down and constrained from moving.

JHG

 

[MintJulep]

Start for Tmoose for this brilliant observation:

Whole body motion won't effect the part if the spindle and workpiece are moving together.

(although he gets demerits for not knowing the difference between effect and affect.)

 

[BrianE22]

I am curious if anyone has any "modular" type ideas where I can at least dial in a certain speed or frequency where they can make finish passes with an acceptable finish.

There is machine tool software available that monitors vibration and can increase or decrease spindle speed and/or feedrate to put the machine in a better "chatter free" zone. I can look for some links if this is of interest you.

 

[GregLocock]

Sorry I misread the 50. The bad news is that if it is out of balance then just adding mass to the system won't help - unbalance forces can be very high indeed.

Cheers

Greg Locock


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

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

 

[Cockroach]

Correct Greglock, addition of mass is counter productive. But what you can do is try to home in on the resonant frequency using a deliberation bar, essentially selection spring constants that mitigate vibration.

This I expect, would be an iterative solution. You have no immediate data from which to figure out resonances. I would talk to a manufacturer such as Sandvik or alternate supplier and tap their experience.

Good luck with it!

Regards,
Cockroach

 

[Tmoose]

Hi FAstline - where are you located? Hiring a vibration analyst if you don't have access to the right tools would net frequency information real quick, and might not even require shutting the machine off.

Frequency analysis metaphor/analogy in poor taste.
The sandwich shop plays the Oldies radio station (103.3). When I wait for a sandwich the music I hear dredges up such unpleasant memories it ruins my lunch and I'm depressed for 2 hours. Dragging down the Shopping channel (106.3) broadcast tower with a bulldozer would be fun, but probably won't make me enjoy lunch any more.

 

[BrianE22]

I did a search on "stability lobes" for machine tools. The process itself is a bit involved (new analysis for each tool and workpiece). The following paper goes through the process of improving a machine tool's dynamic capability:

http://www.mfg-labs.com/FileDownload/Papers_Presentations/Leigh AHS paper-old version.pdf

For them, they had a spindle resonance problem at 59 hertz which they mitigated with a tuned mass damper. Maybe a TMD would work for you.