2D-Position/Vibration measurment

[simasa]

Hello,

I want to measure a 2D-Position of a machine part. The movement may be like a vibration.

Aplitude 50-500 µm
Frequency 20-60 Hz (variable, depending upon speed of the machine)

There might be different ways, like using sensors (capacitive, inductive, laser based methods, laser triangulation etc.).

I would be very thankful if I could have some guideline in this form for a suitable approach addressing my problem.

Thanks in advance

Simrs

 

[IRstuff]

I thought you already decided on a laser system from your other post?

What are you measuring the motion relative to? What will the data tell you? How often? How accurate?

TTFN

FAQ731-376

 

[simasa]

Thanks for reply. This would be an additional measurement system that I would like employ. Its applications is as follows.
I am trying to measure the position of a moving/vibrating machine part which is moving relative to machine bed. This motion is actually decoupled from the machine bed as well as from the rest of the machine due to a special spring damper system. This moving part is indirectly attached to work piece mountings and its undesired movement may have influence on our knowledge about work piece. Machine is 3 axes machine. x-y axes on work piece. Z axis on spindle.

The required measurement should most probably be relative to machine/spindle coordinate system. It should be later used to modify CNC code to improve machining process. Measurement accuracy is required up to 2-3 µm.

Thanks and Regards
simrs

 

[IRstuff]

I'm not really getting a useful mental image of your system. If your workpiece is being machined while on this machine bed, I don't see how it could possibly function if the workpiece can move arbitrarily relative to the machine bed.


TTFN

FAQ731-376

 

[BobM3]

Is your "fixturing" compliant relative to the machine table? So you have a rigid machine table, a flexible fixture and your workpiece mounted to that fixture?

 

[simasa]

You are right. But this is a test machine currently installed with some new mechanical functionality for test purposes which cases a small scale relative movement between machine bed and work piece. Now problem is to measure it. This measurement will also be used to correct our knowledge of x-y position of the machine bed.
Regrds
simrs

 

[IRstuff]

I'll make an assumption that the machine bed is on the order of several feet in dimension, in which case, laser interferometry is probably the only way to go. This also assumes that the workpiece only translates, which is not obvious from your description. If the workpiece can change attitude, then all bets are off.

How is the workpiece position sensed currently?

http://cp.literature.agilent.com/litweb/pdf/5964-9307E.pdf

appears to be gross overkill from an accuracy perspective, but the potential 33-kHz sample sounds to be where you might need to be.

TTFN

FAQ731-376

 

[BobM3]

So you should measure the relative movement between the rigid table and the moving workpiece. Or can you not get to the rigid table? (Just wondering why you would mount the measurement system to a remote wall)

 

[simasa]

thanks for the valuable cooperation.
Machine bed, unfortunately, is not big. The total x-y travel is 200x200 (mm2). Additionally, I want to avoid laser interferometry. As the metallic partials interrupting the laser beam, we lose the information. It may also be expansive. High accuracy of the required measurement system is a key requirement.

Currently the work piece position is measured by linear encoders. This is a research machine. To achieve more accuracy, we are testing an additional and unconventional mechanical system, some how, attached with work piece fixtures.

The laser bases measurement system which I discussed in another post, supposed to mount on spindle assembly wall, would be an additional and independent system to this post. The idea of mounting it on wall was to get the measurement in spindle coordinate system. But I am not sure if wall is reliable enough and arm is stiff enough to accomplish it.

Regards
simrs

 

[sreid]

Heidenhain has these XY Grid encoders. They have grids much larger than the 68 mm size.

http://www.heidenhain.com/index.php...uktDB&op=catview&search=category&CS_UID=18366

These were originally used for an electronic "Ball-Bar" test.