Perform a test on Z-axis?

[pernilla]

Hi.

I was told by Hexagon at our last CMM calibration that the balance cylinder in Z-axis was worn out, in our 20 year old CMM - Mistral. I don't recall the exact info but there wasn't anything to do about it at the time because they had nothing to replace it with and they told med the machine would work fine until next calibration (in dec).

But I'm starting to wonder... Often in Z-axis the CMM probes the same point twice. It's like it missed the first time and has to try again.
And also we got one part that uses A45B0 angle to measure in two half cylinders that has to be parallel to the plane that cuts them. And the problem here is that every now and then it shows out of tolerance but when I measure it in the other CMM (Scirocco) it's really good, nothing is oot. If I remeasure the same part in Mistral a while later, it's also fine.

So here's my question; can I test the Z-axis in Mistral to see if there's something wrong with it? And how would one do that?

[JEFMAN]

It depends on what you've got to do it (gauge blocks, ring gages, cylinders, ball bar...).
An easy test to check the Z roll is measuring a gauge block at 45° in the YZ plane (I assume here that Y is the long axis of the cmm) with A90B90 anb A90B-90.
The length should be the same, the Rzz is the difference of length divide by the tip offset.

In your case, I don't think the balance cylinder causes this kind of problem. I would suspect a wiring problem or a motor problem for the double hits, or maybe a probe defect which could also explain the cylinder prolem (but I can be wrong !!!!!)

[pernilla]

I've got gauge blocks and ring gages. And my long axis is the X.
So I'll place the gauge block at 45 degrees, measure the length of it with A90B90 and compare it to measuring with A90B-90. I should probably make a small program to run it in cnc mode and get the same measuring speed and force... But what does Rzz mean?

When I calibrate the probe it seems fine, and using it to measure other parts (with different angles) works fine. And also as I stated above, most of the times it works fine... Maybe you're right. Could be some wiring problem.

[JEFMAN]

Rzz = rotation of Z axis around Z axis = Z roll.
Rxy = rotation of X axis around Y axis = X pitch (..)
There are 9 rotations (roll, pitch, yaw for x, y and z)

[pernilla]

Got it. Thanks!

[vpt.se]

Could possibly be connected to a bad motor in Z...

[pernilla]

Okay... Is there any way to test for bad motors, or should I just contact Hexagon about this?

[vpt.se]

If you have scanning capabilities (continuous scan, not stitch scan) you could try scanning a known surface in the XY-plane, evaluate flatness and enable graphics view on. Look at the shape of the scan to see if it can tell you something. Perform the test in both of your machines and compare the result of the graphical representation. In the "bad" machine the shape of the actual scan can look like a sawtooth pattern caused by a struggling motor. Maybe even the flatness results differ that much between the machines to suspect a bad motor.

[JEFMAN]

I think you should call Hexagon.
Maybe there's a possibility to test the motor efficiency with a dial comparator instead of the probe, and some move points along z, but I'm not sure about this (the probing circuit must be closed to do it).

[pernilla]

Thanks a lot. I'm going to call Hexagon and ask. I don't have scanning capabilities unfortunately.

[pernilla]

This issue was resolved today.

After contacting Hexagon they came by and replaced the balance cylinder and motor in Z-axis, and a few other parts that Google won't translate to English for me.

So far it's working better than ever!