Rotary Table

[[Pa...]]

Hey guys,
I’m trying to measure a part using the rotary table. I haven’t had rotary table training, so I’m struggling a bit. I’ve included a picture of the part to help with my explanation. The part is aligned on the rotary table with DATUM B in the X axis. DATUM A is through the bolt holes (yellow). There are (42) holes equally spaced on the side DATUM C. The first hole on the side is in line with DATUM B, see picture. The part is two halves bolted together, they want the top half circle measured and compared to the bottom half circle. They’re asking for true position and circularity of the upper circles and their deviation from the lower half circles. I measured the first hole(#1) and made a pattern of the (42) holes. It physically runs all the features with no problems. It measures hole #1, and then rotates the part to measure the remaining holes in the pattern. True Position starts out evaluating YZ work plane as it should, but changes to XZ work plane, then back to YZ. Even though it rotates and orients the holes as it should, it’s evaluating the holes like the machine is moving to the hole, instead of the hole rotating to the position of the first hole which is YZ ?

HOUSING RING ASSY_B932307.pdf

[[Pa...]]

Any ideas ?

[[De...]]

The simplest solution is to make separate true position characteristics and modify the rotation value once you have clicked the special button. Of course this would mean manually creating 42 different characteristics - ugghhhh! 😮

Another solution is to make a right click in the rotation spot and choose formula. Then using PCM create a formual that modifies the angle based on the LOOP1 variable that is created when using a pattern. LOOP1 will automatically increment for each feature. 😃

In this case the formula would be something like
((360.0/42.0) * (LOOP1 - 1.0))

I can't recall without looking if this will require rad() like follows or not. It might simply be a straight angle - its always fun trying to determine whether Calypso wants its angles in degrees or in radians.

rad(((360.0/42.0) * (LOOP1 - 1.0)))

[[Pa...]]

Please sign in to view this quote.

Derek, thank you very much for your help with this. I will try your ideas when I get back to work. Is there a source I can use to learn more about using formulas and the rotary table? Again, thank you !!

[[SH...]]

Please sign in to view this quote.

Best way to learn things is reading, so read the PCM and rotary table manual.....

[[De...]]

Unfortunately, there really is not much for examples that I have found.

I saw a reference to the LOOP variable in the documentation somewhere and tried it to see how/if it worked with patterns when I was in a similar situation. The part I was working on at the time had a feature that repeated something like 144 times around the perimeter. I was so disgusted with the idea of making 144 different characteristics and manually setting the rotation values that I spent some time testing different things to try and find a different solution. The formula is what I came up with.

BTW, I guess I should have stated this, but in case you are wondering, the (LOOP1 - 1.0) in the formula is there because for whatever reason, the LOOP1 variable in Calypso is based at 1 instead of 0, and the goal is to have the value at 0 the first time through the loop and then increment each time through the loop as the value of LOOP1 increments.

[[Pa...]]

I will try the recommendations tomorrow. I'm not sure if we have a PCM or rotary table manual. I've got some investigating to do. I sincerely appreciate the advice, will let you know how it goes.

[[SH...]]

All inside in your computer c drive...

IMG_20200526_113334.jpgIMG_20200526_112528.jpg

[[Pa...]]

Well after struggling and realizing I was out of ideas, I called Zeiss support. I was adding the formula to the feature (circle) with the pattern. I was instructed (Zeiss support) to use the formula in the True Position characteristic. Still not working right, I gave up at the end of the day. My coworker has been trying to help figure this out. He just called to say he found out what we were doing wrong. We were putting the formula in the wrong place inside the TP characteristic. The guy from Zeiss called back to check on us, and he got us straightened out. I feel like a burden just lifted off my shoulders. I can't thank you guys enough for taking time to reply to my post.

[[Pa...]]

Please sign in to view this quote.

Shabu, I appreciate you pointing me in that direction. Will check it out.

[[Pa...]]

Well I looked at the results from the rotary table part this morning. All the holes are now being evaluated in the YZ plane as it should. However the Y axis nominals are creeping up from "0" at hole #1 to about .0001 of an inch plus per rotation. The Y nominal increases until it plateaus, then starts decreasing by the same increments until it reaches zero again. Acting like it's not centered to the rotary axis. I'm open to suggestions.

[[Me...]]

I don't have a rotary anymore so going off memory.

You want to use the part itself as the axis for the rotary table. For example, if you were checking a camshaft you would create a feature like a 3D line from the axis at the beginning of your program and in the rotary dialogue box you would select the feature as the axis for the table. This will take the wobble and off center condition and correct it based on the part itself, not the rotary alignment program.

[[De...]]

I would look at the pattern. What is the alignment the pattern is rotating around?

If it is different than the center of rotation of the holes, that is a likely source of error.

You should make an alignment that uses the same features in your true position characteristic, and use that in the pattern, so that the center of rotation (origin) of the pattern is the same as the center of rotation (origin) of the true position characteristic.

Also, make sure the X and Y values of the center or rotation are set to 0.0

Unless you changed it, the alignment for the pattern is likely set to base alignment, and the origin of the coordinate system of the base alignment probably differs from the origin of the coordinate system of the datum reference frame used in the true position characteristic.