probe offset

[Homer-Jay]

I am new to polyworks, and am trying to get the basics down before i go to training. I have a question I think may be related to the way the arm compensates for the probe tip (romer infinity 2.0). If I probe out a rectangle, lets say a gauge block, and then probe two opposing points along the long axis with a known width of 1 inch, the two probed points will lie outside the boundry of the rectangle. the measurement of the width of the rectangle will be very close to 1 inch, but the two points will be up to .1 wider. I have held the tip perpendicular to the measured surface, parallel, at a forty five..etc and still cannot duplicate my micrometer measurement to tolerance. If I lay off a rectangle, or circle, the dimmension will be very close. But if I lay off two points, and create a distance feature, the measurement is usually from .05 to .1 off. I am carful to create a coordinate system that is consistent with my part, and use the feature control to display the correct dimmension. Am i supposed to qualify my probe with the artifact like i would the laser, or is it automatic? would appreciate any help, as I have learned the software well enough to do simple layouts on green castings, but don't know what to do about this error.

[Admin]

H-J,

I wish I could help you, but my probing skills are below par and i rely on the laser scanner for all of my inspections.

[Homer-Jay]

That may be what it comes to. I already plan on using the laser to do things like pattern wear, jig and fixture wear, tool and die wear etc. and probably even my machined parts. but i would like to do a point to point layout for my green castings. i would just use the quick check tools, but i cant export my list to excel then copy and paste to a controlled form like i can in polyworks. I am attending polyworks 5 day training later this month and will be able to better understand the intent of the software. because i dont think it is really intended to do what i'm wanting to do.

[Homer-Jay]

i believe what i was missing was a cad model to use as a reference.

[Jeff]

Your calibrations come from you calibration files. You do all the calibration in Cimcor arm utilities. When you take a point in polyworks you can compensate for the point in different ways. In the sub-menu pull down if you use standard and then in the compensation method us compensation point. When you take a point it will ask for a compensation point. Just pull the probe away from the part normal to surface and hit the except button again. This is nice because you can use any part of the ball and not have to be normal to the surface to take the point. If you want to add nominal you can do this after you take the point by right clicking on the point and then go to Define nominal then select numerically.  I hope this helps. Jeff

[jrayself]

the measurement is usually from .05 to .1 off.

If these deviations you speak of are in inches then I suggest you check your mounting setup for your arm and then re-calibrate.  If they are in milimeters then I also would suggest a re-calibration, but a deviation of 50 microns very well could just be the measurement error of your arm.

[Homer-Jay]

I will re-check my calibration. Also i will attempt the compensation point method. There are a few of our customers that may not be able to provide me with cad models to use as reference. so if i can do some point to point layouts without a cad reference, it will save me having to draw the models myself from their 2d prints. And i was in inch units.

[Homer-Jay]

Should i also change from probe orientation to compensation point in the probing options menu?

[jrayself]

Even without a reference model, there are very few things that can't be accomplished.

Should i also change from probe orientation to compensation point in the probing options menu?

I would, I usually use a compensation point or compensate to a plane.

[Jeff]

Yes change from probe orientation to compensation point in the probing options menu.

[Homer-Jay]

Jeff and J-Ray, This is fantastic news about being able to go around the reference models to to some probing. Based on what i was seeing in the tutorials, and another thread i had read on here, i was pretty sure i would need reference models to accopmplish anything. I'm so glad that i can simplify the process somewhat to do my green castings. After all we are mostly concerned with having plenty of machine stock, and verifying that our cores stay put. I have been successful in taking what you posted as far as using the compensation point method, and drastically improving my method. And i am also positive that the mounting surface i am using is sub par. we bought a facility next door to our current facility and will be getting a brand new inspection office there. so until i get my new surface plates, and my new inspection table ( hopefully hand scrubbed steel with tapped holes on centers) for the green castings, i am making do with what i have. which undoubtably has some minute flex in it. we got some training seats thrown in with our purchase, and this information, which has given me a head start will undoubtably allow me to get more out of the training classes. thank alot-Mike

[Homer-Jay]

on mine, i dont have to hit the trigger again after i pull the probe away from the surface i am laying the point on. it does it automatically. I'll play with the set ups some more and see what kinds of different setting you can ues in the compensate method. This advise has enabled me to go from .1inch error to .oo23 error in the worst case, so i am very pleased. my gauge block is taped to a bigger block to keep it from moving, and just sits on a carpeted floor. My arm is mounted to a 2X2X.125  piece of steel to keep iit from tipping over. so of course there is error. i hsed you guys advise and was able to get some pretty close measurements. i set the tolerance at .005 . i'lll be getting some better surface plates soon