Cylinder/shank/shaft probe Revisited Attn:Matt H

[Mike N.]

I have been challenged/instructed to use a shank probe (2mm) for measuring an edge on stamping. Up until the supplier crashed the die we were recieving accurate parts based on my acceptable program using a standard 2mm ball probe. The tolerance of this part (profiles of .010 to ABC) didn't require us to split hairs on the measurments. However now that the parts are on the fringe of the tolerance we have now been questioned (by the supplier) on my measurment techniques. I have since incorporated a shank probe and found similar variables using this method of obtaining data. Some of the things i have to consider are....edge break of corners, scale from Heat Treat, and die break of the measured surface.

What I ask of you Matt H and all others is..... where is the compensation of the probe being applied shank/ball ? What is the shank offset value for if it doesn't compensate from there? There isn't any way to tell "it" to compensate at the shank for this hit and from the ball for all other hits!

any input would be greatly appreciated and help with my debate

[NZ CMM Girl]

It is my understanding that the shank offset you enter when you qualify the tip is the only position at which it is calibrated. Of course this doesn't help when you don't know exactly where the tip will hit the part. When I used it, I checked it at different depths on a sphere and decided that the variation was within my measurement uncertainty.

Lenice

[J Temmen]

The shank qual basically does nothing. When you use a cylidrical probe to measure a opint it reads to the center of the radius at the end as if it were a ball probe. It will work great as long as the point you are measureing is perp to your probe, otherwise the greater the deviation from perp the grgeater the error you are introducing.

[Matthew D. Hoedeman]

When you get down to the absolute bottom line, I am not sure if it 'knows' anything about the probe between the ball and the spot on the shank it calibrates. So, unless you are hitting either on the ball or at the exact 'ring' on the shaft, how do you know if it 'knows' where you are touching. If you are using the ball and surface sample hits and the correct depth, then you 'know' that Pcdmis 'knows' where it is making contact. I think that the shank calibration and the cylinder probe is strictly for checking the edge of a flat piece of sheet metal, say a die cut or laser cut blank, where there should be no surface variation (which means there should only be a minimal amount of variation), so the actual depth will not matter so much.

To check the validity of your cylinder measurements, do this simple test:
1) Qualitfy the cylinder probe and align the part
2) Check all the various edge points using your current settings (which should be correct for depth of touch)
3) Check the exact same part, without re-aligning, but change the depth so that it matches the depth you used for the shaft calibration
4) Compare the numbers, I think that you will see a difference
5) Qualify a ball probe and align the part
6) Check the edge points using the correct depth value for the shear/breakage
7) Compare those values to the previous two sets. It should match the first set but not the second.

The more off 90 degrees you are to the part, the greater the difference will be between correct depth and the shank depth. In all honesty, I do not know that the shank calbration will really do anything for anyone and any time, except when checking a flat piece of sheet metal. That is the only time I can think of that a cylinder probe and the shank calibration would be useful, but then, like you said, how do you tell Pcdmis you are using the shank and NOT the ball? I think it would have to be either/or, not both, when you calibrate. If you calibrate the ball on the end, that is where ALL the data will come from. If you calibrate the shank, then it is up to the operator to make sure that ALL touches are at the shank depth, and the only time I can see that happening is, like I said, on a flat piece of sheet metal. So, for formed metal/plastic/carbon/misc-stock, use the ball and use it correctly.

[Paul Sarrach]

For sheet metal and Plastic, There is a need for the shank. And if done correctly they work. The important thing is that you have to square up your Ph_whatever_9,10 And you must square up you fixture and or part, to be true to the machine and probe, which I have seen most programmers do not do today. So that makes me want to do a poll.

[\v/inston]

the problem is, per matts example in the old forums, that the shank probe could accept bad parts or reject good parts, if the trim area you are checking is right at or near the USL or LSL.

http://www.brownandsharpe.com/anyboard/forums/pcdmis/posts/30134.html

[Rick Coleman]

Cyl stylus

The ONLY time I will use the cylinder or barrel stylus anymore, like Matt said, is when checking a flat blank. Be it laser cut, die cut or water jet, I trust what I get here; material .8 to 2.5 mm thick. With some of the crazy draws and angles on the stampings we produce, I would much rather take the extra time and use an edge hit with one surface hit where I can control the depth of measurement with a ball stylus. Much better results.

[\v/inston]

only bad thing about using the ball probe is if your parts have a large variance, then you may have to adjust your edge points for every part.

[Matthew D. Hoedeman]

I have yet to see parts with more variation than the prehit/retract that I use (and the Pcdmis defaults to) of 5mm. You have to use automove to ensure it misses going from the sample to the touch, but the little extra time it takes (even when checking 200 edge points) is sure a cheap cost for piece of mind, and I can use as many pieces as I can get, let me tell you! In plastics, you might run into that, but I have never seen it in sheet metal. Sure, you might have to make some wierd adjustments if it is a checking fixture it sits on, but you would have to make the same adjustments if it were a cylinder probe. One surface sample hit, one hit on the trim edge with the automove and I report BOTH the edge deviation and the surface deviation, so by doing it this way, I have eliminated a check point on the surface, so did I lose any time, really? I like the T&S axis' for edge points (but I love T& something else!).

[\v/inston]

A

i have had to adjust my edge points alot, but the die makers were making huge changes to the die that was majorly effecting the parts.

poll are you a T man or an A man?

[tenndoc]

A man!