Very basic fundamental question regarding CMM (sanity check actually)

In the picture, does the 3D coordinate measuring machine's touch probe register the measured point at point a. or point b. ?

I think we need to know more about the process you are using the CMM for.If you are establishing an X or Y datum you would use the probe on the sides of the workpiece and if you were probing the surface to establish a cutting strategy for a similar radius ball end mill,you would be using the centre of the spherical radius to establish a toolpath.For inspecting a 3D surface I would expect a parameter in the CMM software to enter a tip radius/diameter from which surface offset could be established for comparison with the surface CAD file.

Quote:

---

Originally Posted by myhäje

In the picture, does the 3D coordinate measuring machine's touch probe register the measured point at point a. or point b. ?

---

Do't you just tell the machine / cam etc what the radius of the ball is (offset). Then it deducts the offset and sets the probe smack bang in the center once it touches?.
Now any tool diameter installed will work as it should.

That's how I thought it was done anyway.

Quote:

---

Originally Posted by myhäje

In the picture, does the 3D coordinate measuring machine's touch probe register the measured point at point a. or point b. ?

---

If your machine is calibrated then you set up your part referance, or fixture, from that information the CCM can measure where ever the probe touches is then a recorded point, an angle you would be probing multiple places along the angle to determine the angle and the position of the angle

So your question is the contact point of the probe when it is triggered is the measured point, all calculations are done from what has been setup in the CCM control with the probe diameter that you are using Etc.

hy myhaje :) that image may be somewhere in between these scenarios; if that measuring would be on :
... a cnc, is neither a, nor b
... a cmm is a 3d vector, that is fully located inside the image that you shared, only if there is no component among a direction that is perpendicular on the image

i once had to deal with a similar situation, and only the automated process of measuring, computing and showing data was not clear enough, but luckly i could inspect the measurements file that was generated by the measuring process, before being sent to the interpretation software

if possible, try to analize measuremant data ( before being interpreted ) from a sphere as you change altitude; look into the files that are created into the task/job folder / kindly :)

OK, maybe a better question would have been "Is the measured point always the same as the contact point regardless of direction/approach vector when using a touch probe?", assuming everything's been calibrated properly. It is a simple yes or no answer.

mostly yes, it should be, but you have to verify this

Quote:

---

Is the measured point always the same as the contact point regardless of direction/approach vector when using a touch probe

---

more precise, the point is always the same, or it is within machine accuracy, as long as direction/approach vector is within precision cone

the precision cone represents the range accepted for the aproaching vector direction, within which the accuracy is guaranteed

for example, you can never go perpendicular to a surface, because perfect perpendicularity does not exist, but the measurement will be ok as long as you are within 90degress±accepted deviation, thus you can approach a little tilted, but not too much

you can detect this error : for example, probe the upper surface of a sphere only from above, then from polar directions, then compare data from those 2 sets

the calibration process of a cmm, when using that ceramic sphere, begins with measurements paralel to machine axis, also coaxial with the polar directions ( like at the equator, and the top, maybe also just a bit offset from their location ), and after sphere center is found, it may move to directions that are located somewhere arround the middle of the upper half, or right under the equator, and these final testings are done in order to calibrate the precision cone, or the direction cosines

if you need numbers, then that cmm should be tested, similar to how you would test/calibrate measuring on a cnc

if i may ask, what happened, why are you digging this ? depending on part's shape and desired accuracy, maybe you can measure it with an alternaitve method, then compare results to the ones from the cmm / kindly :)

hello again :) i have been thinking, try a simpler test, that does not involve searching for measuremnt data, but only use your cmm software as usual

pls check attached :
... left : use a prism with a known width, put it aligned with the cmm axis and measure it's width; the cmm should return a value pretty close to the real one
... right : repeat the measurement a few times, increasing the tilt angle, and, after a value, the cmm measurement should fail / kindly :)