Vertical runout of a torus

[[Me...]]

What would be the best practice to measure vertical runout of a torus? I have been measuring distance "d" in 10degree increment around the big diameter of the torus. (Using cartesian distance with self centering turn on)

overview.PNGsection.PNG

[[Ri...]]

I can see nothing wrong with that approach. If I were checking this on a surface plate, I would locate bottom face on a surface plate, place a gage ball of an appropriate size in the torus and measure over the top of the gage ball at various locations.

[[Da...]]

You could also use a formula and scan the bottom of the torus. If that's what you're looking for, I can explain. 🤠

[[Me...]]

Please sign in to view this quote.

Thanks for your input Richard. We apply that technique to verify cmm result. But its going to need alot of manual measurement to get the data since we do inpsection on all parts.

Please sign in to view this quote.

That sounds like what we are doing. Could you explain more about it Dave.

What do you guy think about measure lowest point of small radius(of a torus) with self centering? How reliable it is?

[[Ri...]]

Huy, a couple of questions need to be asked. First, what kind of tolerance are you trying to hold? Second, how reliable is your manufacturing process? If you have generous tolerancing for this Runout, then several self centered points at the bottom of the torus will work very nicely. You can use Cartesian distance from the bottom face and Min and Max Characteristics to get the high and low. Then Result element to add those to two values together and there's runout. Now, if your tolerance is tight, then you will need to be more thorough. The same can be said for the reliability of your manufacturing process. If it is virtually impossible to produce a bad part, then the self centered points at the bottom of the torus will be fine. If the manufacturing process isn't rock solid and there is a significant possibility of a out of spec condition for this callout, again, you will have to be more thorough.

[[Da...]]

Well, I am not exactly sure if you can recall the radius checks that you are doing into a torus feature.

But even if you can't you can use RECALL for several of the circle section that you checked, It will recall the centers of the circles. ReCall them into a circle. We are only using this recalled circle to get its X and Y locations. Now you can create a Plane. Fill in the Z height and Type up coordinates that match the position of the circle. Now, in the plane go to strategy choose a circle path. When you open the strategy select the circle size and Left click to bring up a formula. In the formula window select the Actual of the RECALLED circle then select Diameter. This will now use the actual diameter of the Recalled circle , and use that for the circle path of the plane. Next , do the same for the X and Y value of the Recalled circle. Now you have a circle path that is pretty close to the bottom of the torus, it will adjust for variations in center or diameter...

Ask for axial runout of the plane to whatever your datum is, that should do a pretty good job. Because of form errors it won't be perfect but it should give some pretty good numbers.

sorry sometimes my explanations get hard to understand .

[[Ri...]]

Under Resources/Features Settings Editor/Probing/Self Centering Probing it might be possible to turn that on for a Plane Feature with a Circle path on a plane. I am not sure how that might work. That's one I've never tried.

[[Me...]]

Richard
Tolerance for this vertical runout is not to excess .0015" every 30degree along the torus diameter. Our manufacturing process is pretty reliable. We design special fixture for both mill machine and cmm to make sure bottom plane is restrained to fixture surface. Inspection result going off recently so im questioning if its because of fixture worn out or my cmm program is not good.

Dave,
What i did is recall 72 cirlce section to get the center of circle. Then i create 72 self centering point base on diameter provided in blueprint. Your method seems to produce better result because it measure from the actual diameter. I will try that and compare the result.

[[To...]]

Not sure how well the self-centering will actually find the lowest point because you're probably driving in the vector direction and so I don't know how much sliding will occur in X and Y, but maybe at that point, there is not a lot of difference, especially for the Z direction.