Roundness of Sphere

[Matthew D. Hoedeman]

Well, I just had to carbon-burn a 1/4-20 tap into a 3/4" tooling ball. I want to check to see how distorted this might have made the ball. I am checking it with 99 hits on 7 levels. The tap goes in 1/2" so there is 1/4" of metal minimum (except where the tap went in). Does this seem to be enough or should I use more hits/levels to check this? The resutls I got were 0.013mm (or 0.0005") and tht is less than the machine specs. Opinions, anyone?

[#2#]

Well, I just had to carbon-burn a 1/4-20 tap into a 3/4" tooling ball. I want to check to see how distorted this might have made the ball. I am checking it with 99 hits on 7 levels. The tap goes in 1/2" so there is 1/4" of metal minimum (except where the tap went in). Does this seem to be enough or should I use more hits/levels to check this? The resutls I got were 0.013mm (or 0.0005") and tht is less than the machine specs. Opinions, anyone?

How much of the sphere did you cover with this? The whole thing, the top hemisphere only, only the part that matters?

Are you saying that .013mm is unacceptable? Or are you just questioning to ensure that your measurement was adequate and .013mm is the worst case scenario?

I know, so many questions...

[cmmguy]

Did you check it before the burn and if so what were the numbers like.
It would seem that the number of hits are sufficient. I just dont know how much probing/lobing error plays into this type of measurement.

[John Kingston]

Have you tried measuring a tooling ball that you haven't modified? Maybe measure some other sizes and see what kind of results you get? Measure one, smack it with a hammer and measure it again, see what change you get in measurement. Just some crazy ideas that would at least for some variation into the measurement.

[Matthew D. Hoedeman]

I only measure the top 1/2 of the ball and until I tap it, I really have no way to hole the dang things to check them. The spec sheet on them calls for 0.0002" Size and roundness tolerance. Machine specs are 0.0007" max, but calibration usually shows right around 0.0005". Using a TP2 so the lobs can add up, I didn't use anywhere NEAR that many hits to calibrate the tip. The hit spacing looks pretty cool on the screen, I was just wondering what all y'all thought about it and if I should use more hits. As for the 'end' tolerance for these, 0.002" is it, so the roundness equals 1/4 of my end tolerance. And no, these will NEVER get sent out to see how they check on a better machine, sorry, nor are they for any customer, they are for me and MY stuff here at work. I am going to mount them on studs and use them as scanning targets.

[#2#]

Can you use hot melt glue to hold them for a "before tapping" inspection? I know it isn't the best but it might give you some numbers for comparison.

[Matthew D. Hoedeman]

Using 5 other balls and 9 hits / 3 levels, they are up to 0.006mm out of round, so I measured the worst one, using 99/7 and got 0.012mm. So, I guess it is not putting enough heat into the ball to distort it.

[Andrew Gardner]

We burn a lot of threads, and depending on the oil temp, usually there is not a problem.

[BOB PARSONS]

Tooling Balls

Matt,

If you are measuring just the top hemi-sphere can you get a
plate w/ some .740 dia. holes made up so that the balls are captured in
the holes. I have done this before and the balls are held quite
well for measuring purposes.