Angle Checks

[Darren Johnson]

We have a Microval Pfx with pcdmis ver3.0864 that we are having a few problems with. I am wanting to know how close this machine and software can check an angle. A little bit more info on the machine, it is a manuel machine with an older computer. We have a part for a customer of our the we have to check an angle within 9min. Is this machine capbable of that or would be wasting time and money on this. Would a dcc machine be much more capable of checking the angle with the tolerance we have? We are trying to find the best way to do this. Thanks for your time.

[bw_bob]

We have a Microval Pfx with pcdmis ver3.0864 that we are having a few problems with. I am wanting to know how close this machine and software can check an angle. A little bit more info on the machine, it is a manuel machine with an older computer. We have a part for a customer of our the we have to check an angle within 9min. Is this machine capbable of that or would be wasting time and money on this. Would a dcc machine be much more capable of checking the angle with the tolerance we have? We are trying to find the best way to do this. Thanks for your time.

How much of an angle are you checking. It its like 35 degrees then you would be hard pressed to check on a manual machine. I know of no way of compensating for the sin error in your hits. A dcc machine is the way to go. My .02

[Darren Johnson]

The angle we are trying to check is a 35deg angle +0 -4min. I know this is something very hard to check on a manuel machine due to the fact that I can not hit the same exact points everytime. Thanks for the reply.

[Bill **********]

It isn't even hitting the same points thats a problem. The problem will be cosine error. If you aren't vectoring into the touch on the vector of the surface you're measuring, you're inducing cosine error into your measurement. The ruby isn't hitting the measured surface on it's equator. For a tolerence like the one you're trying to hold a manual machine is a total waste of time in my professional opinion.

Bill

[Darren Johnson]

Thanks for the answers, I really do appreciate them and anymore that might come up. You guys have definilty answered my question. Thanks again for your time.

[#2#]

You could alway set up a 35º angle with a sine bar and measure to see how repeatable you are. While that won't necessarily give you and answer to overall accuracy it will tell you if your repeatablility is outside of +0 -4 min.

So, if you can't repeat within say 15 min then I definitely would not trust the results.

[Jan d.]

You could alway set up a 35º angle with a sine bar and measure to see how repeatable you are. While that won't necessarily give you and answer to overall accuracy it will tell you if your repeatablility is outside of +0 -4 min.

So, if you can't repeat within say 15 min then I definitely would not trust the results.

Great idea. This is a good way to evaluate your skills and technique. Refine until you get within what you need.

The machine is accurate enough to measure what you need. It is just a matter of skill and technique. Practice until you get it right.


Jan.

[cmmguy]

It isn't even hitting the same points thats a problem. The problem will be cosine error. If you aren't vectoring into the touch on the vector of the surface you're measuring, you're inducing cosine error into your measurement. The ruby isn't hitting the measured surface on it's equator. For a tolerence like the one you're trying to hold a manual machine is a total waste of time in my professional opinion.

Bill

NO cosine error in the measurement of planes or lines. Only single, individual, compensated points can have cosine error...

[craiger_ny]

I agree that the tolerance is demanding (±0.033º) for most features, and in the hands of the wrong CMM operator of a manual machine problems could arise. However, I must say that a manual machine will comp well and cosine error will be very clean (practically non-existent) on a plane. Manual machines are great and even better in the hands of an astute operator. If you have a manual machine give it a try. Go ahead and vector poorly to a plane and check your results. I grew up on manual and I do realize you can get negligibly more accurate results when vectoring properly (how much better depends on the feature type) but the comp math in PCDMIS and most microprocessor controls does a very nice job with 3D features.

Craig

[Paul Barriage]

Measuring the ANGLE will do well based on the length of your lines. The location may require a bit more justification.

[dwade]

However, I must say that a manual machine will comp well and cosine error will be very clean (practically non-existent) on a plane. Manual machines are great and even better in the hands of an astute operator. If you have a manual machine give it a try. Go ahead and vector poorly to a plane and check your results. I grew up on manual and I do realize you can get negligibly more accurate results when vectoring properly (how much better depends on the feature type) but the comp math in PCDMIS and most microprocessor controls does a very nice job with 3D features.

Craig

I too grew up on a manual machine and found my greatest error in measuring angular surfaces was consistency of speed and force. Make sure you measure the same direction (up/down/sideways) with the same speed and force on all of your points.

Duane