?Is there any information available in regards of the quality of regitrars? ?How can I know if a registrar is competent (or still is since it was accredited)? ?Who qualifies them, and how?
Before going any further, there are two important questions to be answered... (1) Do these tolerances apply in the "restrained" or "unrestrained" condition? (2) Are you measuring "restrained" or "unrestrained"?
I have been down this road, and assure you that you need to look at the entire design, not just the ring gear to assure manufacturing integrity
You really have a tough one on your hands with that kind of tolerance. You did not mention the environmental conditions or the accuracy of the CMM or the roundness tester. Don't forget that one of your biggest enemies when trying to maintain that tight of a tolerance is going to be the room temperature and variation as well as the ability of your equipment to measure the feature accurately. Have you done an R&R on both machines? How about determining the uncertainties? If these factors have not been addressed, you are only looking at half the problem. I think you are on the right track with the Tallyround. I only know of a few (and very expensive) cmms that are capable of that kind of accuracy.
One place I used to work at had a SIP CMM and a M&M Gear Analyzer. The Lab was maintained at a strict 68 degrees F, with a limit of 4 people in the room at any given time. The parts were allowed to soak in the Lab for several hours prior to any measurements. (The manufacturing floor was also temperature controlled, but not as closely as the Lab.) I should tell you that the production was more of a job shop type atmosphere with normal orders of less than 5 pieces. (The product was highly specialized.)
Don't give up on the getting the spec openned up even more. Sometimes engineers ask for tight tolerances even though it is not necessary. They seem to think that the tighter the tolerance the better the product. If they still insist that the tight tolerance is necessary, inform them that the cost will reflect the added requirement. Gage manufacturers do it all the time. The normal price difference between a Class XX and Class XXX plug gage for example is double the price!
I hope this has helped out with you situation. Good luck.
I assume that your part is toleranced consistent with the ASME Y14.5M 1994 standard. If so, size and form are two seperate issues. Lets look at the size (diameter) tolerance first. If you consider two concentric cylinders, one at the maximum diameter and one at the minimum diameter, each with perfect form, all of the points on the part must lie between those two boundaries. The GD&T standard would recommend a go-nogo functional gauge to insure that no points on the part violate the limits of size.
However, attribute gauges are not conducive for process control or capability, so a number of variable gauging methods can be used to define the size. These can range from two point static methods (calipers/I.D. micrometers) to dynamic multiple point evaluation of the surface at one or more levels (tallyrond, scanning CMM, etc.). The correct one to use depends on the part and the process. If form is an issue, for example lobing with odd numbers of lobes, a two point gauge may not see the variation in the part.
Form errors such as circularity are determined in a number of ways. One way is to look at the maximum and minimim points from the centerline of the inner diameter as established from the least squares circle.
If I were to do this, I would evaluate capability for size based on the least squares circle, as this is one of the most stable algorithms. I would also look at the circularity (specifically the max and min points) to insure the limits of size are not violated.
Community metadata Could not be loaded. No pre-configured community forum parent found for the incoming fid: . You need to add this forum into custom_code/community.sql
Send to friend
Comments
almetzel 4/6/2005
Before going any further, there are two important questions to be answered... (1) Do these tolerances apply in the "restrained" or "unrestrained" condition? (2) Are you measuring "restrained" or "unrestrained"?
I have been down this road, and assure you that you need to look at the entire design, not just the ring gear to assure manufacturing integrity
rcade 4/5/2005
You really have a tough one on your hands with that kind of tolerance. You did not mention the environmental conditions or the accuracy of the CMM or the roundness tester. Don't forget that one of your biggest enemies when trying to maintain that tight of a tolerance is going to be the room temperature and variation as well as the ability of your equipment to measure the feature accurately. Have you done an R&R on both machines? How about determining the uncertainties? If these factors have not been addressed, you are only looking at half the problem. I think you are on the right track with the Tallyround. I only know of a few (and very expensive) cmms that are capable of that kind of accuracy.
One place I used to work at had a SIP CMM and a M&M Gear Analyzer. The Lab was maintained at a strict 68 degrees F, with a limit of 4 people in the room at any given time. The parts were allowed to soak in the Lab for several hours prior to any measurements. (The manufacturing floor was also temperature controlled, but not as closely as the Lab.) I should tell you that the production was more of a job shop type atmosphere with normal orders of less than 5 pieces. (The product was highly specialized.)
Don't give up on the getting the spec openned up even more. Sometimes engineers ask for tight tolerances even though it is not necessary. They seem to think that the tighter the tolerance the better the product. If they still insist that the tight tolerance is necessary, inform them that the cost will reflect the added requirement. Gage manufacturers do it all the time. The normal price difference between a Class XX and Class XXX plug gage for example is double the price!
I hope this has helped out with you situation. Good luck.
mkomarmy 4/4/2005
I assume that your part is toleranced consistent with the ASME Y14.5M 1994 standard. If so, size and form are two seperate issues. Lets look at the size (diameter) tolerance first. If you consider two concentric cylinders, one at the maximum diameter and one at the minimum diameter, each with perfect form, all of the points on the part must lie between those two boundaries. The GD&T standard would recommend a go-nogo functional gauge to insure that no points on the part violate the limits of size.
However, attribute gauges are not conducive for process control or capability, so a number of variable gauging methods can be used to define the size. These can range from two point static methods (calipers/I.D. micrometers) to dynamic multiple point evaluation of the surface at one or more levels (tallyrond, scanning CMM, etc.). The correct one to use depends on the part and the process. If form is an issue, for example lobing with odd numbers of lobes, a two point gauge may not see the variation in the part.
Form errors such as circularity are determined in a number of ways. One way is to look at the maximum and minimim points from the centerline of the inner diameter as established from the least squares circle.
If I were to do this, I would evaluate capability for size based on the least squares circle, as this is one of the most stable algorithms. I would also look at the circularity (specifically the max and min points) to insure the limits of size are not violated.