Nereth1
Mechanical
- Feb 2, 2014
- 136
Hi everyone,
I have a reasonably long fabricated shaft with a bearing mounted to each end. The critical design constraint being that these mounts are aligned roughly coaxially (Ø1mm) and quite accurately parallel (2 minutes of arc). The mounts are welded on to the shaft so the welding distortion is likely to pull them out of good alignment right at the joint before the bearing despite the rest of the shaft being straight.
I'm new to GD&T (first time using it on a real project is this project) and no-one else at my company uses it, I'm basically teaching myself straight off the internet+ASME Y14.5, so I am not very confident in my chosen method.
What I have chosen to do is set one bearing mount axis as a reference datum, and then hold the other one coaxial to 1mm and parallel to Ø0.2mm, but this just doesn't feel very robust - I'm not too happy about measuring angularity using a distance instead of the actual angle, for one, if we modify the length of the whole welded spigot for some reason that tolerance needs to be changed, and it will be different on every slightly varying product in the range.
Based on my reading last night, there are a few other ways I can think of and they all might not be valid:
1) Use a parallelism tolerance that grows with distance, i.e. tolerance it Ø0.2/150. I'm not sure that method works the way I think it does. I'm not sure if it means "entire surface must lie within 0.2*(length/150) of a line parallel to the reference", which wouldn't work because subsections of a length longer than 150 could easily be more angled than it implies, or if it means "every 150mm long section of surface can have maximum 0.2 diametrical deviation from parallel), which would work as it holds every subset of the length within angular limits.
2) The standard describes making a conical tolerance zone by enforcing a wide diametrical tolerance on one surface and a narrow one on the other, and the reader interpolates. But that feels just as if not more awkward than what I'm doing currently.
3) I can enforce parallelism on only the bearing mount surface itself using a chain line to delineate it, rather than the whole spigot, but if we change the bearing spec to a bearing of different width we have the same issue with the tolerance needing to be changed.
I've never seen any reference to it yet in the standard or any of my reading (although I'm not done going through either) but wouldn't it be just great to be able to specify an angle in a geometric tolerance instead of a distance? Would completely solve this if I could just have a tolerance that reads "parallel, Ø0.033°, to reference axis A", but I don't think this is supported?
Thanks for your help and advice.
I have a reasonably long fabricated shaft with a bearing mounted to each end. The critical design constraint being that these mounts are aligned roughly coaxially (Ø1mm) and quite accurately parallel (2 minutes of arc). The mounts are welded on to the shaft so the welding distortion is likely to pull them out of good alignment right at the joint before the bearing despite the rest of the shaft being straight.
I'm new to GD&T (first time using it on a real project is this project) and no-one else at my company uses it, I'm basically teaching myself straight off the internet+ASME Y14.5, so I am not very confident in my chosen method.
What I have chosen to do is set one bearing mount axis as a reference datum, and then hold the other one coaxial to 1mm and parallel to Ø0.2mm, but this just doesn't feel very robust - I'm not too happy about measuring angularity using a distance instead of the actual angle, for one, if we modify the length of the whole welded spigot for some reason that tolerance needs to be changed, and it will be different on every slightly varying product in the range.
Based on my reading last night, there are a few other ways I can think of and they all might not be valid:
1) Use a parallelism tolerance that grows with distance, i.e. tolerance it Ø0.2/150. I'm not sure that method works the way I think it does. I'm not sure if it means "entire surface must lie within 0.2*(length/150) of a line parallel to the reference", which wouldn't work because subsections of a length longer than 150 could easily be more angled than it implies, or if it means "every 150mm long section of surface can have maximum 0.2 diametrical deviation from parallel), which would work as it holds every subset of the length within angular limits.
2) The standard describes making a conical tolerance zone by enforcing a wide diametrical tolerance on one surface and a narrow one on the other, and the reader interpolates. But that feels just as if not more awkward than what I'm doing currently.
3) I can enforce parallelism on only the bearing mount surface itself using a chain line to delineate it, rather than the whole spigot, but if we change the bearing spec to a bearing of different width we have the same issue with the tolerance needing to be changed.
I've never seen any reference to it yet in the standard or any of my reading (although I'm not done going through either) but wouldn't it be just great to be able to specify an angle in a geometric tolerance instead of a distance? Would completely solve this if I could just have a tolerance that reads "parallel, Ø0.033°, to reference axis A", but I don't think this is supported?
Thanks for your help and advice.