Assistance with perpendicularity measurement

[Ron Rompen]

Reaching out to the rest of the group for some ideas/concepts on how to measure perpendicularity in a thin-walled tube end.

I am currently measuring this feature using a (manual) CMM, and have identified that the R&R is a huge factor - analysis of this problem has indicated that this is inherent in the limitations of our equipment (manual operation vs NC operation), error in centralizing measurement points on the thin wall, etc.

As a temporary solution I have implemented an attribute gauge which is effective, however I would (ideally) like to be able to present variable data, to be able to show levels of improvement in the manufacturing process, rather than just 'better', 'worse', or 'the same'.

I would appreciate any suggestions as to how this could be better evaluated. The depth of the -A- datum cylinder is 25mm, tube OD is 60mm, and wall thickness is 1.2mm.

 

[Miner]

How are you currently constructing your datum A? How many points per circle, how many circles for the cylinder? Using a datum simulator for the cylinder or the part itself?

 

[Ron Rompen]

How are you currently constructing your datum A? How many points per circle, how many circles for the cylinder? Using a datum simulator for the cylinder or the part itself?

Datum cylinder is constructed of 8 points taken on two levels. Datum plane is constructed of 6 points, at approximately 60° separation. it's not an ideal method, but it is what is realistically acheivable given time constraints, operator skill levels, etc.

 

[Miner]

Would it be possible to fixture the part in a datum simulator and use that instead of the part?

 

[Ron Rompen]

Would it be possible to fixture the part in a datum simulator and use that instead of the part?

That is one idea that I've been working on - the design of the part makes it difficult to make a datum simulator that would allow rotation of the part to evaluate the entire B datum surface. But i do appreciate the input.

 

[Welshwizard]

Would the design of the part make it difficult to use a high quality bevel protractor to sample in at least a couple of places?

 

[John Predmore]

Could you use an Optical Comparator or other optical technology? I realize you can only visualize one diametric silhouette per viewing. If the tube is out-of-round, from bending perhaps, you may know where the major and minor lobes are and chose a radial orientation appropriately.

 

[Ron Rompen]

Would the design of the part make it difficult to use a high quality bevel protractor to sample in at least a couple of places?

The size of the part and the part design prevents me from getting an accurate measurement with that method. And a precision instrument like that is the LAST thing I want to put out on the floor if I can help it. Thanks for your suggestion though.

 

[Ron Rompen]

Could you use an Optical Comparator or other optical technology? I realize you can only visualize one diametric silhouette per viewing. If the tube is out-of-round, from bending perhaps, you may know where the major and minor lobes are and chose a radial orientation appropriately.

Don't have an Optical Comparator, and with the current sales forecast I don't see being able to get one approved for quite some time unfortunately.

 

[Welshwizard]

Bearing in mind the specification and what you have already stated;

Understanding it's the inverse of the datum structure but what if you placed the component on the pipe face and used a small square to measure the lack of perp around datum A in conjunction with feeler gauges?