Perpendicularity to shoulder/face for closely fitting boss and bore

[Burunduk]

Part A has a 22mm bore boss with tolerance of h6(0/-0.013). It mates with a 22mm H7(+0.021/0) bore on part B. Both parts make face contact on flat surfaces designed perpendicular to the boss/bore. The mating faces are designated as datum features on both parts. Should the boss and the bore be specified perpendicular within 0 at MMC to their respective flat face datums to "protect" perfectly perpendicular Virtual Condition boundaries of 22mm and avoid assembly problems?

 

[chez311]

Burunduk,

Without knowing anything more about your application - perpendicularity applied to both components seems like a good idea. On part A there might be a position control which takes care of your perpendicularity - or if further control is needed, perpendicularity could be added as a refinement.

 

[Burunduk]

chez311, I will clarify:

Part A has a boss (not bore as I wrote by mistake - now corrected). The boss in part A and the bore in part B are secondary datum features in datum reference frames in their parts, so that other features on the parts are controlled for position with reference to them. The boss and bore are controlled for Perpendicularity to the flat primary datum features - the flat faces that mate by face contact.
Both Perpendicularity tolerances in the 2 parts are stand-alone, not refinement of Position.
I'm specifically interested in opinions whether or not those Perpendicularity tolerances should require 0 at MMC, considering the H7/h6 fit on diameter 22mm; which means the same MMC size of 22 for both. My concern is that when both are at MMC, any perpendicularity error of either of them may lead to failure to assemble properly (properly means enough face contact on the primary datum features, which have to be clamped tightly together). Sorry for lack of drawing, but I hope the description can be followed.

 

[chez311]

I'm specifically interested in opinions whether or not those Perpendicularity tolerances should require 0 at MMC, considering the H7/h6 fit on diameter 22mm; which means the same MMC size of 22 for both. My concern is that when both are at MMC, any perpendicularity error of either of them may lead to failure to assemble properly (properly means enough face contact on the primary datum features, which have to be clamped tightly together).

If the flat faces which are mating to each other are both your primary datum features for each respective part, then 0@MMC perpendicularity will prevent the situation you're concerned about. If both parts come in at MMC (22) then they will be allowed zero perpendicularity error (hence the name - 0@MMC). Any perpendicularity error beyond that will only be allowed by a proportional deviation of size away from MMC. The amount of face contact will only be limited by form error of the mating flat faces themselves.

 

[Burunduk]

chez311, thank you! This is what I was thinking also, now with you confirming it I am more confident in 0@MMC.

One issue remains for me though: part A for example, with the boss toleranced h6(0/-0.013), when the cylinder produced at middle of size tolerance, there is only 0.0065 mm Perpendicularity allowed by 0@MMC. With AME diameter closer to MMC, even less. This is considerabely tighter than the tolerances I am used to define on my drawings. Do you consider it reasonable to hold by CNC grinding (both cylindrical and flat surfaces ground)? The height of the boss above the flat shoulder - 5mm.

 

[chez311]

Burunduk,

That is quite a narrow tolerance band, I'll give you that. As far as being able to hold that tight a tolerance with conventional machining methods - I'm not sure, that would be something to discuss with your vendor.

If the application requires it, I'm not sure what else you can do. Your other options would be to either loosen up the allowable size range with 0@MMC (results in looser fit at LMC), allow additional perpendicularity error @MMC (results in a part which would have perpendicularity error at MMC which would have sub-optimal contact) or some combination of the two, along with a shift in the size range so that perpendicularity error at MMC would provide ideal, or something between ideal and sub-optimal, contact of the mating faces.

 

[Burunduk]

Thank you, chez311.

An experienced manufacturing person says holding the tolerance is feasible when both the shoulder and the cylinderical feature are ground in a single set-up.

If the vendor will raise any issues, one of the compromises you proposed is what I will be considering. Thank you again.

 

[chez311]

Absolutely, glad to help