hole patterns as datums

[obie18]

I realize ASME Y14.5-1994 allows us to use hole patterns as datums. I am looking for opinions on it's usefullness. Situations when it would be considered best practise and situations in which it would be best to avoid. Pros and Cons.

Thanks

 

[dingy2]

I would suggest that if during assembly there is a pattern of bolt holes and studs on the mating part that are assembled first and then another pattern assembled later. The second pattern really references the first pattern so the first pattern could become a datum if it met a positional tolerance requirement of some sort.

Hope this helps.

Dave D.

http://www.qmsi.ca

 

[CheckerRon]

I just finished a part which was a standoff post with a rectangular base, having a 4 hole pattern, a cross shaped post section in between and the top mounting flange with its' own 4 hole pattern which is on a compound angle relative to the base plate.

I wanted to locate and profile the top plate relative to the base, and the best way seemed to be to make the 4 hole base pattern a datum.

This permits fixing base with at least 2 virtual size pins, and projecting up the center axes of the datum bolt pattern to locate and pivot the top angles in the 2 planes to a fixed basic height for the profile. Then I datumed the top bolt pattern relative to the top flange itself.

It seemed the best way to dimension this nasty part.

 

[PaulJackson]

The function of the pattern is the most important critera for its selection as a datum feature.

For example, some vehicle's wheels attach to their mating hub assembly with just a pattern holes and bolts...while others use a center pilot diameter in addition to the bolt pattern.

According to the function the pattern without the pilot would serve as the secondary datum feature, naturally the mounting surface would be primary, a tertiary datum feature designation would not be necessary since the pattern constrains all remaining degrees of freedom.

If the pilot diameter's clearance/interference with the hub controls the coaxiality between the hub and the wheel tighter than the bolt pattern then the pilot would be declared secondary and the bolt pattern tertiary.

The selections should always address function before any other consideration.

Paul

 

[ringman]

Paul,

Interesting situation you have presented here. This would appear to be a condition where you have a 'double fixed fastener' tolerance consideration. The studs would be fixed and the what seems to be normal taper on the bottom of the nuts would also result in a fixed location.

I would be curious to see an example of just how this might be handled. "VELLY INTRESTIN"

 

[PaulJackson]

obie18,
I hope you can see the logic behind function driving the choice for a pattern of features as a datum feature.

ringman,
The ole double fixed fastener dilemma eh!...If there ever was an ole one? That wasn't what I intended to portray in the example and I am reluctant to help this thread morph into a discussion on the functional designation of "variable limit tolerance modifiers" MMC, LMC... but suffice it to say that neither in this case would be a functional designation for the features that determine rigid coaxiality. For the assembly clearance of the wheel hole pattern's smallest cone clearance ID... I suppose you could look at it as a FIXED FASTENER situation but If we are talking function |B| should be defined as the four/five/eight/whatever conical surfaces that react with the conical ends of the nuts that react with the conical taper of the threads of the fixed studs... Moreover if the wheel's tire bead mating surfaces were controlled for location and orientation with anything other than runout then |B| should be designated "RFS"... In the event the pilot ID to hub OD clearance trumphs the pattern itself's position tolerance... the example is one where the secondary should functionally be designated RFS!

The point is... function always trumphs assembly stack considerations. There is one sure fire way to determine whether "variable limit tolerance modifiers" MMC, LMC should be designated and that is to ask yourself this question; Does function deteriorate as the features are permitted to deviate from basic location? If yes RFS... If no then a modifier MMC or LMC is appropriate.

I am totally opposed to attribute gauges being reffered to as "functional gages" when the modifiers have not been selected according to function. I am although tolerant of non-functional selections when circumstances... such as mandated 100% critical characteristic inspection is obligated.

Paul

 

[ringman]

Paul,

Could you possibly be more explicit with a sketch of perhaps 4 holes, on a dia with a clearance hole for the stud and a taper for the nut and illustrate what you have stated?

Relating the feature to the required datum features with modifiers etc.

 

[PaulJackson]

ringman,

Do you really need a picture of this? Haven't you ever changed a tire? You loosen the lug nuts, jack up the car, and remove the wheel and tire assy, align the hole pattern of the spare with the studs and push it against the hub/brake assy ( getting it on the hub... is a classic fixed fastener-clearance hole scenario), then... while restraining the tire against the hub you finger tighten the cone side of the lug nuts into the chamfered holes of the wheel.

Interference pattern fit,

If there isn't is a hub pilot that mates with the ID of the wheel then any position error in the hub's stud pattern and the wheel's chamfered hole pattern will result in non-uniform contact between the chamfered nuts and the chamfered holes of the wheel. When tightened, however studs bend... chemfered holes deform... and there is some sort of equlibrium established to force the centers of two errant patterns into a locked compromise.

If there are pilots then the entire system is forced into that locked compromise.

Clearance pattern fit,

If there are not chamfers on the nuts or wheel holes but only the pilot diameters (common sometimes with truck lug patterns) then the clearance between the pilot diameters (with gravity) only serves to increase errors in coaxiality and insures some margim of mis-alignment.

If there are geometric controls (other than Runout or concentricity) that reference either the pattern or pilot as a secondary datum feature it should be designated functionally RFS.

Paul

 

[ringman]

I thought I was asking for a sketch of the geo controls that you described to get the wheel produced. I guess I missed my point.

 

[NeoTV]

How would you use a hole pattern as a datum in an assembly, when this pattern comes from a purchased part? Is it really neccesary to have a feature control frame in the hole pattern?

 

[CheckerRon]

msaldamando:
Hole patterns in purchased parts are usually a problem. You commmonly get a catalog picture with betweeen hole dimensions, often with no tolerances or some ± tols,and a hole size which may or may not have a size tolerance.

Thus I don't see how I trust using such as a hole pattern datum.
We usually match our mating hole patterns and positional tolerances with sufficient clearances to allow for variations in the purchased part. If we get ± tols off the purchase part data, we will convert those to position. When we get a real part or two in hand, we will measure it to see how close it is to the vendor data, and sometimes adjust tolerances accordingly.

 

[KENAT]

Ah, the joy of trying to create working mating hole patterns to off the shelf parts with incomplete data sheets. Our current contract checker is driving himself crazy with it.

KENAT, probably the least qualified checker you'll ever meet...

 

[ewh]

Congrats on getting a checker (even if he is contract)!

Believe it if you need it or leave it if you dare. - [small]Robert Hunter[/small]

 

[KENAT]

Thanks, he's upset at least one person already, but that's not necessarily a bad thing! Also, I'm still stuck doing checking (no offence to those that actually enjoy it), he's just in to help with the outsourcing for a few months, year max probably.

KENAT, probably the least qualified checker you'll ever meet...

 

[CheckerRon]

KENAT: If your new checker has upset an engineer or two in your place, I'm not suprised.
Chances are he has told them to do something they should be doing but don't want to do.

I'm confused. I thought I was responding to msaldomando (above) who turned into NeoTV.

 

[NeoTV]

Sorry checkerron, I just changed my handle,
Your answer was helpful, but my question was because I am welding a tube to a purchased valve that has a 6 studs pattern, so I want to use the pattern as datum to assure the tube is perpendicular and in the position I need it to assmble. The parts have to assemble in the pattern and the tube goes into a hole.

 

[CheckerRon]

NeoTV: In that case, one could use the hole pattern as a datum---probably secondary--with the flange face as primary to control perpendicularity of the tube.