GD&T machined octagon centered on shaft 6

[MechTech2000]

Lets see if you guys can figure this one out,
I'm correcting old designs and drawings for my company, and I'm trying to clearly indicate the concentricity of an octagon shape on a shaft.
This octagon is to be press fit into a pinion (68mm h6).
Datum [A] is the center of the shaft.
The ground cylinders that will mate with bearings are: '[CON|Ø0.030|A]'.

I've looked at using basic dimensions: [3x45.00°] & [SYM|0.000(M)|A] except that symetry does not allow the of MMC.

Do I have to use profile tolerance for this application?

 

[dingy2]

MechNorth - I agree with you that concentricity (won't get into symmetry here) should not be used should not be used on a drawing. One should use positional or circular runout.

Most people think that concentricity means centre to centre distance between 2 coaxial features but it doesn't. They also get messed up in the interpretation of the diametrical tolerance zone. This is one of 2 areas where radially tolerances apply.

I am a bit confused when you state a concentricity misconception is that it "only applies to a nominally circular surface". Are you stating that it can apply on flats on a hexigon? Please show me an example in the standard (I have also studied the standard) where concentricty has applied to anything but a round feature of size? Figure 5-57? 5-54?

If concentricity should be replaced with circular runout (you stated total runout), then could circular runout apply to anything other than round features of size that share the same axis?

Dave D (your northern neighbour)

 

[MechNorth]

Powerhound, you are correct, concentricity controls the derived centerline only, and the actual surface must be controlled otherwise. If the surface is circular, Runout/Total Runout is the best solution. I would use Total Runout for a rotary balance situation, it's a lot easier to verify than concentricity.

Dave, the standard is an imperfect and incomplete document. There are a number of sections that illustrate only a couple of the possible situations that are described in the text, and you often have to extend the application described into new situations. Flats on a hex are one such situation. The key here is the radially opposed points. Check the sketches I posted on my site.

http://www.profileservices.ca/files/tidbits/tidbits_missc.html

One problem that I have with using concentricity for this application is that you would need the stylus/probe to be inline with the diametral line to get the correct reading, rather than normal to the surface ... as a result, the radius of the tip of the probe/stylus increases the error of the measurements.

By the way, where are you located, northern neighbour?

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[dingy2]

MechNorth - I think that you meant "diametrically opposed" and I hope that you don't mind but I disagree with your thought. I am using the standard and there is nothing that would indicate anything other than a round feature of size anywhere. Let me know of anywhere in the standard where it reflects, states or implies your thought.

You are saying that one could use circular runout on flats on a hexagon?? Sorry but I absolutely disagree with that and I hope the Designers here never place that on a drawing. Profile of a Surface, right on, but concentricity or circular runout - no way.

I will let you have the last word on this subject.

 

[MechNorth]

Hi Dave,
Rereading my posts, I haven't said to use total runout on the flat of a hex..."If the surface is circular, Runout/Total Runout is the best solution." Perhaps I should have said cylindrical instead of circular. I fully agree with you that you can't use runout/total runout on the flats of a polygonal object. Also, the standard itself uses diametrically opposed and radially opposed interchangeably {...(or correspondingly-located elements of two or more radially-disposed features)...}, see above post.

Some progressive uses of GD&T require the extension of principles from the standard into new applications. The standard doesn't indicate the use of three-level composite surface profiles, but does illustrate a two-level usage...does that mean you can't use 3-levels? No, and it is a very useful control.

The standard does not indicate anywhere in the definition of concentricity that it is only applicable to a Feature of Size; it speaks only of diametrically/radially opposed elements. The standard is quite consistent in indicating a Feature of Size where that is the intended limitation but it does not specify a FOS here. To flip the table, why do you feel that it only applies to FOS?

I would never use concentricity or symmetry, much less in this application. I rather like Profile of a Surface...it's robust and inclusive of other controls.

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[dingy2]

MechNorth - I stated "round feature of size", not "feature of size" and I should have said that the part must be small enough to be placed in a divider head or chuck.

The standard states that "runouts" are recommended in place of concentricity. One would never place a total or circular runout on a hexagon shape or any other shape than a cylindrical feature of size. It just does not happen.

I hope that the Designers here do NOT place concentricity on anything else but a round feature of size. Better yet, don't use it all. Roundness is a factor in concentricity and trying to find the median point is rather difficult. Circular runout should be used in place of conentricity. It is easy and we all can understand it.

Shop Floor - If you see concentricty with a diametrical tolerance zone - tip - just perform a ciruclar runout using the value in the FCF as the circular runout criteria. One does not have to divide the value by 2 and all that stuff. Circular runout does take into consideration the roundness and also off-centre condition.

 

[powerhound]

Wow, what a direction this thread has taken. Although it hasn't been hijacked, it has led to a productive discussion on the application of misunderstood callouts. I've almost abandoned my contention of using concentricity but I'm still hung up on the idea that using diametrically opposed elements to establish a centerline might be a good idea. I know in practice, most people who understand concentricity rarely use it, but that needn't be the rule across the board. Maybe most people avoid it just because they're used to avoiding it. The callout has to be good for something besides a point to argue against using. Just as concentricity is usually reserved for high speed applications, maybe the other reservation is to center a polygon on a round shaft. It's just a thought.

Powerhound
Production Supervisor
Inventor 11
Mastercam X
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[MechNorth]

Dave, I got burned by omitting the first part of the 5.12.1 italicized paragraph. My apologies. Here is the relevant material from that paragraph; "For example, a nominally cylindrical surface of revolution may be bowed or out of round in addition to being displaced from its datum axis. In such circumstances, finding the median points of the feature may entail a time-consuming analysis of surface variations. Therefore, unless there is a definite need for the control of the feature's median points, it is recommended that a control be specified in terms of a runout tolerance or a position tolerance."

The passage is using a cylindrical feature as an illustration, and bases the replacement of concentricity with runout on it being a cylindrical feature.

Powerhound ... DON'T DO IT!!! Concentricity, if nothing else, is a very difficult control to verify. Go towards the light!

Jim Sykes, P.Eng, GDTP-S
Profile Services
CAD-Documentation-GD&T-Product Development

http://www.profileservices.ca

 

[powerhound]

I...I...I feel faint...must get air...a

http://wwww,

forget it. Adios concentricity...I know you're good for something, but what that "something" is, I guess I'll never know.
Okay Dave, I've been clean and sober from concentricity for a full 5 minutes and I already feel better. What was I thinking?

Powerhound
Production Supervisor
Inventor 11
Mastercam X
Smartcam 11.1
SSG, U.S. Army
Taji, Iraq OIF II

 

[ewh]

I think that both MechNorth and dingy2 deserve a star for delving more deeply into this subject.

 

[kbro151]

I believe I have a use for concentricity.

We make inspection gauges for holes drilled in multiple parts. The holes are tightly toleranced and must be concentric. The gauges are frequently stepped to different diameters. The different diameters must be cylindric and concentric. We check runout, but the runout of any individual diameter alone is not enough to ensure the accuracy of the hole we are checking. We normally inspect these manually, not on a cmm.

Responses and comments welcome.

 

[ewh]

The only use I have had for concentricity is similar, used on the bore of a servo-valve (several separate surfaces).

 

[dingy2]

The stepped gauges reflect a positional tolerance. One cannot have an attribute gauge for concentricity.

For the gauge to be used, the positional tolerance must have MMC and also the reference datum must also be in MMC.

Dave D.

PS - I am now finished on concentricity for good!!!!

 

[namdac]

I agree with ewh in the fact that MechNorth and dingy2 deserve a star for thier indepth input. Bravo!! The rest need to read the GDT Manual or take a class. From what I read a class is in order. Check with your local Community Colleges should carry the class. If not let me know and I will give you the ssbn# and you can check Ebay for the Manual.

This was going good until the train left the track.

Regards,
Namdac

 

[kbro151]

I don't think I effectively conveyed what I meant. The stepped gauge is what we manufacture. It is the form of a pin with a handle, with multiple diameters. The pin must be inserted into the holes in the mating parts. This pin verifies that the holes are drilled to spec. The cylinders in the gauge pin need to lie on the same axis. This is a GO/NO GO gauge to detemine the positional tolerance of the actual holes in the mating parts. A NO GO result using this particular gauge could result from either a positional error of one of the holes relative to the other or a hole diameter that is too small. This gauge is next to useless as a check of hole size, since the only thing it can concretely tell you about hole size is if the smallest hole is too small.

Dave, if I were to use a positional tolerance on two holes which have the same nominal center, wouldn't I be defining concentricity?

In summary, this gauge does not measure concentricity. It must be machined such that the different working diameters on the gauge are concentric.

 

[kbro151]

In my previous post, where I said "This is a GO/NO GO gauge to detemine the positional tolerance...", please amend that to "This is a GO/NO GO gauge to assess the relative position..."

 

[dingy2]

Ok - I am back on this one.

Some people call these types of gauges "concentricity gauges" but they are really positional gauges.

Tolerance on gauges is 10% of the part tolerance and to check the gauge, one should apply a circular runout on the steps relative to the datum.

For sure, I am out of here!!!

 

[kbro151]

"one should apply a circular runout on the steps relative to the datum."

Runout relative to a datum axis would yield concentricity, if the profile is circular, right?

I was interpreting most of this thread as calling for runout instead of concentricity, but runout by itself only gives you deviation from a circular profile, correct?