Runout only used for parts that rotate in their assembly? 6

[SMO]

Hi,

In a recent design review, I was told I can't use Runout for a part that doesn't rotate in the assembly.

I applied Runout on the mounting flange, and the coaxial-alignment shaft opposite the flange, of an Axle Pin and referenced the datum axis of the primary shaft.

This part is fixed in it's assembly, but will surely be made on a lathe, and a rotating pulley/bearing will slip fit over the shaft primary shaft.

Since the Axle Pin is completely axisymmetric, with a tapped hole in center, there will be not way of telling it's axial orientation when it is fastened.

In any case, I felt this was a good case for Runout, instead of Perpendicularity, or Profile...

Is there anything wrong with that?





Regards,
SMO (NX10)

 

[SwinnyGG]

Think about gauging- that's the whole point of GD&T.

If this part doesn't rotate in service, how are you going to measure runout on finished parts?

If this part doesn't rotate in service, do you care if the axes of the flange and shaft are precisely aligned?

 

[pylfrm]

SMO,

What the part does in service has no bearing on what tolerances you are allowed to apply per ASME Y14.5-2009. Functionality only becomes important if you're trying to determine what tolerances make sense to apply. I can think of cases where runout makes sense on non-rotating parts, so any blanket statement condemning it seems rather questionable.

If you're looking for input on your specific application, it would probably be helpful to post a drawing or illustration of some sort.


pylfrm

 

[CheckerHater]

No, part doesn't have to rotate. Just ask "experts" to show you where the standard says so.

Imagine part like bushing. You may use runout to control, say, eccentricity of ID and OD. Then you press it in and it never rotates.

The benefit of the runout is that it's compound control that allows you inexpensively measure combined effect of many form/orientation variations at once.

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[AndrewTT]

We have used circular runout on non-rotating parts at my work. We did so b/c we had an inexpensive apparatus that we could give to our suppliers that made the measurement easy. We wanted coaxiality and runout can be used to achieve that. In our case (regular FOS) the circular runout tolerance was larger than the size tolerance so the circular runout control was only yielding coaxiality and rule#1 provided the form control.

 

[dgallup]

Runout only on rotating parts? Ha ha ha ha ha. There are a billion and counting reasons to use runout on parts that don't move at all.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.

 

[btrueblood]

Use runout all the time for seal faces (think o-ring grooves) to control depth and thus o-ring squeeze.

 

[SMO]

All,

Thanks everyone - I was pretty sure I was right, and I still think the part has the correct shape, size, and function to perform a simple Runout inspection.

It's a really small shaft (easy to fit with a collet), with a small flange and, functionally, it has very snug fit with the pulley that slides over it.
There's no way to clock this shaft and once it is fixed, it will in place for approximately 1 million cycles, over 7 years.
I feel you could achieve a nice, and reliable, inspection of the flange, perpendicular to the snug fit pulley/shaft, by using Runout and looking at wobble as a form of perpendicularity.
This should accurately simulate how the flange will sit and how straight it's shaft will be upright.
I care about this because the timing belt (horizontal) will have a lot of tension and I want to minimize the coaxial errors between the pulley (forced by the tension) and my fixed shaft.

Of course, if the manufacturer has better capabilities inspecting this part with a perpendicularity callout, I don't have a problem applying that, with the same allowable tolerances.
But getting mfg. involved is the next step, as this was a preliminary review, and I'm already being forced to accept this markup... Frustrating

Anyway so:
-No such rule,
-mfg. capabilities should be considered

I think I got it, thanks again!




Regards,
SMO (NX10)

 

[SwinnyGG]

Sounds like you found what you needed.

If you can inspect for runout, then by all means it can be applied. My post was just suggesting (my firmly held belief) that it is best practice to specify the the GD&T convention (or combination thereof) which provides the easiest (i.e. most reliable) method for inspection. GD&T callouts which are very difficult or impossible to verify via inspection are useless.

 

[powerhound]

I happen to be in the camp that says runout should only be used for rotating components. I'm not saying that it's illegal to use runout otherwise but I believe there are times when runout is not really the best characteristic to use. I definitely disagree with using runout just because you think the part will be made on a lathe. With runout you don't have the power of using an MMC modifier on the tolerance or an MMB modifier on the datum references. Runout is a more restrictive tolerance than position so if your part meets the runout tolerance, it will meet the equivalent position tolerance. But that just unecessarily makes the

My question is that if you use runout on components that don't rotate, why is that? Why not use position and take advantage of the MMC and MMB modifiers when you can? Is anyone arguing that using position on a non-rotating component is detrimental to its function while using runout isn't?

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[CheckerHater]

Is anyone arguing that using position on a non-rotating component is detrimental to its function while using runout isn't?

Nobody is making absolute statement here. GD&T is supposed to save money. Runout doesn't require functional gauge or CMM. It can be done with general-purpose instruments.
Also you missed OP mentioning that even if part in question is not rotating, the mating part, in fact, IS.

Are you saying that part that functions as shaft or axle should never be checked for runout?

"For every expert there is an equal and opposite expert"
Arthur C. Clarke Profiles of the future

 

[TheTick]

There is no mandate anywhere in the standard that states that runout is only for rotating components. There are only personal prejudices.

Runout is an odd duck in GD&T. It is the only control that is based on a specific measuring technique. It does not directly control geometric form. It only indirectly reflects the sum effect of other form controls.

 

[powerhound]

I didn't mean to imply that someone was making the statement, I just wanted to know if someone felt that using position on a non-rotating part, instead of runout, would prove detrimental to the function.

As far as your last question goes, what features on a shaft or axle are we talking about?

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[powerhound]

It does not directly control geometric form.

When the runout tolerance is smaller than the size tolerance, runout directly controls form.

John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[powerhound]

There is no mandate anywhere in the standard that states that runout is only for rotating components.

I agree.

There are only personal prejudices.

I wouldn't characterize it that way. I use runout when function dictates, not preference, and definitely not process. If features are static relative to each other in function, position is the functional way to go. One could certainly use runout legally but it might be more restrictive than could be allowed if position were used.




John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[Belanger]

My question is that if you use runout on components that don't rotate, why is that?

Because the runout symbols are the only ones that control form, orientation, and location of a circular part, but not size, and there might be a functional reason for that, which still doesn't imply rotation.
Position controls location and orientation, but not form. Circularity controls form but not orientation or location. (And profile would add a size element to the tolerance, which might not be desired.)

Is anyone arguing that using position on a non-rotating component is detrimental to its function while using runout isn't?

I wouldn't say that position is detrimental, but runout packs more punch because it controls form too.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[powerhound]

Because the runout symbols are the only ones that control form, orientation, and location of a circular part, but not size,

Runout only controls form if the tolerance is smaller than the size tolerance of the considered feature.

and there might be a functional reason for that, which still doesn't imply rotation.

I agree with this but you can do this with position and refine for the rest. Again, I'm not saying it's illegal to use runout for non-rotating features. If the functional requirement of non-rotating components are such that the tolerance must be RFS, the datum references MMB, and the size tolerance sufficiently controls form, just use position. If size tolerance is not enough to control form then refine for it. If the function is non-rotating then why should the check be?

runout packs more punch

I agree with this too, but I see it as runout controlling more things at once than may be necessary all the while limiting datum reference to RMB only.






John Acosta, GDTP Senior Level
Manufacturing Engineering Tech

 

[Belanger]

I agree with all of that. Maybe I'm just a glass-half-full kind of guy.

John-Paul Belanger
Certified Sr. GD&T Professional
Geometric Learning Systems

 

[TheTick]

Another advantage of runout is that it's cheap. It takes a lot less resources (time, training, equipment) to check runout than to check position, circularity, cylindricity and concentricity, especially in combination.

 

[dgallup]

Totally agree Tick. Cheap and easy. Even when I specify concentricity, 90% of the time I get a runout measurement. If the runout is within the concentricity spec it's OK. If it's above the spec it might be OK or not.

----------------------------------------

The Help for this program was created in Windows Help format, which depends on a feature that isn't included in this version of Windows.