Runout Tolerance always less than Dimension Tolerance?

[mechEhere]

Does runout tolerance always need to be less than dimension tolerance? If bearing surface is dimensioned 1.0000 (+0.0003 -0.0000), then the runout tolerance can only be 0.0001?

 

[PaulJackson]

No! Size could be perfect, form could be perfect, but location and orientation to the axis created by the datum feature(s) could make the runout imperfect.

Paul

 

[dingy2]

Paul is correct here. There is no relationship between the size tolerance and the runout (circular or total) tolerance.

Dave D.

http://www.qmsi.ca

 

[jerry1423]

There is always a relationship between the size tolerance and geometric tolerance in the way that the size tolerance does hold geometric tolerances to the extent that they fall within the tolerance zone.
In this case it is hard to tell without seeing how the datums are utilized, but I highly doubt that the runout will be that great with the size tolerance held that close.

 

[PaulJackson]

There is always a relationship between the size tolerance and geometric tolerance in the way that the size tolerance does hold geometric tolerances to the extent that they fall within the tolerance zone.

Wrong

Paul

 

[KENAT]

6.7.1.3.4 of ASME Y14.5M-1994 is probably applicable.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[jerry1423]

Show me how I can be wrong.

 

[NomLaser]

I'm pretty sure Paul already did with his example of a part measuring to the dimensional requirements but out of spec due to being off axis, thus not meeting a runout tolerance.

 

[PaulJackson]

Jerry1423,

Position RFS, Runout, Total Runout, Concentricity-pre 94, and Conentricity-post 94 (that is evenly non-uniform) all disregard size while constraining orientation and location.

Paul

 

[dgallup]

Runout does not have to be a refinement of other dimensions (unlike say parallelism). Runout of a diameter can be greater (even much greater) than the diameter tolerance.

 

[TheTick]

Visualise runout on a corkscrew.

 

[PaulJackson]

Kenat,

In the case of bearing journals being controlled for orientation to one another as is suggested in 6.7.1.3.4, I prefer using Position at MMC. Dave will probably fall off his chair hearing me say that in this case an attribute gage is the best solution.

As in figure 51 you would constrain -C- to [pos|0(M)|C-D(M)] and -D- likewise [pos|0(M)|C-D(M)]. The reason is that making the total runout measurements for the journal is frought with problems... typically very small tolerances, traditionally large repeatibility and reproducibility errors in the measurement and typically very good processing that grinds both journals either simultaneously or in the same setup.

If someone were to make such a gage it would probably collect dust because the journals would always fit given sizes targeted away from MMC. If a little slip-fit allowance is needed to accept MMC conditions... that amount can be specified rather than zero... but ... I caution designers not to do that but rather apply the additional allowance to size (assuming that size is controlled statistically) because some stupid quality guy is going to use that tiny allowance as the position tolerance RFS! Therefore I have recommended this same type of control on most of the transmission component journals at my former employer.

Paul

 

[KENAT]

Paul, I just brought it up as it talks about use of runout with other controls which I thought might help jerry 1423.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

 

[PaulJackson]

I know Kenat... I am not criticizing. It is just when questions arise about runout applied to the same features that establish the axis for measuring the runout itself… (not that MechEhere posed the question that way) people are typically trying to constrain orientation of one journal to the other. The situation is quite common in all products that have rotating parts from shavers to airplanes to toys to cars yada yada. I was just offering a method that I have often used to capture the functional relationship in a specification that has little or no wiggle room for quality overseers to demand variables data capability… since the position spec is Zero at MMC.

Variables data capability can actually be done on it but very few would know how to approach it analytically.

Sorry if my response was offensive.

Paul