Review - PR 242750 Various GD&T : VC (BONUS) on datum features in TP

huangyhg

Various GD&T : VC (BONUS) on datum features in TP seems very high<< Anthony Griggs  --  08/29/06  10:53:52>>
I see that I now have the ability to analyze my TP with a VC on the datums! That is great.
Looking at the numbers though, it seems to me that the numbers are on the aggressive side. Again, this may be an educational issue. So please clarify.

Here are the issues:

1: General calculation of maximum allowed shift on a hole.

Please go to datum-R- on the report. It measures 0.18713. The size of the feature can be between 0.1873 and 0.1871. I always understood that that leaves me with a bonus of 0.00003 at MMC. It would be 0.00017 at LMC. Why is this seemingly reversed? I checked all other cylinders, and they also seem to have the same issue. Could it be that PC-DMIS thinks it is measuring an OD????

Based on the TP of 0.001, I am allowed a shift of 0.00103.
If my datum were to be SMALLER than the minimum allowed size (let? say it were to be 0.1870), then the feature would be reported as out of tolerance, but still I would report an allowed shift of 0.001-0.0001 = 0.0009?

2: Datum shift on a cylindrical feature.

I also think that since I measure the feature as a cylinder, perpendicularity MUST be taken into effect, so I think that the maximum allowable shift must be lowered by the perpendicularity. That would make the maximum allowable shift on datum-R- 0.00103-0.00001 = 0.00102.

Can you also report the maximum allowable shift, like you report the BONUS? Maybe this is the +TOL plus the bonus in the TP report fields????? Which, I think, does not take perp into effect yet???
Which therefore I think is incorrect, but I may be wrong here.

3: Calculating the gauge pin size for a datum feature that has 2 virtual conditions: one of location and one of orientation. This is the condition that we have everywhere.

Let? go to feature CYLSH7C5 E6.
I think the bonus ought to be 0.00002, the perp is 0.00005 so the maximum allowed shift ought to be 0.00097 (0.001+0.00002-0.00005).

The size of the gauge pins for datum ?- and ?-: If there was NO second virtual condition on them, the size of the gauge pin ought to be 0.1871 (at MMC) ?0.001 or 0.1861.

HOWEVER!!!!!!!!! I have 2 virtual conditions on the datum: Location (TP) AND orientation (perp).

So for location, the gauge pin size ought to be 0.1861. But for orientation, the size would be 0.1871 (at MMC) ?0.0002 (tolerance on perp) = 0.1869.

Per the books, I think that in my case (when have 2 virtual conditions on a datum feature), the orientation takes precedence. (I am not sure but it seems logical to me that the largest one takes precedence, however my book is clear on this; it states that orientation always takes precedence).
So in my case the gauge pin size must be ONLY 0.1869! And not 0.1861.

I suggest that you report the calculated gauge pin size per datum feature.

So now I have a gauge with 2 pins at datum R and S that are only 0.1869. I would first calculate whether both pins can go in the actual holes. In my case, I can see many instances where it would already fail that portion of the fit test. If not, I would report that the 2 datum gauge pins can NOT go into the datum features, therefore TP can NOT be calculated. Or maybe you revert to RFS on these holes at that time, but I would like to know when you do that.
I think reverting to RFS is wrong though. What would be better would be to calculate a new gauge. One that has an RFS pin on the secondary datum and another RFS pin at the tertiary, moved in a way that both pins just can go in. I would report the new position of the tertiary pin and the sizes of both pins. Then calculate TP to this new gauge.
But let? say (let? be optimistic here) that they do fit into the actual datum features and you have some room left to wiggle. Now you can wiggle and try to make the actual TP as small as possible.
Therefore I believe that your reported TPs are way too generous.


Steps to reproduce:
show above

Results:
shown above

Expected Results:
described above
<<END>>

<< Changes made by Neil Kay -- 07/23/08  16:32:32>>
Action:  Paola Pallo to Yanhua Huang
<<END>>
<< Changes made by Neil Kay -- 07/23/08  09:52:24>>
Action:  David Petrizze to Paola Pallo
<<END>>
<< Don Turcotte  --  09/25/06  10:13:09>>
Merged into V42R.
Files inserted to server
------------------------
V42R\DIMENS\DATUMREF.CPP
<<END>>
<< Changes made by Don Turcotte -- 09/25/06  10:13:16>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Tim Wernicke  --  09/25/06  09:01:45>>
Please merge this fix into V42R.
<<END>>
<< Changes made by Tim Wernicke -- 09/25/06  09:01:49>>
Action:  David Petrizze to Don Turcotte, Status:  REVIEW to OPEN, Priority:  Critical to Stop Rel.
<<END>>
<< Don Turcotte  --  09/25/06  09:49:09>>
I have reviewed the XactMeasure algorithm for calculating bonus refinement on a datum.  The smaller of (tp tol, orientation tol) is used for the refinement.  I have, therefore, modified the PC-DMIS refinement calculation to use the smaller of any true position or orientation (perpendicularity, parallelism, angularity) tolerance on the datum.  So in Jan's case where there is a perpendicularity tolerance on the datum which is tighter than the true position tolerance on the datum, the perpendicularity tolerance will take precedence.
Files inserted to server
------------------------
V41B\DIMENS\DATUMREF.CPP
V42\DIMENS\DATUMREF.CPP
V43B\DIMENS\DATUMREF.CPP
<<END>>
<< Changes made by Don Turcotte -- 09/25/06  09:49:19>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>
<< Don Turcotte  --  09/11/06  11:38:45>>
Sent e-mail to Jan :
"Jan,
Relative to your questions/comments below :
1. The cylinders in question are indeed OD.  If you look at the edit window commands, you will see for the cylinders that they are OUT which means outside (or pins).  Also if you F9 on the constructed cylinder, you will see that the OUT radio button is selected.  You can change these to IN to get the MMC calculations that you expect.  For a hole at MMC, the bonus is Measured size ?lower size limit as you noted.  If this calculation results in a negative value, PC-DMIS sets the bonus to 0.0.  If this calculation results in a value which is greater than (upper size limit ?lower size limit), then PC-DMIS sets the bonus to (upper size limit ?lower size limit).  
2. When the considered feature is a cylinder, the PC-DMIS ?itter?always checks both ends of the cylinder axis to verify that they are within a straw whose diameter is the tp tolerance + the size bonus on the considered feature.  With a primary plane datum, the perpendicularity is not affected by the fitting process.  If the considered feature is within its perpendicularity tolerance, then after ?itting?it remains within its perpendicularity tolerance.
3. With regard to a datum with two virtual conditions : one of location and another of orientation : PC-DMIS reports the bonus on the datum at material condition from the bonus due to size + the tp tolerance on the datum + the orientation tolerance on the datum.  This is reported in the Feature Control Frame ?ize?table for the datum feature.  In the ?osition?table, PC-DMIS reports the remaining bonus on the datum (after the fitter has ?iggled?the datums to minimize the deviation on the considered feature).  Consider the case where the datum size is at MMC so the size bonus is 0.0 and there is no location tolerance but there is a perpendicularity tolerance.  The total bonus in this case should be the perpendicularity tol since at perfect orientation the datum should be allowed to shift by this amount.  So it makes sense to add the orientation tolerance to the overall virtual condition bonus.  The PC-DMIS ?itter?is ?iggling?the datums based on available degrees of freedom and checking that the endpoints of the datum cylinders are within their bonus ?traws?and at the same time checking that the endpoints of the considered feature cylinders are also within their bonus ?traws?   If no position of the datums can be found which will bring the considered features within their tolerance (tp tol + size bonus), then the fitting is not done and the true position is reported with the datums at RFS.
Best regards,
Don
"
<<END>>