PR 260020 Random failure of Roundness Plots....

huangyhg

Random failure of Roundness Plots....

<< Tim Wernicke  --  02/23/09  16:33:41>>
Random failure of Roundness Plots.

Steps to reproduce:
Take a look at P6_LSQ_DIA_S4 in the Text_Only Report Window.  Note that the plot shows a solid (thick) green line and a measured result of 0.000.  When looking at the raw data which is a Constrcuted Circle from Basic Circle Scan there appears to be variation.  

This does not fail 100% of the time, but when it does their only recourse is to re-run the part program for (4) hours!
<<END>>

<< Tim Wernicke  --  03/03/09  12:12:16>>
We had a conference call with the customer last week.  They were not happy with the answer here but seem to have moved on.
<<END>>

<< Changes made by Tim Wernicke -- 03/03/09  12:12:19>>
Status:  RESOLVED to CLOSED
<<END>>

<< David Petrizze  --  03/02/09  14:37:28>>
Reviewed.
<<END>>

<< Changes made by David Petrizze -- 03/02/09  14:37:34>>
Action:  David Petrizze to Tim Wernicke, Status:  REVIEW to RESOLVED
<<END>>

<< Changes made by David Petrizze -- 02/26/09  08:12:48>>
Action:  Don Turcotte to David Petrizze, Status:  OPEN to REVIEW
<<END>>

<< Don Turcotte  --  02/25/09  08:35:47>>
Tim responded that he "...will discuss this with the customer".
<<END>>

<< Don Turcotte  --  02/24/09  16:40:41>>
Tim expressed the following concern so I took another look.

"But, did you notice that the green line in its original state appeared to be very thick compared to the rest.  That seemed a bit odd.  Why are other plots with small deviations showing thin lines and this one seems much heavier or solid?
"


I replied with the following e-mail:

"Tim,



If I simply load the program in V432 (or V44B) and display the report, the ANALYSISVIEW command reports as in your jpg with the diminfo showing MS of 0.  If I F9 on the ANALYSISVIEW command, the plot looks correct (not a thick green circle) and the diminfo will show an MS of 0.00029.    Since the ANALYSISVIEW command serializes the bitmap picture and does not recreate it when the program is loaded, I suspected that the ANALYSISVIEW command was created in a different version of PC-DMIS.   I tried loading the program in V42 MR3, but got a serialization error.  So I loaded in V432 and did File SaveAs to V42.  When I load this program in V42 MR3 and F9 on the ANALYSISVIEW command, I get the circularity plot with the thick green circle and the diminfo shows an MS of 0 as in your jpg.  So I can only conclude that this program came originally from V42 and some changes were made in V432 so it was saved as a V432 program.  If you don’t re-execute the ANALYSISVIEW commands, then you are simply copying the bad V42 bitmaps over to V432.  Executing the ANALYSISVIEW commands in V432 (or V44B) or recreating them seems to work correctly.
"
<<END>>

<< Don Turcotte  --  02/24/09  13:43:13>>
Sent the following e-mail to Tim:

"Tim,

I loaded the program, then unmarked all, then marked P6_CIRC_S4 (the circularity FCF) and the ANALYSISVIEW command following the FCF.  I changed the OUTPUT= on the FCF command to REPORT and refreshed the report window.  The MEAS value of the FCF is 0.00029 whereas the MS value of the diminfo in the analysis view is 0.0000 as in your jpg (seems odd that the FCF reports to 5 places but the diminfo reported to only 4 places).  I then placed the edit window cursor on P6_CIRC_S4 and executed from the cursor (Ctrl-U).  I got the following in the report window.  Both the FCF and the diminfo in the analysisview report to 5 places and both MEAS in the FCF and MS in the diminfo are 0.00029.
"
<<END>>