Simple Re-measure Loop Question

[pjleitner]

I am new to PC-DMIS and need a bit of help with what I think is a very easy loop.

I want the operator to measure the part. If any dimension is out of tolerance, I want to prompt the operator to remeasure the part (from the beginning) 1 additional time. If the part is still out, I want to prompt the operator to tag the part with the dimension that is out (display the dimension that is out) and end the program. I have attached file with a flow chart of what I am trying to accomplish.

[Matthew D. Hoedeman]

Well, first of all, is this a DCC (computer controlled) machine? If so, you should be doing your measurements in DCC mode, not manual. The only thing done in manual should be the manual alignment, FOLLOWED by a DCC alignment. So, if this IS a DCC machine AND you are doing the alignment in the best, most accurate way (DCC) then re-measuring the part is not going to fix an out of tolerance dimension, unless it is SO CLOSE to the in/out border that the machine/hardware variation could make it go either way. In any case, DCC is the way to go, no need to remeasure, just measure and report.

For the second part, you CAN set Pcdmis to display JUST the out of tolerance dimensions in the REPORT window (not sure about doing it in the graphics window) BUT, if there are not too many features, you can use DIMINFO in the program and this will put various dimension values on the screen, in labels, where the checks are made at. You can make them change color (mine are green if in tolerace and red if out) and it makes for a quick, easy 'see' as to what is in or out. Move the labels and a leader-line will connect the box to the feature so you can spread them out. So, you can get a screen dump of the cad data along with the features measured WITH the minimum dimension info you need for in/out and it makes a nice little picture to put with the part.

[craiger_ny]

Sounds like a good one for John King(of code)ston. He just learned how to do this today. You'll just need to make a counter to keep track of how many times it is out of tolerance.

Craig

I could do this too if you'd like. Is your program already set up with a loop or is it your intention to measure each part as it's own execution routine? You might want to throw a comment up telling the operator to clean and refixture the part because that is the most common reason for out stuff to be found in. Like Matt said it is very likely you will get the same or similar results. I however am not opposed to measuring manually.

[John Kingston]

This link should give you examples of how to do this. Craiger_NY, aka MasterCoder, showed a number of different ways to do this.

http://www.pcdmisforum.com/showthread.php?t=2645

[pjleitner]

This part is being measured on a manual machine that is susceptible to operator error with miss loads and other issues. That's why I want to do one addition measurement from the beginning if the part is out of tolerance.

No loop in the current program. Sill in the developing phase.

[Matthew D. Hoedeman]

Well, that makes me feel better, I know now that you aren't just wasting time re-checking for the he11 of it, trying to get it in for whatever reason. Manual machines really suck.

As for the DIMINFO and the picture showing IN/OUT, this is how you can get it set up:

1) EDIT ~ GRAPHICS DISPLAY WINDOW ~ EDIT DIMENSION COLOR
(I have mine so that all in tolerance is GREEN and any out is RED)
2) EDIT ~ PREFERENCES ~ SETUP (or use F5) and go to the DIMENSION tab, mark the box, FEATURES USE DIMENSION COLORS. This will make the labels on the screen match the colors in the report, and if RED is OUT, then the label will be red as well.
3) Use DIMINFO in the report. INSERT ~ REPORT COMMAND ~ DIMENSION INFO and follow the prompts. You can include a LOT of information (headers, axis, labels, etc.) or you can include just enough critical to give someone a good idea of what they need to do to make it right. When working with sheet metal, I mostly use JUST the "T" axis of the points, whichs gives the (T)otal deviation and I do not include ANYTHING else, so all you get is a number in a box on the screen. All the monkeys here are trained to know what they are seeing, and it makes fixes and adjustments go real fast. As for showing JUST the out of tolerance on the graphics screen, not sure IF it can be done or IF it is worth the effort.
4) EDIT ~ PREFERENCES ~ EDIT WINDOW LAYOUT is where you go so that the REPORT will show ONLY the out of tolerance dimension. It will be under the REPORT tab in the pop-up.

This may sound a lot harder than it is, but it IS a pretty good method for getting information BACK to those who need it quickly.

[tking]

Thanks for the step by step instruction, Matt

TK

[dfiola66]

Always amazes me how so many here take the time to help out. Thanks again!

[craiger_ny]

The more I think about the whole prompting a re-inspect the more involved this is getting. First I think it is more work than you may be willing to get into but I know the frustration you are trying to avoid.

Consider the following scenarios and let me know how you'd like to tackle them.

• The operator encounters an out of tolerance condition. Do you want to continue measuring then prompt a redo at the end of the program or prompt a redo as soon as the condition is encountered?
• If you prompt a redo as soon as the condition is encountered and the condition repeats, do you want the item to be rejected for that condition only or to continue the inspection to detect other possible out of tolerance conditions?
• If you elect to continue the inspection and do a redo only after the entire routine is complete. Will the part be rejected only if the original out conditions repeat? What if diferent out of tolerance conditions surface? Should a redo be prompted for those?
• Upon being prompted to redo shall the entire inspection routine be re-done or only the out of tolerance features? If so what if the condition was encountered not because of the feature being dimensioned but because the datum (or other alignment feature) was not properly cleaned/measured?

I guess what I am getting at is the flow chart you have makes sense when veiwed from the standpoint of measuring and dimensioning a single feature but as soon as you consider all of the scenarios that could be encountered with multiple features it becomes quite complex. Not impossible but very complicated. I'm not entirely sure it would be appropriate to code it. Perhaps work instructions or an SOP of some kind would be more appropriate. As a proponent of getting the operator out of the picture this sugestion is contrary to that but I think careful consideration needs to be given to all of the possibilities of coding it and how involved debugging alone will be. Let us know if you still want to take it this way. I'd be willing to help but as involved as it is I am not willing to present an "out of the box" solution.

Craig

[pjleitner]

Answers to your questions:

1) The operator encounters an out of tolerance condition. Do you want to continue measuring then prompt a redo at the end of the program or prompt a redo as soon as the condition is encountered? Prompt a redo at the end of the program and redo the entire program. The program only measures two elements for distance and surface profile.

2) If you prompt a redo as soon as the condition is encountered and the condition repeats, do you want the item to be rejected for that condition only or to continue the inspection to detect other possible out of tolerance conditions? See above answer.

3) If you elect to continue the inspection and do a redo only after the entire routine is complete. Will the part be rejected only if the original out conditions repeat? It will be rejected for whatever is out after the second run through. What if diferent out of tolerance conditions surface? It will be rejected for the different out of tolerance conditions as well as anything else thats out after the second measure. Should a redo be prompted for those? No.

4)Upon being prompted to redo shall the entire inspection routine be re-done or only the out of tolerance features? No, start from the beginning. The program is short.

5)If so what if the condition was encountered not because of the feature being dimensioned but because the datum (or other alignment feature) was not properly cleaned/measured? See above answer.


This is more complicated then I thought!

Thanks so much for your help and everyone else's help!