True position question when dealing with fixtures.

[Jeff Hunt]

When you have a fixture/gage for a sheet metal panel, & you align your program to the fixtures tooling balls or corner, how do you set up to measure the holes for true position in relation to the datums on the part? I ask this because usually the datums are either pinned or clamped, so it is hard to measure them proper. With the surface datum features, shold I just measure as close as I can to the clamp?

Also, how fo I go about referencing my hole to a datum when that datum contains more then 1 feature that is on more then one surface (think of an inner car door where the fixture nets are around the seal surface.)?

[Paul Barriage]

Our "final" alignment after using the tooling balls or fixture edges for reference, uses the pins for rotation & origin (offset of course for car position). Once that is established, we save the alignment. Then start a new program & recall the alignment. At that point, you are STILL aligned to the pins. You can FORCE the nominal dimension for the pins by creating offset points.


864-577-8317

[\v/inston]

you could use an iterative alignment using your datum features or if your 2way, 4way, and A nets check good to the tooling ball alignment then you could just go off that. if -B- and -C- are under the clamps either report them as 0 (because they will be) or tell your customer that you had to check the part constrained on the fixture and were unable to check the datum features because they were under the clamp.

[Matthew D. Hoedeman]

With 17+ years sheet metal, using fixtures, with NO issues from customers, you simply align the fixture, check the fixture locators, if they are wityh fixture tolerance (usually less than +/-0.002" or 0.050mm), then you simply check the part, the part IS already located to ABC with your alignment of the fixture. Make the alignment a program all by itself, save the alignment to a file, the in a new program, recall that alignment and program your part checks.

You do NOT have to make them different programs (align and check) BUT, if you have to make changes to the alignment for WHATEVER reason, there is no chance of you screwing up the check program, eithe rby telling it YES UPDATE all the points, or through some yahoo turning OFF that warning and having it automatically update everything in the program. At that point, you are so far beyond screwed and the LIGHT from screwed will not get to you for 20 or 30 years.

[Jeff Hunt]

I see what you are saying matt, but my question is, how can I reference the datums when I dimension a hole for TP? Are you saying:
1. align to tooling balls
2. check datums on fixture without panel on
3. put panel on & inspect
4. dimension holes referencing the datums checked in step 2

That sounds like a plan to me, but how do you deal with datum A having more then one datum? I.E. A1, A2....Datum A for me usually consists of a series of nets that hold the panel on a plane. Sometimes these nets are on the same surface & angle, sometimes they are not...Should I create a feature set from them?

Sorry for the dumb questions. I am embarrassed to admit that I hav been programming for 13 years & I have been to all of the PCDmis classes, including power users, but I never got a grasp on how to do TP. My company has always used box tolerancing, but I want to try to move away from that. I have noticed that many of the PCDmis classes are not really geared twords sheet metal.

Anyways, I appreciate all of you guyses help (like that?) & I owe you all a beer.

[tking]

Anyways, I appreciate all of you guyses help (like that?) & I owe you all a beer.

Well I didn't help you out at all, but hey, make it a homebrew!

TK

[Matthew D. Hoedeman]

First of all, an easy answer, if your current alignment already matches the TP callout, there is NO NEED to use datums in the TP dimension.

As for the rest, here is a LONG answer. I use a 3 step process:

1) I create an alignment program, manual then DCC, CAD=PART. Once the alignment is done, I check the datums, and it does not matter where they are or what they are. If they check in tolerance, I save the program, both as an ALIGNMENT program and as a FIXTURE CHECK program.

If they do not, I adjust the alignment offset values (I don't use ITERATIVE alignments) until the datums check within tolerance, BEING VERY CAREFUL not to update after changing the alignment. You must also do a CAD=PART after you adjust an alignment (I only adjust the DCC alignment) since Pcdmis does not update this when you change a value in an alignment. I will also open a feature and do a FIND just to make sure that none of the values change, insuring that nothing was updated to the new alignment.

2) Once I have it saved as both a FIXTURE and an ALIGNMENT program, I open the alignment program, and remove all the fixture check data, then save the alignment to an external file.

3) I create a part check program, first recalling the external alignment.

I now have a saved, external alignment for the fixture and I can then make any 'extra' programs I need to do 'spot checks' or 'Gee, I just want to see this' programs or what ever, all recalling the external alignment, WITHOUT the need to align the fixture for every program. I can make FULL CHECK, SPC CHECK, Gage R&R, Trim Development and any other programs I may need, and I do not have to align the fixture more than once.

As for TP, if your TP callout is to ABC and you have the fixture aligned and ALL the ABC datums are in fixture tolerance, then you WILL be doing TP to the ABC datums simply by doing the TP WITHOUT DATUMS, it will automatically use the cirrect alignment to do the TP, and if your current alignment has the ABC datums within tolerance, that is all you need to do. I NEVER use datums for TP because I always have my current alignment equal to the callout of the TP dimension. Pcdmis has a VERY bad habit of 'shifting' and 'slifing' the datums around when using MMC so the values you report are not correct, as far as I am concerned. If your BC datums are getting 0.25mm of bonus due to MMC size, then Pcdmis will SHIFT the alignment as much as it can to give you the smallest deviaiton values for the TP as it can, instead of showing you the bonus for the datums and adding that to the TP tolerance, like it used to do and still should do. Kind of funny story about that too, Pcdmis used to do it by adding the datum bonus to the TP tolerance and we were told by B&S/Wilcox that it was the correct way to report it (and most of us agreed), but then they changed to the current method and once again told us that it was the correct way to do it. So, which is it? Before or After?

ANYWAY, as long as you current alignment matches the TP callout, ther eis NO NEED to use datums when doing the TP dimension, and chances are, you will be better off that way.

Sheet metal fixtures all have a tolerance that they must meet to be acceptable, usually 10% of the part tolerance to a maximum of 0.002" (FORD standard). As long as it checks that way, there is no need to do an iterative alignment on the fixture locators, you can do plane-line-point (or whatever) off the bases of the fixture. IF all those locators, no matter where or waht, then check within tolerance, you are done, no need to mess with anything else.

Of course, there are many other ways of handling the alignment/fixture check/part check. This is the way I have set it up since I can have up to 5 programs for a single part on a single fixture and I do NOT want to have to run the alignment 5 times (What a waste of time!). You can, if you want to, do the manual alignment, the DCC alignment, check the datums, (adjust the alignment if needed), then once the fixture datums check good, simply delete them and program the part.

[Jeff Hunt]

I think I get you matt. The only part that had me loopy was the external alignment thing, but I understand it now.

The way I write my programs is like this. Each part gets 1 "Program". This program includes a menu system at the begining asking you what you would like to run. The program itself is broken up into seperate sections, for Ford, I would have a menu like this:

What program would you like to run?

1) Level 3 & 4 Points
2) Level 2 Points
3) Holes
4) GRR
5) Alignment
6) End

Then the program goes to whatever the label is coorisponding to the users input. Its all run thru IF & GOTO statements.

Thanks for the help though. I finally understand.