Mirror Program

Step by step CMM-Manager tutorials
User avatar
medupriest
Moderator
Moderator
Posts: 418
Joined: Fri Mar 13, 2015 11:59 am

Re: Mirror Program

Post by medupriest » Tue Sep 19, 2017 3:04 pm

We are looking into the issue for a long term solution, but I think I may have it, or at lease I think I have a strategy for you to do it.

I am not sure if the translation of the coordinate system is the cause, but I swapped out the lever-align-origin-translate with an RPS (relative point system) alignment. The results will be identical, but this basically allows you to lock individual axes and apply the nominal offset in the same procedure. I have attached the file. Give it a look when you have the opportunity, but be very cautious if you want to execute it as I have no way of proving it out of course.

After I did the RPS alignment, I mirrored the program and then did an align part by flipping the part 180°. It looked okay on my end, but we will have to see.
You do not have the required permissions to view the files attached to this post.

User avatar
medupriest
Moderator
Moderator
Posts: 418
Joined: Fri Mar 13, 2015 11:59 am

Re: Mirror Program

Post by medupriest » Tue Sep 19, 2017 3:43 pm

I see in my project file that sign on X flips rather than Y. It may be an executable program, but the reporting will show the sign flipping on X rather than Y. You can probably just rotate your part coordinate system prior to reporting and that should then give you the results you are looking for.

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Wed Sep 20, 2017 9:20 am

We will give it a shot. Thanks for the input.

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Thu Sep 28, 2017 12:07 pm

So, to update, I successfully mirrored a program yesterday. The only alignment was a CAD alignment at the beginning of the program. I wouldn't normally use CAD alignment ONLY, but because of the nature of this particular project it was good enough.

Today, we set up a fixture for programming (RH) and it functioned as it should. Here's the sequence we followed.

Measure Mode

Align Fixture

Program Mode

Free position alignment followed by DCC alignment
Measure features
RPS alignment
Tolerance features

This worked fairly well with some caveats.

Mirroring

At the END of the program, mirror around ZX. The display shows the mirror plane properly and seems to work well. The program runs, but the nominals for everything are wrong. By wrong I mean rather than the sign on Y being swapped, Y is still positive and Z is negative. Looking at the RPS alignment, the MirrorProgram operation totally borked it. The RPS alignment was then deleted and reconstructed (prior to that step the axes seem to be ok) and suddenly the CAD is 180°on the table upside down and even Sync CAD doesn't corect it. We have used every available tool that we are aware of from update report to reset nominals to transforming elements and it just doesn't want to work.

This is the same behavior that you described here :
I see in my project file that sign on X flips rather than Y. It may be an executable program, but the reporting will show the sign flipping on X rather than Y.
Since you see the same behavior, its seems that this is definitely an issue that shouldn't be.

We can attach this current RH program if you wish. Perhaps we are just missing something.

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Thu Sep 28, 2017 1:09 pm

So, here's an interesting conundrum. Precisely the issue that I've had over and over.
Capture RH.PNG
Capture LH.PNG
Notice how the mirror routine reverses the direction of the Y axis? This is mathematically wrong. 3d CS convention is (AFAIK) "the secondary axis is always CCW from the primary axis when looking down the positive third axis".

This breaks that convention. The dimensions on the left side of the vehicle are now positive, not negative. When working with car positions this is kind of important.
You do not have the required permissions to view the files attached to this post.

ken60
Power User
Power User
Posts: 69
Joined: Fri Mar 21, 2014 12:32 pm

Re: Mirror Program

Post by ken60 » Fri Sep 29, 2017 2:31 pm

Since you see the same behavior, its seems that this is definitely an issue that shouldn't be
Nikon will have to take care of that, to get by I would Reset the cad alignment frame, set it to call out and then Place Cad to match it

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Mon Oct 09, 2017 2:41 pm

I am working with another part, but a similar scenario. This is a free state part check, so I am having to use CAD align as well as Datum Align in order to attain the highest level of accuracy. The RH program went very well. It's 11 points checked on a swoopy warped plastic surface, and checking the same part five times, the results gave me a max range of .04 for X,Y,Z,3D. That's fantastic considering some of the points were 1.8mm from nominal.

The problem happened again at mirroring. Not only did the relative direction of the Y axis change, the nominals for the features used for alignment (a hole and a slot) were different. Not by a little; by 100mm or so in the Y direction. Mirroring around my ZX plane did not change the sign on my Y values (as should be the case), the nominal feature locations were off and when I attempted to CAD align, the CAD reverted back to the RH version.

In order to save time and trouble I am going to rewrite the LH program from scratch, mimicking the working RH and using the actual LH CAD (which matches the RH CAD number for number, BTW).

I would like to know whether these issues (and they are issues as noted above) are being addressed. We rely heavily on mirroring programs, and if this function isn't working properly CMM Manager isn't as useful for us as I hoped it would be.

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Thu Oct 12, 2017 9:32 am

Possibly related issue ... again, I could certainly be missing something.

This is a picture of a mirrored program. The steps I took were to mirror the entire program, switch to measure mode and redo the CAD alignment step to put my cad in the proper position. The measured elements which should follow along did not.

Am I missing one important step?
You do not have the required permissions to view the files attached to this post.

JVanderveen
Regular User
Regular User
Posts: 19
Joined: Mon Jan 18, 2016 9:58 am

Re: Mirror Program

Post by JVanderveen » Fri Nov 10, 2017 12:42 pm

Another program issue today.

It's a small part. CAD alignment, Datum alignment, measure 18 points. That's not much to ask.

After mirroring the program and after the CAD alignment is run, the CAD reverts back to the original un-mirrored state.

This is a known issue. Why does this not just work?

User avatar
medupriest
Moderator
Moderator
Posts: 418
Joined: Fri Mar 13, 2015 11:59 am

Re: Mirror Program

Post by medupriest » Fri Nov 10, 2017 1:52 pm

Which version of the software are you using?

Post Reply