hexagon logo

When to update noms?/

I have been through training and I am still confused on how to use this. After I alter an alignment for example it will ask me if I want to update noms. I always click no because from experience I found it overrides previously programmed nominals. On the other hand I updated an existing program to a new datum scheme, the part had an all over profile that checked bad in the new alignment until I literally re-created the entire profile in the program and now I am wondering if clicking "yes" to update nominals of related dimensions after I changed the alignment would have resolved this. 

So when do I click yes or no to "Do you want to update nominals of related...?"

  • If you ever hit yes when modifying alignments it'll move everything below it to match changes to the alignment. Most of the time you probably want to hit "no". There are occasions where hitting "Yes" does help if you need to move everything below the alignment to the updated alignment information. 

    There is instances where I use the "yes" option as a tool for myself. 

    For example, if I need to hit tabs every 90 degrees on the tab features below that is no where near my origin. I will align to the circle, and find the nominal 0, 90, 180, 270 hits on the tabs. Then I will delete the alignment I created to origin to that circle in the middle, and hit YES so that it moves my hits relative to my original alignment origin. It may sound tedious, but it works really well for me, and saves me from trying to figure out math.

    I have used "yes" for nominals update if I have created a few location dimensions with similar nominals/tolerances, and forgot to set the tolerances in the dimension window. It'll update them all for me so I don't need to manually edit them all. I usually say no first so I can make sure that it'll be okay. 

  • Saying "yes" to the Demon is always a risky proposition.
    I try to never let it update the nominals as we use move points, and it will happily change every one of them in your program.  Normally make a copy of the program, then delete everything after the alignment, fix the alignment and perform a "code-ectomy" and just paste the after alignment code back in. 

  • Now that you mention it, I do that as well from time to time. I just don't realize it. I usually hit "no" and always save before I do risky maneuvers like I mentioned above, just in case I screwed up. I have also recalled alignments to keep things from changing downstream if I'm doing changes mid program(bad idea I know) but it has not shot me in the foot yet. Although I have lost lots of progress due to forgetting to save, and when I added a reporting format command PC-DMIS didn't like it and crashed.

    I have also been burned by 3D Best fit alignments backtracking and deleting features from it. Not sure what would cause that.

  • There will probably be many examples given before this thread dies out, but I feel like the short answer to your question is... when the origin changes.  At least that's when I click yes.

    Say you programmed a part in it's entirety.  Then later you decide to insert an alignment in the center of the program.  And let's say the Z origin when you wrote the program was on the top face of a part and the alignment you insert in the center of the program changes that Z origin to the bottom of the part.  All points below that alignment, the Z location is relative to the top face.  So, if you're probing from top to bottom in a line you're counting down in -Z.  When your Z origin changes to the bottom of the part, if you do not click YES, then the points below that alignment are still gonna be -Z counting from the bottom now which will make the probe go below the actual part.  Clicking yes in this instance will change all points below the new alignment to the +Z equivalent so they probe in the correct place.

  • Tagging off of what Hightech said, I often program parts offline and then come on-site to run them or someone else does. Occasionally we will have a part production supplied that has some sort of roadblock that prevents the feature I chose from being in the alignment.

    Recently we had a tank part come in before it was turned +R which means reuse, and it had damage on a feature that the CAD did not show. I setup my basic manual alignment off the top of the turret and it was not level, so I had to choose another plane to call that level back to. In this case I selected a new plane, it was lower than the original choice, and created level off that. It prompted the same above Alignment prompt and I hit yes. After that, you want to ensure your move points/clearance planes are where you want as they will change.

    As for the 2nd prompt about nominal changing, as stated in here it will only be when you want to change the origin of something, otherwise it will not alter the location.

  • Thanks for the info. I think this answer is what I was looking for. Do you know if the point locations imbedded in the cad model are dependent on the current part alignment or if they are pre determined by the origin attached to the model when it was created? I'm also trying to figure out why having a slightly off location origin changes profile despite the fact it relies on cad for nominal info. Feel free to ask more questions if that is not clear. 

    Thanks. 

  • For the first part of the question... if I understand correctly what you are asking, I suppose the answer could be either.  For example, the initial points placed onto a model immediately after importing the model will be relative to the model origin when the model was created.  After you create an alignment and move that origin, all subsequent points are relative to that alignment unless you recall the startup alignment.....

    As for the second part of the question, I don't deal with profile so cannot answer intelligently.  But if profile is like everything else, A rock solid alignment makes for a rock solid program.  If there's error in your alignment, I just almost have to believe that error will be carried to your measurement results and maybe in certain instances even compounded.