Iterative alignments (again)

I'm guessing it's because the vectors from points taken on a blade would probably violate the iterative alignment rules: https://docs.hexagonmi.com/pcdmis/2023.1/en/helpcenter/index.htm?rhcsh=1&rhnewwnd=0#t=mergedProjects%2Fcore%2F18_alignment_topics%2FIterative_Alignment_Rules.htm

• The normal vectors for the first set of elements must be approximately parallel.
  
• The features in the ROTATE set must have normal vectors that are approximately perpendicular to the vectors of the features in the LEVEL set.
  
• The features in the ORIGIN set must have a normal vector that is approximately perpendicular to both the vectors from the LEVEL set and the ROTATE set.
  

For more complex parts (like blades) I've found that I usually need to code my own version of an iterative alignment using IF/GOTO commands and labels. I've attached a document that explains the process.

Thanks Neil, I’ll have a look and give it a try tomorrow.

I like this method. A real example might be more understandable .

Nope, sorry, don’t follow! I have more questions than answers now! No offence, merely a reflection of my intelligence, or distinct lack of!
1) Step 2 - Saving an alignment before I’ve even probed anything? Or should I have probed something?
2) Step 3 - Creat generic points, if I do them at the extreme ends of my part, one will be a nice flat land, the other the aerofoil tip? Alternatively, what I am actually interested in is two points on the dovetail, could I use these, as they simulate how the part sits in the respective mating part? I’m been told these points MUST be 11mm apart, so could I iterate to achieve this?
3) Step 4 - Measure ALL features. The datum targets? Or the 2 generic points at the extreme of the part? If it’s the datum targets, I don’t understand why this won’t just fail to iterate again?

I’ll be in the corner crying!

Sorry for the delay in responding but I'm currently out for Easter vacation.

1) The reason you save the alignment before entering the iterative loop is to create the file, the external then alignment gets recalled at the start and updated at the end of each iteration. Saving the alignment before iterating prevents PC-DMIS from updating THEO values with each iteration because it does not "think" the alignment has changed.

2) In my example, the generic points are used to check how stable the alignment is. The further apart the points are, the more stable your final alignment will be. It's the same principle as clocking along a straight-edge to check if it's level. The bigger the distance you clock across, the more accurate your assessment. Similarly, the further apart your generic points are, the more accurate your final alignment will be. As for your "11mm apart points", it's impossible for me to say without seeing a drawing and knowing exactly what is required. My gut feeling is that the generic points will yield a more stable, accurate alignment from which you can then create your "11mm apart" points.

3) The generic points are not supposed to be measured, you recall your external alignment and create the generic points at the start of the iteration. You then measure all of the features and perform all of the steps required to create your alignment and update the external alignment with this current iteration, before dimensioning the generic points to check how much of an effect the current iteration of your alignment had on them. If the deviation of the generic points is more than your allowable tolerance, you loop back and perform another iteration. If the deviation is within your allowable tolerance then the part is successfully aligned.

No worries Neil, hope you had a good break?
Just to update! The design team didn’t want us to use the Datum targets after all, wanted us to use a 3-2-1 alignment on some nice, lovely perpendicular planes!
The 11mm apart requirement remained though, I can’t share the drawing unfortunately, it’s a conical shaped dovetail, at a fixed distance from the edge/datum we scan down the dovetail face on both sides, out of those points we have to fix somehow a distance of 11mm across the faces, this distance is apparently the exact distance the dovetail locates into the female dovetail, appreciate this is tricky to understand without a drawing!