The purpose of Looping Base Alignment

[[ng...]]

Just out of curiosity, what is the purpose of looping the Base alignment?
---
And also was is the difference between when running Base Alignment:

Program name with no (CNC)
and
Program name with (CNC)

[[An...]]

See attached.

Iterative_Alignment_1.pdf

[[Cl...]]

Just to follow behind Andreas excellent pictorial explanation I will
add that with each iteration (loop) Calypso "best fits" the alignment
until it reaches the tolerance applied to the loop.

Imagine putting on a new shoe. The shoe is made from a cast of
your foot, but .0005 larger all the way around.You put the shoe
on and wiggle it around until it feels just right.

Does that make sense?

[[Er...]]

Just to clarify, I understand your point Clark, or maybe I don't, and what Im about to write is exactly what you said. But no "best fit" occurs. Plane angles and origin is recalculated within ever iteration. And for every iteration, repetabillity gets higher, and you get closer and closer to previous iteration. Hence the often used valueA break condition.
But there are no algorithm minimizing errors. The cmm is just able to take points more accuratly at the xyz (originating from part bs) you have programmed.

How that makes sense. (Sounded better on swedish in my head.)

[[Ro...]]

I loop my base alignment because my fixtures are generally held on a rail with a single 3/8-16 bolt, or bolted down with 1/4-20 cap screws.
taking the fixture on and off means there is a pretty good chance that there will be upto .020 thou difference in fixture location, and im too lazy to do a manual alignment every time i put a fixture up (we are a job shop, so we deal with up to 20 different part numbers on the Zeiss daily)
The first Loop is the "rough" alignment, once the cmm knows "Where the part is, exactly" the 2nd time thru is repeatable within tenths every time, and i bet my future paychecks on that.

[[Cl...]]

I tried to put it in simple "layman's" terms.

[[De...]]

Starting with the second part of the question first, the difference between CNC alignment and not CNC alignment.

When you run the alignment with base alignment name (CNC), Calypso looks for the part in the location of the last run CNC alignment.

When you run the alignment with base alignment name (with no cnc), Calypso looks for the part in the location of the last MANUAL alignment.

With a CNC alignment, if you moved every part 0.1 in +X direction, Calypso would look for the 6th part at X+0.5 from your original alignment.

With a non CNC alignment, every time you run your part, Calypso reverts back to the manual alignment to determine where to find the part.

Now as to the iterative alignment portion of the question, the answer is pretty simple really, given the fact that the part may be loaded out of position to where Calypso thinks it should be, the probes may contact the part in such a way that the probe tangency calculations are not correct based on the difference of where the part actually is to where Calypso thinks it is.

So first time through Calypso does the math based on where it thinks the part is. The valueA or delta value is the difference between the 'original' base alignment location where Calypso thought the part should be and the 'actual' base alignment location where it just found the part based on its probing of the part.

Now it runs the part again, the second time through it scans the part in relation to where it just found it based on probing the part, since this time is based off scanned data and not where Calypso 'thought' the part should be, it is much more accurate and this time around the valueA or delta value is much smaller.

If necessary the loop will continue until it satisfies the break condition. Unless the part is moving or not repeating for some other reason in which case it may never break the loop.

[[To...]]

Here's how I used to describe it in training classes. Let's say you measure something and the person you're inspecting it for doesn't like the results. What? It COULD happen. Anyway, he asks you to measure it again. This time you get a different answer. He then responds with "Well, which result should I believe?" One of the main reasons that you don't get repeatable results is due to not measuring the features on the same molecules. Extreme? Maybe, but that's where I want you to focus. The main reason you're not measuring on the same molecules is because your Base Alignment is not zeroing in on the same molecules. Sure, your base alignment is being calculated but it is likely a bit different every time you load the part on the machine. The looping process helps zero in on the same molecules of the base alignment which means the rest of the features in the program should also end up measuring on the same molecules, hence the improved repeatability. Looping the alignment is not something you have to do on every part because some parts have features with good form for the base alignment. For example, a part that has been ground or is very flat, you're not likely to get a lot variability for your Spatial Rotation, etc.

Hope this helps.

[[ng...]]

Thank You, very informative. I learn new stuff everyday, never ending. And Thank You for those who replied back without being smarta****.

[[To...]]

Please sign in to view this quote.

❓ ❓ ❓

[[Cl...]]

Must of had a bad experience in the forum at some point?

Sometimes it's hard to interpret the tone of a written response. 🙂

[[ng...]]

Just saying in general. No one in this thread was being one.