Profile without FFS

[[To...]]

I seem to recall someone coming up with a work around for measuring this profile without Free Form Surfaces. I believe it involved doing a geometry best fit alignment of 4 planes and 4 cylinder and using the GBF alignment in some kind of evaluation. Can anyone close the gap on this?

Profile of a Surface_No DRF.jpg

[[Ju...]]

Is this what you're looking for?
viewtopic.php?t=9879

[[To...]]

Please sign in to view this quote.

No, but thanks. It involved create a geometry best fit of the 8 features.

[[Da...]]

My take on it would be to create a geometry best fit of the features of interest, then export nominal and actual points of each feature with respect to that best fit. Subtract the actuals from the nominals and calculate the magnitude of each resulting vector. Find the max deviation and multiply by two for your surface profile result.

[[Ja...]]

I think easier way would be to create Geometry best fit for all features with all boxes checked in evaluation constraints dialog box to allow them for best fit in all DOF and then create profile characteristics for each features referencing the GBF and report max result of that.

[[Ro...]]

Ok i just spent the last 2 hours doing nothing at work, coming up with a solution that in my program works fine, but its not fast or easy. I'm also working on an offline seat with err distribution turned on.
I used 10 points, 1-5 in Y and 6-10 in X

I did a quick formula in a result element of moninal - actual, and filtered it thru an absolute command so the result is always +

For the Y points abs(getNominal("Point1").y-getActual("Point1").y)

For the X points abs(getNominal("Point6").x-getActual("Point6").x)

In the final result element i did a Max value, and found the max deviation, and multiplied it by 2 for Profile result.

2*(max( getActual("Ref Result Element1").actual, getActual("Ref Result Element2").actual, getActual("Ref Result Element3").actual, getActual("Ref Result Element4").actual, getActual("Ref Result Element5").actual, getActual("Ref Result Element6").actual, getActual("Ref Result Element7").actual, getActual("Ref Result Element8").actual, getActual("Ref Result Element9").actual, getActual("Ref Result Element10").actual))

I also Grouped those same 10 points into a Free Form and did a Profile that way, the results were the same.

The final result kinda sucks, because even if you mask the Ref Result Elements they still always show up in the report. So i named them "REF" because around here that's the only way to get operators not to care about them.
This was all done without PCM.

dgfhjdgfdgfdgfghdg.PNGdvfmndrfswhuidrferfew.PNGdasdsdvgfghhk.PNGdfsdfskljkljdfeskljdfskljdfkljs.PNG

[[Je...]]

Please sign in to view this quote.

.
Very impressive, Roberto. If it works, you get my vote for measuring hero, although Uncle Carl would shun you.

I strongly identify with the neurodivergent behavior you showcased here. Innovate while procrastinate is my theme.

I often attain my greatest achievements while tip-toeing around the actual assignment. I used to get down on myself for this. However, now I understand that ADHD can be, paradoxically, kryptonite or a superpower. So, I employ tools like timers, deadlines and automated reminders to keep my brain🧠 calibrated. A few noteworthy ADHD brains: Elon Musk, Albert Einstein, Emily Dickinson, Leonardo Da Vinci.

.

[[Ma...]]

I would guess - if your approach would be single points, then you can use "Space point distance" [list=1]

Space point distance

Maximum result - fill all points to get worse deviation

Result element - double value from Maximum result or own formula

[[To...]]

Please sign in to view this quote.

Not sure if you were trying to solve my issue but you're results are tied to the BA. My profile has no DRF

[[Ja...]]

I think I know what you are looking for if it's the space point method. It should work the same with planes and cylinders though. You basically have to measure the features or space points twice. Measure them once relative to the base alignment. Use them to create a 3-D best fit alignment. Copy the features/space points and change the alignment of each one to the 3-D best fit alignment. I remeasured that way. I'm not sure if you can just recall them. Then you can check the distance from nominal of each one, use maximum and minimum result to get the worst case. I have only done it with space points. I am not sure how it would work with planes and cylinders. I think I might remember someone explaining how to it though.

[[Ro...]]

Please sign in to view this quote.

Create a secondary alignment.
use 2 lines (or 4 lines) to align each pocket.
change the alignment of all the points as shown into the secondary alignment and it would basically be datuming to itself.

[[Ke...]]

Please sign in to view this quote.

Not much of a gap, imo.
-Measure your planes & cylinders
-pull them into a Geometry Best Fit
-evaluate each plane & cylinder to the constructed Geometry Best Fit
-report the maximum result of the evaluated features.
The results were almost identical to the Free Form results, when I did a comparison in the same program.