hexagon logo
Register Sign in
  • Replies 41 replies
  • Subscribers 17 subscribers
  • Views 984 views
  • Users 0 members are here
Options
Related

What Was Your Mentally Exhausting Program??

cheesiechanman
The reason I ask is because of the program I am working on right now. I have 3 fixtures with 4 parts that I have been working on. All of them have complicated Datums / Poor design. The fixture that holds two parts is the most complex of course. The gage is the only one built to hold the part vertically. It has a hole in the back of the gage to probe the bottom of the part through. So the part sits with the rear surface in the CMM's +x direction. The part sits on the net pads for datum A about 120mm from the back surface of the gage (where the hole is) There are 18 small clips spaced around the outside edge of the part that I need to touch in the -x direction. This means I need to articulate my probe to, at times, an angle of A-115B35. This shortens the distance from the end of the probe to the Z axis column. Most of the time I am less than 2mm from crashing into the gage.

The part also only has one tiny slot to control rotation on one side of the part. So when placing the part on the gage it can rock, roll and slide everywhere. So that means I need to create local alignments on each clip to ensure that they hit correctly which means dancing around that almost crashing point the entire time. And I need to create alignment after alignment to ensure that I am hitting it correctly and then be sure that I recall a fixture alignment before proceeding because if I forget to revert back to the fixture and the next part changes in size/ location/ rotation by a mm or 2 I am guaranteed to crash. And this doesn't include the complexity of some of the dimensions and the callouts. For example a profile without datums with a unequal tolerance called out..... or a position on a single plane.....

So with all that being said I have been programming these parts for a couple weeks now and my head at the end of the day is just mush. That's where the Forum comes in. So I want to thank all of you for giving me a source of relief from my mental torture chamber over here.

Also, What are you more exhausting program memories. What is that one job that really kicked your butt?


  • At my old job we did a few prototype parts for Boeings experimental drone tanker and one of those parts was a bottom skin piece except it wasn't just skin, it was structural with flanges, holes, stepped floors and walls, ect. We had to profile every feature of that part. Add to that, the part was about 12' long and 6' wide which was about 1' wider than the table. To get this done, I designed some triangles so I could set the part op at a 45 degree angle so it would fit and run and it had to be run with the tooling tabs attached. We also had to install insulated double doors so we could get the part on and off the machine. The part also had to be run on both sides. In total it was about 35k points and 1400 circles.

    I think that is about the time my hairline started to recede.
  • Years ago I had to do a full layout program for a drive shaft yoke forging. It was symmetrical left to right side and upper half to lower half. 2/3 of the dimensions were called out in all 4 quadrants. There were close to 300 measured features on the final report. I had no CAD, I had to fabricate my own fixturing with stuff around the shop, and I was 3 weeks out from my CMM 100 courses at Hexagon. There were a lot of short arc radii and distances to and from short arc radii. At the time, I really didn't understand what a short arc radius was. Took me 2 weeks and most of my sanity but, I got it done.

    I guess it was only bad because I was so new to everything. It didn't help that the layout was a month past due and everyone was rushing me to get it done. I still have a little PTSD from it.
  • Once I got so upset with a part that my face turned red and it fogged up my glasses. It was when I very first started programming. We had a routine for leveling our Rayco posts for a few pretty critical parts. I created the program and added in the routine. The mistake that I made was instead of leveling off the sub plate and z-org off the top of the post I leveled off the small Ø0.5" post and then tried to adjust a post that was a few feet away to the same height but the level was all skewed. I couldn't figure it out for hours and everyone was busy at the time so I was on my own.

    Honestly a great alignment learning experience.
  • Another one that was a ripper was a landing gear bulkhead for an older military jet. Everything was legacy and I was doing it on a laser tracker. hundreds of dimensions on a jacked up model that wasn't even 70% correct. It took me 8 days and the part was screaming hot. I was a company man at the time, so that last day I stayed up there until about 10:00pm that night (6:00am - 10:00pm) finishing it up.
  • I started at a new facility and they brought me on a little late. But the facility wasn't near completion either. So in short, I had to used a portable air compressor for my first couple weeks. Said facility had moved the CMM (Horizontal Arm) from a shuttered plant. They planned well and had a pit made to recess it so we could walk right onto the plate. Cool, my first horizontal. They were a Teir 1 to one specific GM facility. All was good except for the floor plan. It was the rear section, from the front of the back seat to the rear bumper. Around 60 stampings, 1300 welds, 70 weld studs and we have not even started talking about the profile requirements. And I had to measure something from all 6 directions. My fixture was a plate with 4 risers only to allow access and they were still in the way. I worked my first 43 of 45 days straight and I still hadn't got the program totally done. Originally I was leapfrogging the measurement routine so I would measure one half, spin the fixture and measure the other side. This was my first sheet metal facility so I learned the hard way about the need for the NI probe body. I think it was 3 months before I had the program finally proven out. Then I found out after the fact that the arm they moved was originally a dual arm setup. The sad thing was they ended up purchasing a refurbished dual arm a couple years later. And it took them 4 years to put in a hoist. That part neve got above a 70% pass rate, shoot the main stamping from GM never got above a 50% pass rating. Start with crap, don't expect a diamond in return. In the end, I had to reprogram it after receiving the dual arm machine. Which was another first for me.
  • We made this one part close to 10 years ago for a big aerospace company. I was still working part-time and going to college full-time. It wasn't a very big part, maybe 12" X 12" X 4" or so... but it had a ridiculous amount of features in it. The print was like 12 pages long, with over 800 dimensions, and it went all the way up to datum BE or something (through the alphabet twice). This was before we had Discus so I had to create an old school bubble print and by hand with a pen and a washer and manually create the AS9102 FAIR, and it ended up being almost 100 pages long. The program was so big, that if I had to edit something, after pressing F9, it took literally a minute to open the AutoFeature window (my computer at the time wasn't all that powerful either). It got to the point that I just cut sections of code out and pasted them into a "new" program so I could edit it, and then I would paste it back in to the actual program once the edits were complete.
  • Main gearbox housing for aircraft engine.

    Part was loaded into an inspection fixture and then put onto the CMM. Workpiece itself filled up entire measurement volume of my SF7107 with only about 2inches to spare on either side. The normal "home" for this job was our 152614 but they had me write a program on the 7107 to correlate some data output between the two CMMs.

    Program had to be stopped in 6 different areas for the fixture to be moved around so I could probe different areas. This job had a crap ton of EQUATE alignments & viewsets.

    All said and done the PCD rpt is 42 pages (RIP Amazon rainforest). There are hundreds of holes & different surfaces. Datum scheme goes all the way up to CQ (going from AA, AB, AC, AD, AE).

    I wish I could share pics but they are classified.
  • 17 foot long part on an machine with only 6 foot of travel. Lucky for me, it was before they "fixed" the EQUATE ALIGNMENT function. 3 setups, only 2 of which were pointed the same way, 3rd setup require the fixture to be turned 180 degrees, fixture hung over and mostly covered the desk for one setup.
  • One of these.



    But now I know it's because the program didn't have PNT31 to level and rotate to.

    ¯\_(ツ)_/¯
  • I have worked on some big complicated parts with thousands of features to report, but really the most mentally exhausting thing I've had to program was a couple of relatively simple components for some high pressure valves. The prints were just so over toleranced that half the parts measured out of spec no matter what machining processes they tried. Unfortunately, some of the parts did not work when pressure tested even if they measured to print and each manager had different theories about what needed to be changed and could be changed. I had modified those CMM programs and the related inspection documents more times than I could count for all the special test runs and revisions that were made on those parts. I spent so many hundreds if not thousands of hours documenting nonconformities, attending design review meetings, and creating many different reports that no one bothered to read. Those parts and the company dysfunction surrounding them were the main reason I decided to leave that employer after having worked there for 16 years.
© 2024 Hexagon AB and/or its subsidiaries. Privacy Policy | Cloud Services Agreement