Bryan Carpio Felsher
So...I typically just work with what is available in Catia and around the 10th attempt get a path that looks acceptible. When I pick paths from my process catalogue, I can usually get a nice path by the 3rd try or so...LOL! But seriously, I really don't think there's any way a CAM system can outthink the path a good programmer would come up with if he were to write every single move out by hand.
Okay, in most cases, the time this takes isn't worth the trouble. I usually try to find the balance, as time=money.
I'm willing to admit that probably 50% of my problems getting a good path out of Catia is operator error. The roughing in Catia is a bit odd to set-up, yet- this is just a matter of learning to use the interface properly. At least for me it is...
That being said, the most important thing I would ask for is simply for Catia to generate the paths quicker.
A few projects ago, I worked on some 180 x 20 x 10 inch structural titanium parts for a 5-axis multi-spindle gantry. I consider these types of parts to be my personal specialty. Well, at least, I very much enjoy programming big hog-outs on machines with 100 h.p. heads. Anyways, a "roughing" path in Catia would take literally about 10 minutes to generate. This is when I usually hit replay on several paths and either a) take a crap b) put the clothes in the dryer c) rant and rave screaming at my monitor to hurry the heck up like the angry german child on you-tube. (always keep a couple spare keyboards...)
Well, for an experiment (simple y-axis back and forth style roughing entering outside the part), I tried to create the exact same type of roughing path in another CAM system that is considered to be child's play compared to Catia. This system isn't even apt based and spits out only a binary CL file. (maybe those are quicker?) Anyways, the system created the path on screen in about 5 seconds! Now why can't Catia do that? Remember- 180" x 20" x 10" stock 90% radial stepover back and forth in Y-axis, with automatic "H.S." cornering- and the "other" system did it on screen in only 5 seconds and the tape was as pretty as I could have programmed every single move by hand! Good thing I have a converter to APT, because I can make good money by creating the roughing path with another system, then importing the aptsource back into Catia so it's all still in Catia and I can watch it run on screen, post out a complete tape, etc...Then program all the finesse multi-axis finishing in Catia....tricks of the trade...
Bryan Carpio Felsher
I even have a customer I visit, who programs literally everything in APT- doesn't even have roadmap generators, or any kind of graphical software at all- everything done by hand on paper....the guy is pretty old....but really cool to watch. He has a windows XP PC, but he writes his programs with the stock DOS text editor!!! Now that is SERIOUSLY old-school!!! I told him, his boss ought to buy him some paper program cards and build him a mainframe! That guy is so freakin' cool! I mean, he is programming really complex 5-axis parts with complex ruled surfaces, polyquadratic surfaces, and everything- all with no graphics! It's truly awesome...he's retiring in a couple months...the last of a dying breed....NC programmers really used to have to be really smart- he makes me feel like a serious simpleton! But in a good way...one of the sharpest guys I've ever known.
I really love this stuff. I love hearing old stories about the old days...Fortran, APT-360, mainframes, John Parsons with 3 guys cranking handwheels with 3 guys behind them yelling out values from a calculated print-out to prove how NC machines could work...making templates for helicopter blades for Sikorsky...seriously, a movie about the life of John Parsons would be so awesome!!!
A bit of history, when COE was in Traverse City, Larry Jones (old school post guy) called John Parsons and got him to come speak to the NC group. We all got to shake his hand and say thanks. Really cool. Like meeting the Wright Brothers or Henry Ford.
Bryan Carpio Felsher
Larry Jones is a good friend of mine- and he's told me about the Jim Parson's story. He said that for him, it was like meeting Mickey Mantle. Larry's still really busy, and he's written nearly all of my post-processors and special applications. He retired from IBM, and now owns "NCDATA Services." Really great guy, and he's taught me a lot. We went out to lunch together a few months back...hopefully he'll make his way out here soon again! Indeed, it's a small world. Larry's told me some really cool stories about his days as an NC programmer- with time/shared mainframe computers, etc...by the way, I have some very cool articles about John Parsons, and I will copy and paste them onto a post soon.
I missed out on all that excitement, and often think of how it must have been. I think I would have been pretty good at it. But I am glad to say, that at least I got to see the big explosion of graphical systems, and the consequential wage increase to programmers. (I started out as a manual programmer with a compass, calculator, and paper using Easytalk!- talk about slow and stupid...we didn't even have APT and I was just a kid working in a shop where the boss only came in for an hour a day at lunch break from his day job-started out running a band sander, then snuck onto the few Acroloc CNC machines we had...embarassing to admit...a few crashes...but eventually became a decent 3D programmer after getting sent to a school that I begged to go to.) I think maybe it just got back to the kind of wages 5-axis contract APT programmers used to make to create programs in the days before every shop had their own system.