DKing
50+ Poster
Posts:129
 |
| 02 Jul 2009 03:13 PM |
|
Is there a way to use a T-Slot cutter when the part surface is NOT a plane but a surface? In the following picture I used Multi-Axis Curve Machining and the cutter drove fine but when I ran the program through Vericut it wanted to gouge the top of the slot a little bit, which I would expect. I'd like to be able to drive a cutter so that it would respect the top of the slot even if I had to use two thinner cutters and cut the top and bottom of the slot seperately.
Thanks,
Dave
 |
|
Dave King
TECT Aerospace |
|
|
ROGER_BOMBASSEI
1000+ Poster
Posts:1060
|
| 02 Jul 2009 03:31 PM |
|
Did you try lead and tilt tool axis? Lead and tilt set to zero
|
|
Roger Bombassei -
Online CATIA MFG Instructor
roger@catiaonline.com |
|
|
ROGER_BOMBASSEI
Occasional Poster
Posts:0
|
| 02 Jul 2009 03:31 PM |
|
Did you try lead and tilt tool axis? Lead and tilt set to zero
|
|
|
|
|
BFELSHER
1000+ Poster
Posts:1385
|
| 02 Jul 2009 03:32 PM |
|
Hi Dave,
For that type of stuff, (rather than fighting Catia) I usually extract/create the surfaces into a new geometrical set, and then drive what I need using a regular endmill with a user-representation, since that way, all the regular toolpaths will work, and I can at least video replay and visually see/compare what the cut will look like.
To control the upper surface, I usually would just "cheat" and offset the bottom surface or leave stock on part surface (just calculating what it should be based on the tool). I haven't been able to get it to compute off the top of a T-slotter correctly any other way. Maybe someone else knows a good trick. I've heard of guys doing some stuff, like cutting with the opposite side of an endmill (with a user representation) and then manually editing the aptsource to change the tool axis vector, but that seems dangerous to me. Of course, you could also build a custom post macro to do it...but "cheating" as I do, seems the easiest way. |
|
Bryan Felsher
True Precision
|
|
|
DKing
50+ Poster
Posts:129
|
| 02 Jul 2009 04:19 PM |
|
Roger,
I did not try that tool axis. The picture makes it hard to see but I really can't tilt the cutter because of the undulation of the floor.
Bryan,
I've done similar things when we used NCL. We'd just play with the thick on the bottom surface to get the top to come in, but since the floor goes from convex to concave the bottom of the tool won't gouge but the top has to since you're basically trying to bend a disk (cutter) around a corner and something has to give. If I have a narrow cutter and just cheat the bottom I THINK I will get the same results. The flipping the cutter probably works the best but I'm like you... that's scary stuff. However in this case I've set the part up so the tool axis of the cut would be 0,-1,0 so it would be easy to change but still. |
|
Dave King
TECT Aerospace |
|
|
BFELSHER
1000+ Poster
Posts:1385
|
| 02 Jul 2009 06:31 PM |
|
I know what you mean, Dave. Do you have much control over your post processor? No matter which direction the tool axis is pointing, it's pretty simple to flip the axis with the post.
It's possible to create a special mode if you do, and I could help you out with the correct transformation matrix. It's not overly complicated (yeah, right...)
First Translate the point. To do so, you just need the direction cosines (which you can get from the original GOTO)
Then if the T-Slotter is .250 thick, and you want to translate it therefore .250 on those direction cosines, then do .250 * I, .250 * J, .250 * K. Now, you have the new points.
Now, to flip the axis, you just have to multiply each vector by -1.
You could call this Translation and Flipping "Macro" (I think some posts would call it) by inserting something like:
MODE/FLPTOL,TRANS,.25 $$ AND STORE THE VALUES AS MODALS SO TURNING BACK ON DOESN'T HAVE TO REPEAT THE VALUES
MODE/FLPTOL,ON
MODE/FLPTOL,OFF
Or you could set it to translate by a set number of degrees. But there are different rotation matrixes depending on what plane you are rotating on. I could help you with that, if you need. It's also useful for 90 degree heads, of course. It's really simple. Just a couple of multiplications on each vector, and then translate along the new vector the same way I described above.
Then you'd program that area with the bottom of a tool, and the post would flip the axis and translate by the amount you set in the MODE statement. Of course, in order to see the result in Catia, you'd definitely have to use a user rep, with the user rep flipped 180. At least that way, you wouldn't have to do any manual editing of the aptsource and just have to put in a statements in your aptsource file to turn the mode on and off. Once you have the new mode set-up in your post, it wouldn't be tooooooo painful....
I'm building posts, so all this stuff is really fresh on my mind.
But...wouldn't it be easier if the CAM system just calculated based on the actual T-slotter? I think that NCL maybe used the CUTTER height to test for check surface violations? I don't know....never used it....maybe that's too much to ask. Anyways, just a thought. |
|
Bryan Felsher
True Precision
|
|
|
DFRANK
1000+ Poster
Posts:1269
|
| 03 Jul 2009 01:09 AM |
|
Dave,
Just for giggles, try setting your machining tolerance to .0002, and your max descreeeeetizzzzzation angle to .25, and see if the gouging improves.
hth
Dave |
|
Dave Frank Chief Instigator, Pac-Men Group. === Programmers Advising Catia, Make Enhancements Now 
Catia users+Power users+ the COE DPC's+the COE Forum>>> to brainstorm and collaborate = better Catia tomorrow |
|
|
DFRANK
1000+ Poster
Posts:1269
|
| 03 Jul 2009 01:09 AM |
|
sexy part by the way
how come you get all the good jobs? |
|
Dave Frank Chief Instigator, Pac-Men Group. === Programmers Advising Catia, Make Enhancements Now
Catia users+Power users+ the COE DPC's+the COE Forum>>> to brainstorm and collaborate = better Catia tomorrow |
|
|
DFRANK
1000+ Poster
Posts:1269
|
| 03 Jul 2009 01:18 AM |
|
One more thiing, if the cutter is same as the part geo, I would expect gouging with twist, ahead and behind the cutter, as it it not .... what do they say a rubber end mill If the cutter is say 2 x .5 x .250 Rad, and the slot is .500 wide, with any twist, it will have to undercut somewhere. If that is the case, a thinner wheel, .495, etc may be required hth Dave |
|
Dave Frank Chief Instigator, Pac-Men Group. === Programmers Advising Catia, Make Enhancements Now
Catia users+Power users+ the COE DPC's+the COE Forum>>> to brainstorm and collaborate = better Catia tomorrow |
|
|
ROGER_BOMBASSEI
1000+ Poster
Posts:1060
|
| 03 Jul 2009 02:00 AM |
|
Posted By DKing on 02 Jul 2009 04:19 PM
Roger,
I did not try that tool axis. The picture makes it hard to see but I really can't tilt the cutter because of the undulation of the floor.
Using lead and tilt tool axis set to zero is the same as normal to part surface. I was assuming anything other than normal to part surface would gouge the top.
|
|
Roger Bombassei -
Online CATIA MFG Instructor
roger@catiaonline.com |
|
|
ROGER_BOMBASSEI
Occasional Poster
Posts:0
|
| 03 Jul 2009 02:00 AM |
|
Posted By DKing on 02 Jul 2009 04:19 PM
Roger,
I did not try that tool axis. The picture makes it hard to see but I really can't tilt the cutter because of the undulation of the floor.
Using lead and tilt tool axis set to zero is the same as normal to part surface. I was assuming anything other than normal to part surface would gouge the top.
|
|
|
|
|
DKing
50+ Poster
Posts:129
|
| 03 Jul 2009 09:08 AM |
|
Roger,
You are correct... Lead and Tilt set to zero is exactly what I did. I programmed this path sometime ago when we were quoting the job and didn't remember that I had to use that tool axis to get normal to part surface. We haven't had a lot of need for that tool axis for a while and I must have assumed I'd used Normal to PS like I did in NCL.
Bryan,
I like your post idea. I do have control of the posts. We use G-Post from Austin N.C. It's a full apt processor so we can use all the fun stuff from they apt/ncl days. Naturally they always need new jobs completed NOW so I'm always pressed for time. I may take you up on your offer to help. Thank you.
Samarinder,
I'm getting about .004 undercut with the current tolerance settings. I may try Dave's idea to cut the tolerance down and see if it makes any difference. We actually have a profile of surface tolerance of .015 on this slot but I still don't like undercutting the part. Thanks for your testing. My cutter is actually .525 diameter with a full .06 radius and a .25 shank. We tested with a small diameter so that we could not only climb cut both sides but also because like Dave says, they don't make rubber cutters. I'm also attaching another picture so you can better see how much the slot undulates (my big word for the day).
Dave,
Trust me... most of our parts aren't this "sexy". I got all excited when we actually got the job.
I hope answering everyone in one post is not violating any forum etiquette... it seemed the most efficient.
|
|
Dave King
TECT Aerospace |
|
|
ROGER_BOMBASSEI
1000+ Poster
Posts:1060
|
| 03 Jul 2009 10:34 AM |
|
You'd have a difficult time cutting that even with the absolutly increadible UG  or the bulletproof NCL  . lol
I was just trying to answer the question at hand. But MAFC is probably your best bet.
Your post is fine. Great topic. |
|
Roger Bombassei -
Online CATIA MFG Instructor
roger@catiaonline.com |
|
|
ROGER_BOMBASSEI
Occasional Poster
Posts:0
|
| 03 Jul 2009 10:34 AM |
|
You'd have a difficult time cutting that even with the absolutly increadible UG or the bulletproof NCL . lol
I was just trying to answer the question at hand. But MAFC is probably your best bet.
Your post is fine. Great topic. |
|
|
|
|
DKing
50+ Poster
Posts:129
|
| 03 Jul 2009 10:49 AM |
|
I'm not really familiar with UG but I'd have to cut it the same way in NCL. It didn't watch the top of a cutter like this one either. While I miss some of NCL's functionality, I love programming with solids. |
|
Dave King
TECT Aerospace |
|
|
BFELSHER
1000+ Poster
Posts:1385
|
| 03 Jul 2009 11:59 AM |
|
G-Post is a great post. Best bang for the buck, for sure. I almost bought it, before I commenced to write my own routines and glue them together with FreeBasic. It's taken me longer than the price of G-post, but it's fun to write them from scratch, and a good learning experience.
Off topic a little, sorry, but is the CL file generated from G-Post similar to Catia? What do the circle records look like? Would you mind posting a little snippet? I'm curious because they wanted a lot of money to supply me with a G-post to Catia CL converter. I could easily make that myself with a sample G-post CL file.
As you know, Catia's are just basic APT style.
GOTO / .00000, .75000, .00000
INTOL / 0, 0, 0, 0
OUTTOL/ .0005, .0005, .0005, .0005
AUTOPS
INDIRV/ -1.00000, .00000, .00000
TLON,GOFWD/ (CIRCLE/ .00000, .00000, .00000,$
.75000),ON,(LINE/ .00000, .00000, .00000,$
-.75000, .00000, .00000)
AUTOPS
INDIRV/ .00000, -1.00000, .00000
TLON,GOFWD/ (CIRCLE/ .00000, .00000, .00000,$
.75000),ON, 2,INTOF,$
(LINE/ .00000, .00000, .00000,$
.00000, .75000, .00000)
GOTO / .00000, .25000, .00000
More standard CL files usually look something like this (converted with WPC View):
GOTO / 0, 0.75, 0
INTOL / 0, 0, 0, 0
OUTTOL/ 0.0005, 0.0005, 0.0005, 0.0005
AUTOPS
INDIRV/ -1, 0, 0
CIRCLE/ 0, 0, 0, 0, 0, 1, 0.75, 90, -0.75, 0, 0
AUTOPS
INDIRV/ 0, -1, 0
CIRCLE/ 0, 0, 0, 0, 0, 1, 0.75, 270, 0, 0.75, 0
GOTO / 0, 0.25, 0 |
|
Bryan Felsher
True Precision
|
|
|
DKing
50+ Poster
Posts:129
|
| 03 Jul 2009 12:18 PM |
|
They're great people over there at Austin N.C. too.
FEDRAT/5,IPM
GOTO /45.43803,16.37654,-42.266,0,0,1
GOTO /45.38722,16.37238,-42.266,0,0,1
GOTO /45.33776,16.36001,-42.266,0,0,1
GOTO /45.28609,16.33717,-42.266,0,0,1
GOTO /45.19162,16.33717,-42.266,0,0,1
FEDRAT/10,IPM
GOTO /31.60413,16.33717,-42.266,0,0,1
FEDRAT/5,IPM
OUTTOL/.001,.001,.001
INTOL /0,0,0
ARCDAT/31.60413,16.00257,-42.266,0,0,1,.3346
ARCMOV/CCLW,31.26953,16.00257,-42.266,90.00000105
GOTO /31.26953,15.80257,-42.266,0,0,1
CUTCOM/OFF
GOTO /31.51953,15.80257,-42.266,0,0,1
RAPID
GOTO /31.51953,15.80257,-38.70757,0,0,1
PPRINT OUTSIDE WALLS SMALLEST 2, FAR WALL
RAPID
GOTO /31.49452,20.61893,-38.70757,0,0,1
RAPID
GOTO /31.49452,20.61893,-42.266,0,0,1
CUTCOM/LEFT
FEDRAT/5,IPM
GOTO /31.24452,20.61893,-42.266,0,0,1
GOTO /31.24452,20.36893,-42.266,0,0,1
ARCDAT/31.57912,20.36893,-42.266,0,0,1,.3346
ARCMOV/CCLW,31.57912,20.03433,-42.266,90.00000105
FEDRAT/10,IPM
GOTO /45.20796,20.03433,-42.266,0,0,1
GOTO /45.26604,20.02836,-42.266,0,0,1
FEDRAT/5,IPM
GOTO /45.30823,20.01781,-42.266,0,0,1
GOTO /45.3599,19.99496,-42.266,0,0,1
GOTO /45.4599,19.99496,-42.266,0,0,1 |
|
Dave King
TECT Aerospace |
|
|
ROGER_BOMBASSEI
1000+ Poster
Posts:1060
|
| 03 Jul 2009 12:51 PM |
|
Posted By DKing on 03 Jul 2009 10:49 AM
I'm not really familiar with UG but I'd have to cut it the same way in NCL. It didn't watch the top of a cutter like this one either. While I miss some of NCL's functionality, I love programming with solids.
I only have about 5 years UG experience. 10 APT and 10 NCL.
Any one will get the job done. I usually stick to one at a time.
I Like NCL/APT geometry and motion associativity. Their associativity is more user friendly and predictable. And easier to change.
I could be programming with any of them but choose CATIA V5 and plan to stick with it or V6 until I retire.
|
|
Roger Bombassei -
Online CATIA MFG Instructor
roger@catiaonline.com |
|
|
ROGER_BOMBASSEI
Occasional Poster
Posts:0
|
| 03 Jul 2009 12:51 PM |
|
Posted By DKing on 03 Jul 2009 10:49 AM
I'm not really familiar with UG but I'd have to cut it the same way in NCL. It didn't watch the top of a cutter like this one either. While I miss some of NCL's functionality, I love programming with solids.
I only have about 5 years UG experience. 10 APT and 10 NCL.
Any one will get the job done. I usually stick to one at a time.
I Like NCL/APT geometry and motion associativity. Their associativity is more user friendly and predictable. And easier to change.
I could be programming with any of them but choose CATIA V5 and plan to stick with it or V6 until I retire.
|
|
|
|
|
BFELSHER
1000+ Poster
Posts:1385
|
| 03 Jul 2009 12:52 PM |
|
Thanks a lot Dave! That looks really good, and easier for the post-processor for sure, since the instructions are all just right there.
Thanks! |
|
Bryan Felsher
True Precision
|
|
|