CAD software discussion forum > 3D CAD/CAM > My 1st QM project... Am I doing something wrong?

My 1st QM project... Am I doing something wrong?

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 1 of 11

 My 1st QM project... Am I doing something wrong?
22-11-2004 01:14 . pm | View his/her posts only
I've just started playing with the CAM end of VX. Trying to make an electrode. No an overly complex part, and something we do quite often... Here's the picture of the part.

Now, the 1st rougher, using a Diamon coated DAPRA .750/.031 tool worked well.

But, where I'm having troubles, is rest-roughing from the 3/4 tool to a 1/4" with a .015"TNR.


As you can see QM missed A LOT of areas, that it really should have cut.

Am I doing something wrong? I tried moving the Analysis Acuracy up about 15%, and it took a _LONG_ time to re-crunch. Like 35 minutes. And the toolpath is seemingly unchanged.

on both the rougher and the rest-rougher I've got the path tol at .004, XY surf thick at .01.

Rank: 1

OldForumPost

Newbie

posts: 0

Registered: 2012-1-14

Message 2 of 11

22-11-2004 01:34 . pm | View his/her posts only
Hi Steve,

It looks like you've followed good procedure and I'm not sure why you're having problems. You should email the part to support@vx.com. Also copy to me in case they are tied up and can't get to it quickly enough.

Regards,
BritVic

VicB@VX.com

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 3 of 11

22-11-2004 02:04 . pm | View his/her posts only
Thanks for the timely reply.
Is an 11 meg zip file too large for you and VX support to handle by email?

I also have another question reguarding electrode machining. Now, I *tried* to do a planar offset trode, but it didn't work. That's all we do is planar orbits in our EDM. Complex orbits on the Makino are too time consuming process. What I've always done in the past with PowerMill is to have some tools setup within PM that were called "offset tip radius" tools, when using actual ball mills. Using bull nose tools, isn't a problem obviously, but what does VX suggest when using ball nose tools, when an obvious negative XY offset is desired? I was told by our vendor that negative offset XY was possible, but read in here that it's not a desired process in QM.



Rank: 1

george

Newbie

posts: 0

Registered: 2004-3-1

Message 4 of 11

24-11-2004 08:39 . am | View his/her posts only
Steve,
You can put the file on the VX ftp site. Are you by chance limiting the rest machine path with a boundry? With VX I have noticed that a boundry can not be too complex or it will give you poor results like you are getting. Did you try and use a finish tool path with stock added?

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 5 of 11

28-11-2004 09:57 . am | View his/her posts only
looks like Analysis accuracy is to coarse.

Try to move the slider in the middle in the rest rough path.

If you give me a chance I will fit some nice rest rough on your part and send it to you back to examine what I did.

regards

Rank: 1

Robert

Newbie

posts: 0

Registered: 2011-11-22

Message 6 of 11

29-11-2004 02:19 . am | View his/her posts only
On the outside chance that there's a problem with inch/metric conversion, did you try machining the part in a metric CAM plan? I've experienced some discrepancies in inch/metric conversion.

Sorry that you're having difficulties

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 7 of 11

29-11-2004 09:24 . am | View his/her posts only
Quote

Are you by chance limiting the rest machine path with a boundry?


Nope.

Quote

Try to move the slider in the middle in the rest rough path


Wont that add an enormous amount of calculating time? I moved the slider down about 15% and it took over 30 minutes to crunch, Which, IMHO is unaceptable. Not trying to "dis" QM or VX, but The calculating times seem to be quite a bit longer than anything I've used in the past, on an order of magnitude compared to PowerMill. This is on a Dell 2.8GHZ P4 with 2 gig of ram.

I did the same toolpaths in powermill. The rougher using a .0005" Surface tol, .002" Path tol, and .005" thick, .03" stepdown, .600 stepover calculated in 25 seconds. The rest rougher calculated with the same parameters, save for a .018" stepdown, and a .200" stepover using a .250/.03" tool calculated in 2 minutes and 15 seconds, referenceing the 1st rough toolpath. These times were on PowerMill V4.1 (a 2 year old + release), on a 2.4GHZ P4 with 1 gig of ram.

Quote

did you try machining the part in a metric CAM plan?

No, and I don't want to, nor should I have to.

I've tried making another cam plan, added a QM rougher, and changed nothing but the Z stepdown to .03", and thickness to .01" , VX crunches the toolpath for about a minute, and boom, kicked to the desktop, and VX quits. I've pretty much given up trying to machine this "trode" in VX, and resorted back to PowerMill. This has left a somewhat bad taste in my mouth for VX's cam end of things.

#1.) The GUI *REALLY* needs to be revamped.
#2.) A clearer explaination of the options need to be included in the VX help file.
#3.) Calculation times REALLY need to be an order of magnitude shorter IMHO.

I am uploading the Zipped up file to VX's public FTP server as 1825_61P21.zip as I write this, if anyone wants to look at it. It's not a complex part by any means.

I've also sent the file in question to support@vx.com on 11/22 with no reply as of yet.

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 8 of 11

29-11-2004 02:16 . pm | View his/her posts only
First thank you for your recommendations!

Second the other small issues:
Regarding time calculation QM is around 10-40% faster then PowerMill this was benchmarked more then one time obvious these weren't your test case and I dont have a clue what Delcam did last one year. Feature wise I believe we offer the same feature set maybe eventually AFC, Caching, much sensitive rest place detection, as an extra. I think our production-milling cycles, nurbs cycles and 5 axis cycles are a welcomed addition not found in PowerMill. Adding our CAD and price capabilities we hope will make VX suite a better choice than Delcam one.

Regarding "Calculation times REALLY need to be an order of magnitude shorter IMHO. " I still pretend, overall, VX.QuickMilling extensions are faster in majority of cases then all of our major competitors. Please do not forget in VX you are retessellating each time a given tolerance for a given part isnt found in the cache. VX is a dynamic CAD/CAM package that regenerates paths if a part is altered. In many more static CAM systems retessellation and projector creation (core object used to generate toolpath) can be processed before toolpath generation halving the time calculation, in many of them this is explicitly done at import time unfortunately we cant afford this in VX do its dynamic nature. However you can do the same thing in VX if you dont alter the part or tolerance and youll see speed improvements of 10-70%. I want to ask you to generate for a complex part somewhere in 1000-10000 surface range a parallel lace in QM with a given step and after this change the step to a 10 times smaller one (this will give QM opportunity to act as a static CAM system Delcam in this case) and benchmark it against Delcam. Now both will have to do the same calculation effort. Be kind and disable QMs HSM extensions rounded corners. I will really appreciate the 2 times QM second generation VS Delcam. Just to remind you QM static generating 10 times denser path can be in this case 5 times faster the QM dynamic generating a 10th of the path.

I fill sorry you didn't find your way using the product yet but really QM generates good path fast. Actually I'm curious to find a product that generates at the same tolerance path faster then QM in same milling conditions. Don't try to remind here about: Machining Strategist, Work NC, PowerMill, MasterCAM, SurfCAM, EdgeCAM, Tebis. Regarding enumerated products QM was in general faster in complex molds generation scenarios. Of course to use a product first you should know how to use it and definitely years of PowerMill usage are better the days of QM usage. I really dont want to start a war now about who is a good and bad because is a complex issue.

Regarding missing a lot of areas: You should instruct QM to be more sensitive if you want to not only via (slow) increased analysis resolution but via Min Rest Height setting on limiting tab. Did you try to divide by 10 that parameter if sensitivity is your concern ? I wonder how many time youve used that parameter to tweak rest rough detection.

Not all parts and users are the same and thats the reason you have parameters there to tweak the product to your needs. In past we had many complains about too sensitive rest rough detection and now we move it to a bit more ignorant detection. QM is used for parts sizes ranging: (5x5x5 mm^3 & 2000x2000x2000 mm^3) Is very difficult only looking at this to find general purpose settings.


I will still appreciate to have your part and generates some paths on it. Rest roughing can be simulated with Z Level Rest finishing too.

Regards

Rank: 1

Dan

Newbie

posts: 0

Registered: 2002-8-26

Message 9 of 11

02-12-2004 03:06 . pm | View his/her posts only
This part has a bug that should be fixed in VX10.6.

Thank you very much Steve for reporting it.

Rank: 1

Steve

Newbie

posts: 0

Registered: 2004-10-9

Message 10 of 11

07-01-2005 05:42 . pm | View his/her posts only
Thanks to billator, I've gotten past the QM bug, and have started making what looks like pretty efficent tool paths now. I'd also like to take this oppurtunity to appologize for being a bit over-critical of billator's work

I've been playing around with QM the past month learning the different strategies and settings. While I still think the GUI sucks, and *REALLY* needs to be revamped, the calculation speeds, now that I've gotten past the bug are quite good, and in many cases faster thean PowerMills speeds. Just 2 points IMHO *REALLY* need to be adressed in the near future with VX cam developemnt.

1.) Better/Complete documentation. There are several parameters/setting/tools that aren't in ANY documentation that VX supplies. Take for instance, try to find what a "wheel" tool is. It's NOWHERE in any form of documentation. This is unacceptable IMHO.

2.) Revamp the GUI. Am I sounding like a broken record yet?

Rank: 1

OldForumPost

Newbie

posts: 0

Registered: 2012-1-14

Message 11 of 11

11-01-2005 07:53 . am | View his/her posts only
Hi Steve,

I'm glad that you're having more success with QM. I"m surprised that you find it so bad after being a Powermill user. I've played with it a bit and and know a lot of people that think it's GUI is awful.

I am a sales engineer in Indiana and am trying to work with our customers and development to improve our GUI.

Could you be a little more specific, but not detailed, about some of the things that you don't like in the GUI?

Many Thx,
Britvic
See also