performance from a different perspective
Posted: Sun Feb 15, 2004 4:17 am
Hey gang,
The previous post on performance struck my curiosity..It's interesting to know how some users will actually time toolpath calculations from version to version.I think its a great way to keep mecsoft on their toes!!I am not sure exactly what they were getting at with the insert cutter thing.I understand what mecsoft was talking about as far as the insert cutter choice on the tool creation page,just un-clear as to what bennefit they were trying to achieve as nobody ever tells me anything...Possibly trying to make the software understand an insert cutter cant go as deep in a pocket condition as say a center cutting endmill..Not sure!
Anyways,I thought it would be interesting to go one step further and take Visualmill on the road with me to a friends shop who is running mastercam and really give this thing a test!!!I currently use 4.0 so I took my 5.0 demo along.
Before continuing I would like to say this is in no way meant to make mastercam look like a weak package..On the contrary,it is a mature product with some awesome toolpaths.Like 10 differen't rough algo's alone!!What the heck is a morph spiral anyways;-) The test was just for fun and just as much about my big mouth and his as it was a science experiment...As my friend always pushes Mastercam and I am always telling him about the one to watch for...Visualmill
This test was done focusing solely on 3d toolpaths, Same 30in*30in Mold surface with some fairly complex geometry.Computer was a dell with a 1.5 mhz pentium 4 512 megs ddr,Nvidia graphics card.
Toolpath calculations are on top,simulation times underneath..Thats where the biggest surprise lied,as I spend a lot of time with the sim model..Both simulations were roughly set at the same display quality and lines of code@10,000.
Heres how it went:
Horizontal rough
1.0 dia. bull .0625 c'rad\tolerance .001(--crazy tol.--)\step down .020\step over 50%
M.C. = 3 min crunch time V.M. = 4 min crunch time
M.C. = 6 min sim.time V.M. =***1 min sim. time***
Horizontal finish
1.0 dia. bull .0625 c'rad\tolerance .001\step down .020
M.C. = 3 min crunch time V.M. = 4 min crunch time
M.C. = 4 min sim.time V.M. = ***50 second sim. time***
parallel finish
.5 dia. ball\tolerance .001\step over .005(ouch)
M.C. = 12 min crunch time V.M. = ***4.5 min crunch time***
M.C. = 5 min sim.time V.M. = ***50 second sim. time***
pencil trace
.250 ball\tolerance .001
M.C. = 1.5 min crunch time V.M. = 1 min crunch time
M.C. = to fast to worry about V.M. = to fast to worry about
pencil trace
.250 flat\tolerance .001
M.C. = 1.5 min crunch time V.M. = 1 min crunch time
M.C. = to fast to worry about V.M. = to fast to worry about
conclusion: Visualmill did outstanding!! I was really pleased at the difference in simulation time,Model quality was the same but wow V.M just really has that nailed right down.Cutting times went to Mastercam with a slight edge. Each had there strong points. V.M. excelled in the parallel finish,Where mastercam just has the nicest flat cutter pencil trace around. Visualmill looked kind of ugly there.Ball pencils looked identical.
Roughing out smaller area's the 1 in. couldn't get down in, goes to mastercam it spanked V.M. in that area.It took seconds to regenerate seperate windowed areas,while V.M took the same time it took to crunch the whole job..Definately an area that needs a major overhaul... As far as surface finish goes,,that has as much to do with the programmer and the conditions as it does the software.However Visualmill looks every bit as sweet and our bench guys don't complain..
With all the slope machining stuff and curve stuff I had an answere for almost everything that he pulled out.
I really didn't want to get to much in his face as I was afraid he would whip out the dreaded----CONTOUR_RE-MACHINE----;-)~
All in all it was a fun day and I left the shop in one piece still wearing my Visualmill shop coat...
Greg
The previous post on performance struck my curiosity..It's interesting to know how some users will actually time toolpath calculations from version to version.I think its a great way to keep mecsoft on their toes!!I am not sure exactly what they were getting at with the insert cutter thing.I understand what mecsoft was talking about as far as the insert cutter choice on the tool creation page,just un-clear as to what bennefit they were trying to achieve as nobody ever tells me anything...Possibly trying to make the software understand an insert cutter cant go as deep in a pocket condition as say a center cutting endmill..Not sure!
Anyways,I thought it would be interesting to go one step further and take Visualmill on the road with me to a friends shop who is running mastercam and really give this thing a test!!!I currently use 4.0 so I took my 5.0 demo along.
Before continuing I would like to say this is in no way meant to make mastercam look like a weak package..On the contrary,it is a mature product with some awesome toolpaths.Like 10 differen't rough algo's alone!!What the heck is a morph spiral anyways;-) The test was just for fun and just as much about my big mouth and his as it was a science experiment...As my friend always pushes Mastercam and I am always telling him about the one to watch for...Visualmill
This test was done focusing solely on 3d toolpaths, Same 30in*30in Mold surface with some fairly complex geometry.Computer was a dell with a 1.5 mhz pentium 4 512 megs ddr,Nvidia graphics card.
Toolpath calculations are on top,simulation times underneath..Thats where the biggest surprise lied,as I spend a lot of time with the sim model..Both simulations were roughly set at the same display quality and lines of code@10,000.
Heres how it went:
Horizontal rough
1.0 dia. bull .0625 c'rad\tolerance .001(--crazy tol.--)\step down .020\step over 50%
M.C. = 3 min crunch time V.M. = 4 min crunch time
M.C. = 6 min sim.time V.M. =***1 min sim. time***
Horizontal finish
1.0 dia. bull .0625 c'rad\tolerance .001\step down .020
M.C. = 3 min crunch time V.M. = 4 min crunch time
M.C. = 4 min sim.time V.M. = ***50 second sim. time***
parallel finish
.5 dia. ball\tolerance .001\step over .005(ouch)
M.C. = 12 min crunch time V.M. = ***4.5 min crunch time***
M.C. = 5 min sim.time V.M. = ***50 second sim. time***
pencil trace
.250 ball\tolerance .001
M.C. = 1.5 min crunch time V.M. = 1 min crunch time
M.C. = to fast to worry about V.M. = to fast to worry about
pencil trace
.250 flat\tolerance .001
M.C. = 1.5 min crunch time V.M. = 1 min crunch time
M.C. = to fast to worry about V.M. = to fast to worry about
conclusion: Visualmill did outstanding!! I was really pleased at the difference in simulation time,Model quality was the same but wow V.M just really has that nailed right down.Cutting times went to Mastercam with a slight edge. Each had there strong points. V.M. excelled in the parallel finish,Where mastercam just has the nicest flat cutter pencil trace around. Visualmill looked kind of ugly there.Ball pencils looked identical.
Roughing out smaller area's the 1 in. couldn't get down in, goes to mastercam it spanked V.M. in that area.It took seconds to regenerate seperate windowed areas,while V.M took the same time it took to crunch the whole job..Definately an area that needs a major overhaul... As far as surface finish goes,,that has as much to do with the programmer and the conditions as it does the software.However Visualmill looks every bit as sweet and our bench guys don't complain..
With all the slope machining stuff and curve stuff I had an answere for almost everything that he pulled out.
I really didn't want to get to much in his face as I was afraid he would whip out the dreaded----CONTOUR_RE-MACHINE----;-)~
All in all it was a fun day and I left the shop in one piece still wearing my Visualmill shop coat...
Greg