You would think it would slow things down too much, but it doesn't have to. The machine could be modeled using simple surfaces like autocad used to use to model a house, basically called polygon modeling. The graphics involved to show that stuff is minimal to say the least. Basically use video game modeling to do the machine. It would be super lightweight. On top of that, it's what directx and opengl was designed to show on the screen. Simple triangles.
The trick would be for the graphics card to differentiate between simple triangles and complex visual data from the solid model of the part. Something the software writers of the cam software could easily provide to the graphics card driver writers?
"vinny" wrote in news:h6qju0$rq6$ snipped-for-privacy@solani.org:
Yup. With CAM software that comes with the machine. Runs in XP and you have to have a legit copy of Excel installed of all things.
Probably comparable to 10 yo video game 3D quality wise. But if it were better a lot of shops wouldn't have a computer to run it on. LOL.
Esprit does a pretty nifty job too. I've got a customer using NX that I gave some solid models of the machines too as well. I haven't seen the end result yet but I imagine it will be better than our little CAM program.
I would think you could get models of your machine from the builder.
There is a need to do it the other way, just to check the code and setup. I cannot trust Edgecam, simulation is fine, post processing is poor. We add macros to posted code and check the programs with Vericut.
Vericut scares me. There aren't any videos on how to learn it. I just make a scrap part and then graphically edit out the mistakes and then make another scrap part. This process repeats itselt endlessly, but since I am at a not-for-profit shop, it doesn't make any difference.
Re: my recent reply to Zymrgy, can you tell my hand is still painfully sore and that I have missplaced my Midol? I'm a real bitch when that happens!
I did simulation tests on my design - used 8 Sun computers to split the tasks up and ran it over the weekend. It finished just after Sunday midnight. Ready to be looked at Monday for checking for errors.
If you get a chance load up on of the NX ISV demos on your system to run. I use the ISV on a couple machines we have (Hermle c30 & old SNK). It's VERY lightweight on the system resources. Both are modeled fairly detailed. Back when I was at Haas I used Vericuts Machine Sim. That system uses very simple modeling tools to keep overhead light as well. Of course model components were very easy to come by there.
I do program with no machine in mind unless it's 5 axis since we only have one of those.
The system has verification and simulation. Simulation is machine specific, you have to pick a machine file. I was just curious if that technology is being used out there yet? I try to run it and it goes all bonkers. Maybe I'm doing something wrong lol
na, can't be me, it has to be a bug in the software!
D Murphy wrote in news:0n5km.275139$Ta5.163450 @newsfe15.iad:
Easier said than done with some MTB's, especially Japanese companies. I've had no problems getting them from European companies. Mazak has them, but others it's like pulling teeth.
The Taiwanese are the worst. They apparently believe that solids for a simulation indicates a desire to make a copy of their bad copy of someone else's barely adequate machine tool. Frankly, I think it's a projection of the culture of intellectual thievery they are so proud of.
I guess being the losing side in your civil war messes with the brain pan. One of these days Formosa will pull up it's pants.
Not so wrong approach actually... Over simplified explanation (very deep area): In NX you have a module called Machine Tool Builder (free with cam btw). You create the kinematics there. You create the models as normal NX assemblies with some MTB work. You then go to Machine View and select the machine. Your mcs, geometry, operations, and tools are all children of the machine. Since your working in assemblies they must be below the machine. It is integrated into the part file. That said, you can start with the generic machine then swap it later. The machine will attach to the mcs. It's really easier to know which machine your going to be on first though.
It's not as easy as Vericut to get a new machine built. Try the sample files first. There are two types, the newer CSA turnkey style integrated (and encrypted machine files) and the older MTB style which I use.
I wished I had a class in the Machine Tool Builder as I really enjoy building them but have had to learn them the old on-the-fly method over the years. They have a few samples as well on the GTAC download area also.
When set up it's pretty cool. Because it's running through the MTB post (which is setup using Postbuilder), as it simulates it's running the gcode (not cls). You can't interrogate the as cut like Vericut (yet) but works well for showing complex axis moves. I like to use it for simulating shrouded impellers as you see just how much axis rotation your creating based on tool axis (Hermle very fast, Bosto, very slow).
Can UG do something like this? In Mastercam I could use check surfaces to control things. For instance, I am cutting a floor finished, but I'd like to stay away from the wall. So I would make the wall face a check surf and leave stock on it. This makes me think using a workpiece for the parent with the part in it is not the way to go. Leave the stock block in the workpiece parent, but I think I should select the part for each operation as I go. This would allow me to select the part as faces, allowing me to remove any faces from the part as needed to use for check surfs, etc? Then I would only have one workpiece for the whole job. And have the flexability to use the part as needed.
Is it standard to have a workpiece with the part and blank in it as the parent?
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.