A New Commercial-Grade NC Verification and Simulation System for the Masses

[revygcam]

We are currently looking into developing a new NC Verification and Simulation (Virtual Machining) software system that would have many of the features and look-and-feel of commercial-grade systems at a more affordable price for the masses.

A brief introduction of NC Verification and Simulation systems for those that are not familiar with them:

Many NC programs are verified through trial machining which are slow, costly and sometimes dangerous to the machine and/or the operator. CNC users can overcome this by using software tools to simulate and verify NC programs in a Virtual Machining environment before making the actual cuts on their parts. Also, such programs give the added benefit of allowing the optimization of cutting parameters which in turn speed up the machining process, improve the quality of parts, and increase tool life. Commercial-grade systems allow users to visualize the cutting process and the part in fully interactive 3D modes (including viewing from any angle, zooming into tight areas, use of transparencies and 3D sectioning, etc.), get feedback on cutting times, accuracy and mistakes (e.g. gouges and undercuts), and directly debug and optimize NC programs.


There are currently a few commercial systems in the marketplace that sell for many thousands of dollars and are feature-rich. There are also smaller software systems that sell for much less, but lack the accuracy, speed, features and look-and-feel of their commercial equivalents. We want to create a system that fills this gap by delivering many (but not all) of the features of commercial-grade systems at lower-costs to allow mainstream CNC users access to the benefits of a more useful NC Verification and Simulation system.

Since the development of such a system involves a great deal of effort, time, and expense, please help us in our decisions by responding to the survey and making any related comments in your messages.

Thank you.

revygcam

(Please see a general overview of the software in my followup posting below)

[qtronpowerson]

* Does it verify the G-code program?
* Does it support advanced CNC control features used by the NC program, such as macro language, branching, probing and variables?
* Does it create an accurate in-process model?
* Can it verify cuts with different tool orientations, and can it remove material with a changing tool axis?
* Can it optimize the NC program feed-rates for varying cutting conditions, making the machines run more productively?

[dertsap]

i'm sure you will get lots of interest , if you can offer a prog with similar features to vericut at a better price

[revygcam]

Hi,

There seems to be some interest so I should at least clarify a bit more what I have in mind by giving a general overview about this software. Hopefully I can fill in more gaps in later postings with your feedback.

The idea is to create an affordable software package to be able to simulate your system closely by building a simple yet accurate kinematics model of it once through simple wizards and save it to a library of "machines", together with a library of cutting tools (and the ability to create new ones of course), a library of fixtures and clamps (for collision-detection), and stock (basic blocks or actual 3D models, for castings for example), all based on STL models that can be imported and setup as part of the virtual system. People can make their machines, cutting tools, and fixture definition files available to others so that the cnc community can benefit from them and we would eventually have a free repository that you can go to for existing machines (or tools and fixtures) if you want, so that you don't have to make your own.

At first the simulation/verification system would take in general RS-274 G and M Code (post-processed, not CAM generated APT toolpaths), but I'm hoping to be able to make the system eventually take plug-ins of specific controllers. Some of the other features would be:

- You could run the simulation and watch not only the cutting tool, stock, and important parameters at work, but also your whole machine and clamps/fixtures to detect collisions, etc.
- You would be able to see and manipulate the cutting process in full 3D with the look-and-feel of solid-modeling programs such as SolidWorks.
- You could single-block and edit the G-Code (color coded).
- You could increase the accuracy of the simulation as necessary (at the expense of memory and speed of course).
- You could verify gouges/undercuts using imported STL models of the original designed part for comparisons.
- Error and collisions would show up in customizable colors and formats.
- You could optimize feedrates based on the machine and cutting parameters.
- You could do simple inspections by verifying dimensions dynamically on the models
- And as mentioned, the system would be flexible enough for you to either build your own machines/tools/fixtures/stocks, or download them from a central repository made by other users in the community.

The project will be iterative in that it will initially start as a system for 3-axis milling (with 4 and 5 axis as add-ons later for those who want them), and later for turning, if there is enough interest. Actual RS-274 G & M code would be directly interpreted (not APT or CL output, although that may also be available). Again, more complex G-Code functionality such as variables, etc. would be added in later iterations as the software matures. Of course the price of the software will directly reflect its level of features as it matures. It will also be open to testing and scrutiny of potential users before it becomes a viable system. It would run on Windows 2000 and up machines, with the usual specs for speed and memory necessary for graphic-intensive applications.

I know that all this sounds like too big a mountain to climb, and I'm not naive to think it's easy by any means, but I'm hoping that my current experience will at least be a good start. My background is in both software engineering (3D graphic systems development and solid-modeling) and mechanical engineering (automation/robotics). But I'm going to need a lot of help and am hoping to be able to pull this off using your help. I'd like to get the feedback and collaboration of as many people in the CNC community as possible throughout the designing, building, and testing process of this software so that we all end up with something that is useful and has functions people want.

Let me know your thoughts.

revygcam

[qtronpowerson]

New to CNC: What do U mean by "cam generated APT toolpaths"

[revygcam]

Originally Posted by qtronpowerson New to CNC: What do U mean by "cam generated APT toolpaths"

An APT-CL toolpath file output by a CAM system...


See the following links for explanations on APT and CL:

http://www.nfrpartners.com/cncfaq.htm#whatisapt
http://www.nfrpartners.com/cncfaq.htm#whatisclfile


Also, see ANSI INCITS 37-1999 for the standard at:
http://webstore.ansi.org/ansidocstor...INCITS+37-1999


revygcam

[qtronpowerson]

Hey what's going on?
That's the 6th duplicted email to my mail box... Stop it!!!

[dertsap]

4 or 5 axis simulation would be of far more importance than simple 3 axis machining , the key is to be able to proove complicated setups , such as tool clearances , machine travel and such (possible crashes ) ,
3 axis machine simulation would more be better sold as educational than a true industrial software

[revygcam]

Originally Posted by qtronpowerson Hey what's going on? That's the 6th duplicted email to my mail box... Stop it!!!

This was a computer glitch and I explained the issue to the CNCZone webmaster.

revygcam

[jackson]

my question is can you make a conversational version i dont want to have to draw the part i have a print i dont want to draw the part.

[revygcam]

Originally Posted by jackson my question is can you make a conversational version i dont want to have to draw the part i have a print i dont want to draw the part.

Since the system will be setup as a virtual environment simulating the actual machine, I'm sure it would be relatively straight forward to include conversational functionality. These could even be in the form of plug-in wizards that at the end generate G-Code to be verified and then used to cut the part. Having a plug-in architecture would allow other developers/programmers to create new conversational routines as wizards, adding to a library of routines, creating a large useful base to generate G-Codes to be interpreted, verified, simulated, and eventually used to cut parts.

revygcam

[wildcat]

How do these verification systems generally work? Do they use simply move the virtual tool some very time step and remove the material that is overlapping with the envelope of the tool and continue this process until reaching the end of the program?