When will MasterCAM simulate post processed programs?

[themachinist]

I'm just curious if this is on the road map. You guys have 85% of the work done. There are several open source g code interpreters out there: NIST rs-274 interpreter, smoothie, reprap. Just grab one and work it into the new simulator.

[Watcher]

Honestly... Never say never but I don´t think MC needs to jump into the "We simulate G-Code" wagon...

 

Verification is a very complex subject and CNC Software made the right decision, as many, to license their verification technologies. There´s a great chance that somewhen in the future, their partners commoditize such functionality and then everyone will want to offer it.

 

I remember that a decade ago things like "Nurbs" support was the hype and today we barely hear about it since the CNC controls evolved to a point it´s a breeze for them to chew point-to-point toolpaths and improve them using internal "compressors".

 

There are a number of reasons why just  a few companies are really dominating the art of reversing "G-code" - One of them is that is extremely difficult to support all CNC controls and machines out there, and they´d have to release lots of incremental patches, quite often, to get remotely close to a tool like VERICUT, NC-Simul and a few others. These companies not only have powerful material removal algorithms, but they also have resources to stay focused in supporting every new technology popping out every day.

 

JM2C

[themachinist]

The simulation engine is developed. It just needs to accept new input. Interpreters already exist. Then it becomes an integration problem. A small problem when compared to developing the whole system from scratch.

 

The "wagon" only has a half dozen people in it. VERICUT and NC-Simul are robust, but they are also very expensive and a lot of the market feels priced out.

 

It would be difficult to support every CNC control and machine out there but nobody does that. Most controls are similar anyway because only a few companies manufacture HMI systems. Haas and Fanuc represent something like 40% of the market. I think making a product that can work for 85% of people is better than making no product at all.

[Guest]

... It just needs to ...

 

 

Software development and software integration are no small tasks for something as complicated as CAD/CAM software.

 

I'm with Watcher. CNC needs to stay focused on theor core business and leave that to the Vericut's, ICAM's  and NC Simul's of the world.

 

JM2CFWIW

[Guest]

... Haas and Fanuc represent something like 80% of the market.

 

Reality fiss. :

 

If CNC were interested in "market" they would have developed their MT product around a FANUC controlled lathe first. They chose Mazak and Okuma. Costly IMHO as those two represent a relatively small piece piece of the Multi-Tasking Lathe market.

 

JM2CFWIW YMMV

[BenK]

In all reality I don't want to rely on one system. I plan on getting vericut in the near future. Yes the cost is up there but even if mastercam had a gcode simulation you would have to pay extra for it. So for me I will use a separate piece of software to validate my program.

 

Than you think about development resources, I think cnc software has far better things to devote there time to. I know many people that have a list of enhancement requests waiting for them to be implemented on top of bug fixes people are waiting on. For me I would rather see them spend their time on the overall interface and the mill and lathe products. That includes mill/turn. The more they get distracted from that the longer the end users struggle to find workarounds.

[Watcher]

The simulation engine is developed. It just needs to accept new input. Interpreters already exist. Then it becomes an integration problem. A small problem when compared to developing the whole system from scratch.

 

The "wagon" only has a half dozen people in it. VERICUT and NC-Simul are robust, but they are also very expensive and a lot of the market feels priced out.

 

It would be difficult to support every CNC control and machine out there but nobody does that. Most controls are similar anyway because only a few companies manufacture HMI systems. Haas and Fanuc represent something like 40% of the market. I think making a product that can work for 85% of people is better than making no product at all.

 

Let me guess.... former ESPRIT or CAMWorks user?

[crazy^millman]

In all reality I don't want to rely on one system. I plan on getting vericut in the near future. Yes the cost is up there but even if mastercam had a gcode simulation you would have to pay extra for it. So for me I will use a separate piece of software to validate my program.

 

Than you think about development resources, I think cnc software has far better things to devote there time to. I know many people that have a list of enhancement requests waiting for them to be implemented on top of bug fixes people are waiting on. For me I would rather see them spend their time on the overall interface and the mill and lathe products. That includes mill/turn. The more they get distracted from that the longer the end users struggle to find workarounds.

 

Ben I agree with your thought on this as well. Might look at NCSIMUL it is really good software.

[Watcher]

There´s an american saying that I love: You get what you pay for.

 

VERICUT, ICAM, NC-Simul, are not cheap stuff. It´s true. But the level of quality they offer and how much money spent in headaches and pills they can save a company´s owner, it´s hard to measure.

 

I work in an Oil&Gas company. Since I joined this team, I´ve worked everyday to make VERICUT simulation flawless and comprehensive. In the past 4 years we had VERICUT at our disposal, but it only started to work effectively in the past 14 months. I work on it everyday.

 

In 2013 and 2014 together we had half hundred of programming mistakes that crossed our walls and a couple of millions in scrap and rework. In 2015 we have so far 1 programming mistake VERICUT could not caught, in a very advanced feature of our WFL MillTurns it still unable to simulate (Mostly because of my fault in reporting it properly).

 

We just finished a Tubing Hanger in full Inconel 718 without a single programming mistake going out of our room. VERICUT saved us dozens of times in this component alone. A finished component like that costs around 130K USD.

 

So "pricey" is a relative word. A crashed spindle or a scrapped component will usually be much more "pricey" and can force a company out of business. In our case, it not only helped us to deliver on time and with quality, but also helped to secure more work from the customer, since not many companies out there can do a Tubing Hanger like that without mistakes.

[Watcher]

Ben I agree with your thought on this as well. Might look at NCSIMUL it is really good software.

 

They smoke VERICUT in a lot of areas. If it was my company I´d start from there.

[MIL-TFP-41]

This was asked on the Mastercam forum...here was the response...

 

 

We do not offer G-Code Verification at the moment but there are links available to various packages such as VERICUT and that topic isn't in our current plans for the future, since there are long established solutions that do a great job. Like anything in the future that could change.

[themachinist]

 

a bunch of baloney

 

Whoa there Foghorn. You took my first quote out of context. I was talking about their simulation engine and speculating about what the requirements for offering g code simulation would look like.

 

You paraphrased my second quote incorrectly. I said 40% market share. Not 80%

 

It is not okay to spin other peoples words to suit your argument.

 

 

Let me guess.... former ESPRIT or CAMWorks user?

 

No. MasterCAM since v8

 

 

This was asked on the Mastercam forum...here was the response...

 

Thanks

[crazy^millman]

They smoke VERICUT in a lot of areas. If it was my company I´d start from there.

 

Why we are representatives for them on the West Coast.

[Matthew Hajicek - Singularity]

 

 

They smoke VERICUT in a lot of areas. If it was my company I´d start from there.

 

Do tell?  I'm helping someone make that decision soon.  Target machines are a vertical with a trunion and a Swiss that may have a B axis, medical parts.

[Watcher]

No. MasterCAM since v8

 

Thanks

 

Sorry - This made me to make a wrong assumption: http://the.machini.st/posts/151

 

Nice blog!

[Watcher]

Do tell?  I'm helping someone make that decision soon.  Target machines are a vertical with a trunion and a Swiss that may have a B axis, medical parts.

 

These days they can do nearly everything CGTech does. And graphical power and speed in NC-Simul is far superior.

 

To be honest, many recent things in VERICUT were pioneered in NC-Simul. Don´t know about the quality of the support. VERICUT´s is one of the best in the industry.

[themachinist]

Sorry - This made me to make a wrong assumption: http://the.machini.st/posts/151

 

Nice blog!

Thanks! We use Esprit for our Integrex and screw machine, but I've been using MasterCAM on and off since I was a teenager. (Dad is a machinist)

[mkd]

i wonder if the naysayers against Code Verification in this thread said the same thing about machine simulation?

we now have machine simulation across the board in CAM.

 

adding code simulation is the natural progression. It's a matter of when, not if, IMHO. We already have Vericut interfaces for popular CAMs, so one could argue we are well on the way.

[Matthew Hajicek - Singularity]

 

i wonder if the naysayers against Code Verification in this thread said the same thing about machine simulation?

we now have machine simulation across the board in CAM.

 

adding code simulation is the natural progression. It's a matter of when, not if, IMHO. We already have Vericut interfaces for popular CAMs, so one could argue we are well on the way.

 

I don't think anyone here would see it as a bad thing, but keep in mind that CNC Software is licensing their verification code; they don't own it.  That means they would have to start from scratch, and dedicate a big chunk of resources that might better be used for other development.

[Watcher]

i wonder if the naysayers against Code Verification in this thread said the same thing about machine simulation?

we now have machine simulation across the board in CAM.

 

adding code simulation is the natural progression. It's a matter of when, not if, IMHO. We already have Vericut interfaces for popular CAMs, so one could argue we are well on the way.

 

I do agree with the "...It's a matter of when, not if..." - But it won´t come from CNC Software IMHO. It will be a third party solution, probably from ModuleWorks.

 

MachineWorks or ModuleWorks, whichever commoditize it first, the other will follow, and then it will become popular, like MachSim did.

[mkd]

past history to predict future, i'd have to agree with Watcher

which begs the question why did they choose module works in the first place instead of a 'real' simulation software?

 

IIRC, PTC had integrated Vericut solution. I don't know why that partnership fell apart or how well the integration worked.

seems they may have been ahead of their time.

 

Vericut already has an NCI file interpreter....

[BenK]

i wonder if the naysayers against Code Verification in this thread said the same thing about machine simulation?

we now have machine simulation across the board in CAM.

 

adding code simulation is the natural progression. It's a matter of when, not if, IMHO. We already have Vericut interfaces for popular CAMs, so one could argue we are well on the way.

I don't think anyone here is against the idea. I will be honest I want gcode simulation, who doesn't. At this time though I would rather see cnc software focus on getting their core product dialed in and 100% functioning before taking on anything else.

 

For example if you had a car that only started 50% of the time would you want the auto maker spending development resources on power windows instead of figuring out why the engine doesn't start? It all goes back to the low hanging fruit theory. I feel gcode simulation is at the top of the tree right now.

[mkd]

I don't think anyone here is against the idea. I will be honest I want gcode simulation, who doesn't. At this time though I would rather see cnc software focus on getting their core product dialed in and 100% functioning before taking on anything else.

 

For example if you had a car that only started 50% of the time would you want the auto maker spending development resources on power windows instead of figuring out why the engine doesn't start? It all goes back to the low hanging fruit theory. I feel gcode simulation is at the top of the tree right now.

well when the other guys ALL come out with the new generation of hideous-barn-door truck grill and nobody wants to buy your old (normal) looking truck .... sales trumps functionality every time when you are running a business and need to survive.

 

edit: all joking aside, i humbly don't buy the argument that gets tossed around about fixing bugs before functionality, in this particular instance, because at the end of the day Programmers make code. Verification functionality that ensures high quality code (a programmers core purpose) trumps all other issues. If there are unfixed bugs and tribal-knowledge workarounds that crop up, good code simulation is the filter and should be #1 on any programmers list.

[BenK]

well when the other guys ALL come out with the new generation of hideous-barn-door truck grill and nobody wants to buy your old (normal) looking truck .... sales trumps functionality every time when you are running a business and need to survive.

And that is the problem. I'm not interested in the flash, I want the functionality to be there above all else. I buy a cam software for its ability to drive a machine not because it looks cool.

[Colin Gilchrist]

For what it's worth; you can have had the ability since X7 or X8 to fully "hook up" the ModuleWorks Machine Simulation to read the "input" as "post processed code". Here is the thing: it doesn't support full "G-code" simulation, but it gets you close.

 

What do I mean by that? The ModuleWorks Simulator Engine reads a "ModuleWorks CL File" for it's input. It does not read G-code. That said, it is possible to modify your Post Processor to output both a NC Code File, and to also output the "ModuleWorks CL File" at the same time. There is basically a directly correlation between the G-code moves, and the moves in the CL file. If your post happens to output a 5X "unwind" move, then MachSim would still simulate this movement.

 

The place ModuleWorks hasn't gone yet, is a full "G-code interpreter" like NCSimul and Vericut have. So if you have a G-code or M-code that cause motion, you would have to figure out a way to output the "machine motion" as ModuleWorks CF File events. Not impossible, but quite involved...

 

The big reason that only a few Post Developers even offer this as a service is that any Post Developer that "integrates ModuleWorks CL output" into there post is supposed to pay ModuleWorks a Licensing Fee. (50% of the cost of the service.)

 

So while it is technically possible to get 90-95% accurate simulation (maybe closer to 99% for something like a 5X trunnion without any "special" features), not many people are doing it, but there are a few...