Considering NCSimul, would like input

[jlw™]

I am looking at G-code verification.  After playing with the NCSimul "File Player" I am wondering if it is better than mach sim in Mastercam.

I'm concerned about how the machine/control definitions are setup.  If it is not correct, it's not better than mach sim.

 

I would appreciate input from any one who has been on the build it side for a machine they currently do not support.  The machine I am looking at this for is a 6 axis head/head Mazak V100.

 

Thanks!

[Watcher]

13 minutes ago, jlw™ said:

I am looking at G-code verification.  After playing with the NCSimul "File Player" I am wondering if it is better than mach sim in Mastercam.

I'm concerned about how the machine/control definitions are setup.  If it is not correct, it's not better than mach sim.

 

I would appreciate input from any one who has been on the build it side for a machine they currently do not support.  The machine I am looking at this for is a 6 axis head/head Mazak V100.

 

Thanks!

Very unlikely there's a machine setup they never covered before... It's a pretty decent product and used in demanding niches as well.

I'd say go for it if that's what your guts are telling you. 

One thing I like about Vericut is that I've freedom to change and create new machines. It's well documented as well. I'd look for the same in NCSimul. 

A lock in in any development front from an user perspective is a no go to me.

Two areas where I'm sure it beats Vericut is graphical performance and parsing of NC code.  

[jlw™]

3 minutes ago, Watcher said:

A lock in in any development front from an user perspective is a no go to me. 

Daniel, I don't understand this statement.  Can you reword?

 

NCSimul tells me I can build my own machines and control defs and they have training specifically for it.  It can also be used on any computer.

[Watcher]

1 minute ago, jlw™ said:

Daniel, I don't understand this statement.  Can you reword?

 

NCSimul tells me I can build my own machines and control defs and they have training specifically for it.  It can also be used on any computer.

What I said is that if any type of lock in was present, I'd not go ahead. 

But it's clear now that they assured you no lock in would exist for you. And that's all you had to hear. 

Vericut single platform for example is an entry point for some companies, but they have zero possibilities to tweak their configuration files. 

[jlw™]

I looked at Vericut and at the time we couldn't spend the money on it.  If I had known about NCSimul then I would have bought it for the price they are asking now.  I have now developed/proven my post out in a trial by fire manner and trust it really well but I am looking to make going forward easier.  Mostly to catch my little human errors.  When you have 500+ ops and 300+ hour run times I am bound to and do make simple mistakes.  I want to eliminate this before hitting the machine.  I think that would be money well spent for my company.

[Watcher]

Nothing wrong about finding an alternative for VERICUT. I don´t know the price tags for the two producs in U.S., but I couldn´t care less.

NCSimul is a decent product and their strategy to penetrate in U.S. market is to have a more affordable price. It´s a business strategy and it does not mean they have a lower quality product.

I´m a firm believer and proponent of CAV in our industry. Once I implemented it here, I reduced our rework rate in NC-Programming from 17% to less than 2% in 3 years. NCSimul could have achieved the same I´m sure. It´s the CAV technology and not the brand behind it.

If you company achieve the improvements you aim and you feel you guys are spending the right money on it, that´s what matters.

I think NCSimul is a great product and if their support is as responsive as CGTech´s, you will get a pretty good bang for the buck.

[RaiderX]

I would check into support, other users etc.

also check out there licensing model very carefully. it was a deal breaker for us when I last checked it out but??

product itself looked very capable .

 

Doug

[jlw™]

Thanks Doug, that's insight I'm looking into right now and seeking input on.  From what I hear support is good.  I also know some one who worked for CGTech and I'm not loving their business model or how they treat employees but to be honest the price point is what has me interested in NCSimul.  In a direct comparison of quotes laying on my desk it is 1/6 the cost.  I'm more concerned about them getting my machine correct.  For lack of better words I will be the guinea pig on this machine and I want it to do it's job and save me time, not cost me time.

[Watcher]

3 minutes ago, jlw™ said:

Thanks Doug, that's insight I'm looking into right now and seeking input on.  From what I hear support is good.  I also know some one who worked for CGTech and I'm not loving their business model or how they treat employees but to be honest the price point is what has me interested in NCSimul.  In a direct comparison of quotes laying on my desk it is 1/6 the cost.  I'm more concerned about them getting my machine correct.  For lack of better words I will be the guinea pig on this machine and I want it to do it's job and save me time, not cost me time.

Out of curiosity - What machine?

[Matthew Hajicek - Singularity]

1 hour ago, Watcher said:

What I said is that if any type of lock in was present, I'd not go ahead. 

But it's clear now that they assured you no lock in would exist for you. And that's all you had to hear. 

Vericut single platform for example is an entry point for some companies, but they have zero possibilities to tweak their configuration files. 

I'm using Vericut Single Platform.  If you've only got one machine to verify for it's great.  You can always email support if you need your machine definition tweaked and they're pretty good and quick about it.

[jlw™]

38 minutes ago, Watcher said:

Out of curiosity - What machine?

Mazak V100, 6 axis, head/head

[Watcher]

3 minutes ago, jlw™ said:

Mazak V100, 6 axis, head/head

I´m sorry... you said that before...

Don´t see how they could not. This machine has a RAM, which demands support for collinear axes in the CAV system. They could not emulate HBMs without this support.

They´re Mazak partners as well. Chances are close to 0 that they could NOT support you machine.

 

 

[Watcher]

I just checked again and it seems this Versatech model does not have a RAM, only the two Z columns. Is that correct?

If so, there´s no way NCSimul cannot cover it well. They support a lot more complicated kinematics, like Scharmann ECOSPEED Sprint Z3 parallel kinematics head, or Zimmerman´s ABC heads.

 

[Watcher]

As a general rule, I´d say that any software that can emulate robots reliably from NC code can drive nearly all machine tool models available.

Robots demand a mix of nearly everything.

[crazy^millman]

NCSIMUL can support it. The fact I can drag and drop my models, clamps, work holding and machine inside of the session is a time saver. The fact I can paste code into an in process verification that had already had 11 millions lines of code run through it and I just need to check another 100,000 lines of code and pick up right for where I left off doing so another time saver.

Vericut is a good product, but it needs a rewrite. Java served it's purpose, but needs to move into newer stronger technologies to do CAV. They have a great support staff and I offer world class support.

NCSIMUL is growing their presence in the US and making good head way. CAV is growing in it's role in American Manufacturing. NCSIMUL, Vericut, CAMPlete and ICAM should all be seeing an upswing in sales. Biggest reason is the lack of skilled people anymore to do the complex work. I am really scared for the industry right now. Hopefully with the investment in CAV it will help bridge the gap we are seeing in today's Manufacturing. Time will tell where we are 5 years or 10 years from now, but right now nothing I can see is really making a dent in bridging the skills gap.

[Watcher]

19 minutes ago, C^Millman said:

The fact I can paste code into an in process verification that had already had 11 millions lines of code run through it and I just need to check another 100,000 lines of code and pick up right for where I left off doing so another time saver.

This is really the beautiful part. We do BIG parts on BIG MillTurns and the need to reset is indeed a pain in the xxxx. In-Session files are more a pain than a solution I think.

I really hope CGTech could add this capability to VERICUT. No chance to change systems here - VERICUT is very well customized for our needs, we have dozens of sites using it and it does an excellent job anyway. I´m sure it beats NCSimul in some fronts too.

But if I was picking systems, that part would really have a big influence on my call. I hope CGTech Product Management is reading this.

[crazy^millman]

Just now, Watcher said:

This is really the beautiful part. We do BIG parts on BIG MillTurns and the need to reset is indeed a pain in the xxxx. In-Session files are more a pain than a solution I think.

When you have 20 or 40 IP files and they alone take up almost 40gb on your hard drive yes it get to be a pain real quick.

[Watcher]

18 minutes ago, C^Millman said:

Vericut is a good product, but it needs a rewrite. Java served it's purpose, but needs to move into never stronger technologies to do CAV.

I don´t think Java is their limitation. I think their are bonded to decisions that made sense 20/30 years ago and that today are dragging their feet. Their graphical engine is one of them. Z Buffer Technique to represent pixels made sense in an era where graphic cards were a luxury.

NCSimul was written on the top of OpenGL since day one. That is the real deal.

If CGTech change their graphical engine, I think they could have problems to open files from previous versions. But that´s a half baked hunch as well because all a VcProject file is is the organization of a collection of files in a certain hierarchy. They don´t store binary data like a regular CAD/CAM system.

[gcode]

9 minutes ago, Watcher said:

I don´t think Java is their limitation. I think their are bonded to decisions that made sense 20/30 years ago and that today are dragging their feet. Their graphical engine is one of them. Z Buffer Technique to represent pixels made sense in an era where graphic cards were a luxury.

NCSimul was written on the top of OpenGL since day one. That is the real deal.

If CGTech change their graphical engine, I think they could have problems to open files from previous versions. But that´s a half baked hunch as well because all a VcProject file is is the organization of a collection of files in a certain hierarchy. They don´t store binary data like a regular CAD/CAM system.

Yes.... the last time I went to a Vericut class it was taught by one of the founders of Vericut

The first day of class was spent on the history of Vericut graphics and why it's really not as bad as it really is

I think Vericut is in the same boat as Mastercam was in 2000... Mastercam had a DOS product shoehorned into a Windows world

They've been at it 17 years now and still have a ways to go to leave DOS completely behind  ( the old legacy 3d surface toolpaths)

VC's graphics is rooted in antique technology and will require a ground up do over to fix.

A ground up do over that still supports 30 years of legacy product is a very difficult task, but the sooner they start it, the better

[Watcher]

Not sure if everybody knows what the Z-Buffer technique represents in the CAD/CAM graphical world, but I´ll try to explain it in a very simple and short way:

It calculates the color and position of pixels depending on their depth in the screen, so depending on the orientation of you part, its surface colors changes and have that shadow effect due to Z-Buffer. This is a very very simplified explanation.

Z-buffering

In VERICUT, the position of each pixel in the screen is stored in an internal database. Position, depth, color, etc. When you pan, zoom or rotate a part, VERICUT needs to calculate the new location of that pixel using matrices. Therefore, it needs to be on a pause to do that. And that is also what makes it to require refining in order to have an accurate view, which has the side of effect of being time consuming depending on the complexity and size of your part.

It used to be a very effective way to handle graphics long time ago. It´s not uncommon to see VERICUT graphical performance to be better with cheap cards as they handle pixel based stuff better than high end cards, which are optimized for OpenGL and cutting Edge 3D graphics technology.

 

[Watcher]

9 minutes ago, gcode said:

Yes.... the last time I went to a Vericut class it was taught by one of the founders of Vericut

I think you are talking about Jon Prun. AFAIK he is the solely founder and owner of CGTech.

[gcode]

15 minutes ago, Watcher said:

I think you are talking about Jon Prun. AFAIK he is the solely founder and owner of CGTech.

could be.. he was a very nice guy, very knowledgeable and totally into Vericut 

[crazy^millman]

4 minutes ago, gcode said:

could be.. he was a very nice guy, very knowledgeable and totally into Vericut 

I agree a 1000%.

[Watcher]

22 minutes ago, gcode said:

could be.. he was a very nice guy, very knowledgeable and totally into Vericut 

This one G?

 

[jlw™]

There's no doubt in my mind they can, they currently do not have it in their repertoire yet.