Discussion: Who is responsible for the magic behind CAM software?

[ThickChips]

Hi, 

Want to open a fun discussion on the witchcraft we call CAM software. And specifically, the topic is, who solely or in part is largely responsible for the majority of the mathematics behind CAM toolpaths. We think about 5 Axis simultaneously toolpaths. A lot of math is involved. 

 

Can we attribute this technology to a specific engine, or set of engines? And can these engines be attributes to the works for a few specific people: engineers, mathematicians. 

 

What can YOU offer this discussion?

[Leon82]

It appears that module works licenses their paths to cam software companies.

They may have a patent on it. So if that's the case you either pay the license fee or you have to buy the company for the patent.

 

I should let you know I'm only a self declared expert. ( The best kind imho)

[#Rekd™]

The list of companies licensing ModuleWorks technology.

https://www.moduleworks.com/company/partners/

 

[Aaron Eberhard]

Yes, there are computer science, math majors, and guys who just started on machines and understood what needed to be done and figured it out behind all of this stuff.  All of the CAM software has their wonks behind the scenes doing that.  If you want to meet the people behind the curtain you need to be pretty specific about which toolpath you want to know about. For example, there's a different guy behind Mastercam's Dynamic than, say, Equal Scallop.   And each toolpath is really a team effort anyway.

ModuleWorks in particular was started by Yavuz Murtezaoglu, and he's the guy who wrote the original 5 axis engine for it which does all the surface-based stuff (morph between 2 curves/surfaces, parallel to curves/surfaces, etc.).   He's not the guy who actually coded the newer geodesic engine stuff (in Mastercam's Unified toolpath as guides, automatic mode, or Morph/Parallel set to "exact" mode).  He's also not the lady behind the Swarf Milling.  They now have a lot of programmers!

Depending on how far down you want to go, the guys who apply this stuff to make toolpaths are often standing on the shoulders of giants in the fields of Mathmatics.   Many toolpath ideas are taken from mathmatics and physics white papers.   For example, Yavuz cites Dr. Bronshtein's Handbook of Mathmatics as a fundamental basis for his original toolpaths.

[RonC.]

I found this on Goggle... lol

 

Who invented CAD CAM technology?

 

 

engineer Pierre Bézier

 

The invention of the 3D CAD/CAM is attributed to French engineer Pierre Bézier (Arts et Métiers ParisTech, Renault). Between 1966 and 1968, after his mathematical work concerning surfaces, he developed UNISURF to ease the design of parts and tools for the automotive industry.

 

[#Rekd™]

This gives a deeper history of the CAD side.

https://www.cadazz.com/cad-software-history.htm

[pete_hull]

René Descartes of course, but the game changer was Douglas Ross, creator of APT at MIT in the 50's- the 'preamble' to all CAD/CAM that followed

 

[nickbe10]

12 hours ago, pete_hull said:

the 'preamble' to all CAD/CAM that followed

Yep, I learned programming on an APT system.

Unbounded geometry, you just can't beat it.

3 plane lock, for a long time (if not still) this was the most accurate way to machine a surface.

If I had my druthers, I would run NCL. But I would never advise anybody to install it in their shop. NOT user friendly and good NCL programmers command very high salaries and are as rare as rocking horse manure.

As Pete says, all the so called "multiaxis kernels" are simply a GUI interpretation of the original APT vector matrix. 

[pete_hull]

3 hours ago, nickbe10 said:

 

If I had my druthers, I would run NCL. But I would never advise anybody to install it in their shop. NOT user friendly and good NCL programmers command very high salaries and are as rare as rocking horse manure.

As Pete says, all the so called "multiaxis kernels" are simply a GUI interpretation of the original APT vector matrix. 

Programming with NCL  on an SGI  was the best, absolutely loved it.

[So not a Guru]

20 hours ago, #Rekd™ said:

This gives a deeper history of the CAD side.

https://www.cadazz.com/cad-software-history.htm

This is a fascinating read! Thanks

[cruzila]

"Also in the mid 1960s, the Digigraphics division of Control Data Corporation released the first commercially available CAD software system. The system was a successor to ITEK's earlier CAD software research system (which was named "The Electronic Drafting Machine" and ran on a Digital Equipment Corp. PDP-1 mainframe computer) and as with the Sketchpad CAD software, input was made using a light pen. Digigraphics was priced at $500,000 per unit and only a very few units were ever sold."

 

The first pen driven system was $500,000 in todays dollars that's $4,511,359