Cam Grinding coding

[vmax549]

HIYA ART, Have you ever thought about a CAM grinding function. It would go over BIG in the modeling world.  I have worked on one but cannot get though all the math needed. I have the math just can't get it all together in code.

(;-) TP

[ArtF]

I do seem to remember having a discussion with you at one time on making
grinding code....something about it being complicated by the radius of the grinding wheel
or something?

  Cant say I see myself doing one , at least until a ream of other plans are complete. I have to redo the CNC output section entirely so Im not oppossed, but Id have to say Ill wait till I see light in the tunnel before taking on more plans. :)

Art

[JustinO]

What math do you have, and what math do you not have? Where are you having trouble? Worst case there'd be some numerical methods and/or simultaneous trig equations. Are your equations transcendental?

Sherline makes and sells a cam grinder that comes with software to generate G-code from CAD. It's expensive because it's a single purpose machine, but if you could keep it busy, it's a good value.

You've probably seen these Sherline videos:
Part one:
http://www.youtube.com/watch?v=OIWyGtXEgR4

Part two:
http://www.youtube.com/watch?v=PlZDg0rpIi8

[ArtF]

Hi Justin:

  I think the math is pretty straight forward, its a simple circle/circle intersection
algorithm really.  Watching the video's brought it all back. :) .
I'm not sure many would have a rig for doing it is more the reason it'd be
lower down my list. I wonder how many really desire cam grinding capability,
perhaps its my own inexperience in such matters. Any out there also interested
in cam grinding as an output?

Art

[JustinO]

Hi Art,

Once you work out the positions and tangent angles of the interpolating points along the perimeter of the cam (not fun), it gets really easy.

These positions would be defined in the cam's coordinate system.

Vector add a grinder radius length vector to each point's position vector, perpendicular to the cam's surface, and calculate the angle between this sum vector and the grinding wheel's axis. The two values plop right out of this vector sum triangle -- the angle for the rotary axis holding the cam, and the position of the linear axis moving the grinding wheel.

I should make a drawing...

Maybe you'd sell a few licenses to these cam grinding modellers.

--Justin.

[ArtF]

Justin:

Doesn't sound too bad to do.. but Id wonder how many have cam grinding ability in their
shops. I admit, itd be a cool thing to be able to do,but it may be a bit esoteric for rationalizing the
development time for such a thing....not that it takes much rationalization round here..loll


Art

[JustinO]

It's all about bragging rights.

[ArtF]

good point...

[vmax549]

THERE are a LOT of old Profile Plate cam grinders out there that are waiting on a decent CNC interface.(;-)

It would not be Just for camshaft grinding but Cam lobes in general.

Long ago I had a version working from out of Sheetcam Les helped me with the post. We used the Function of a XY drawing of the Lobe then ran the TOOL around the profile THEN converted the X to A on post. THAT gave you the basic profile. You could STACK the lobes in layers to do an entire V8 camshaft

In mach3 I then used  SUBs to drive the program and stepped into the lobe each loop. You could also do alternate lobes each loop to equalize the wear on the Wheel across multiple lobes. At one point it did just under a Million lines of code to do a V8 camshaft.

Then you also have Crankshaft grinding where you grind the journals. Basic offset grinding there.

THere ARE a lot of peole doing miniture engines these days (;-) .

A cam crank grinder can be maded very easily OR just convert your mill with an Aaxis.

Just a thought, (;-0 TP

[vmax549]

Here is a sample of one of the program to gind a 2 lob camshaft. NOTE the cross over between the 2 lobes. after each loop around the lobe it trnsfered over to the other lobe did one loop around it then return to the first lobe

[ArtF]

Thats clever....

Art

[Mooselake]

2.5D cam gcode could be handy for Kinetic and Rube Goldberg type devices, although that to-do list may already span too many volumes to fit on your (virtual) bookshelf.  Isn't a grinding wheel basically a really big cutting tool, so the output would essentially be the same?

While I don't remember the reference, and google fell short, I read some time back in one those old out-of-copyright books that if gear making hadn't become "easy" that cams would have had a lot more prominent part of machine design.

Kirk

(OT) Really like the celtic knot laser plugin.  Any plans to make a routable version?  Expand it into a CNC Rose Engine?

[vmax549]

Cam grinding SEEMS simple same with Crank Grinding , BUT IT AINT.  When grinding a lobe on a rotary axis the point of contact of the WHEEL changes with the profile of the lobe as it rotates.  There is a formula or three to correct the Z(grinder infeed) so that it translates correctly as the A (rotary) rotates.

Other wise the lobe you think you ground is NOT correct to the original profile. ]

ALSO as the wheel wears it must be compensated as well. That was the reason for cross over grinding a lobe pair( or as many as you select, you can do the entire camshaft at one time if you wanted . It just takes TIME to GRIND.

Just a thought, (;-) TP

[ArtF]

>>(OT) Really like the celtic knot laser plugin.  Any plans to make a routable version?  Expand it into a CNC Rose Engine

  Actually, I intend to add an engraving module to the CNC grouping, so I first needed
an engravable object and celtic knots sounded good to me for that. The knots will appear
shortly..

Art

[Paul Rusconi]

Hi
There is a nice cam grinding program here that is worth a look http://gcam.lucasemail.org/

GCam Forum
https://groups.yahoo.com/neo/groups/GCam_CNC/info

Paul