Cut Rifling Machine Help

This is the basic G code program that I've sort of come up with that makes it move roughly to how we want, with the variables listed like you said.


We have tried this and technically it will rifle something but I'm not too sure how consistent the rotation axis of the tool is locked in with the carriage travel along the barrel. Plus setting the tool zero on the barrel points is not easy as such.  

The X and Z axes should move exactly in synch, that's pretty much the first requirement for a CNC machine.

Personally I would have configured the axes as X and A or maybe Z and C, but that has no bearing on basic functions. Though in either case there is some question about feed-rate calculations. It is probably better to use inverse-time feed rate and calculate a cycle time based purely on barrel length. But that is a detail for later.

For maintainability it might be good to switch internally to named parameters: for example.

If calling the routine from an action button in the GUI the numbers come in on #1, #2 etc. It's probably best to load the internal parameters from those immediately. Though there is the option for the routine to "pull" the values rather than have them "pushed" (Assuming that the Glade GUI is called "rifling-gui" and has a HAL spinbox called "barrel-length")

Yes, I was thinking of renaming the axis for ease of understanding but only quickly wrote out the program just to get basics going. It will be refined as the project proceeds and really this is just proof of concept first. Tbh i've been talking to some senior programmers here and they recon we can minimize the program even further with a few "if" "GOTO" and "while statements, if Linux CNC can run them ? However, that is getting beyond me with my G code knowledge so will see. 

Agree as well, deff start to name the parameters for what they represent and guessing like you said, be simpler to connect them up to the GUI. Still got loads to learn there. 

Plan of attack is to get the program running well and fairly refined to the finished article and then try and connect the parameters to a GUI.

Also got to try and put a pause button in the program to stop the machine at a very specific point to use a dia-test gauge to measure the depth of rifling and then a resume button to continue on from where it is. 

On a separate note i have wrote down a flo chart, as such, for my self to understand how i want the machine to operate and be a sort of instructions guide for new users. Ill post it on here when i have a chance.      

SPC89 wrote:

i've been talking to some senior programmers here and they recon we can minimize the program even further with a few "if" "GOTO" and "while statements, if Linux CNC can run them ?

Yes: linuxcnc.org/docs/2.8/html/gcode/o-code.html

Also got to try and put a pause button in the program to stop the machine at a very specific point to use a dia-test gauge to measure the depth of rifling and then a resume button to continue on from where it is.

That sounds like a job for the M0 code: linuxcnc.org/docs/2.8/html/gcode/m-code.html#mcode:m0-m1

Greetings,

Back to the machining process...

Is this a "cut" rifling process? Not a button rifling?

If cut, does it have multiple teeth to cut the depth in one pass?

Does it cut all the grooves in one pass; or, one at a time?

If single groove cut, how is the cutter returned to the beginning of the part?

This project is most interesting to me and I have been following it closely.

Cheers,

Dave

Thanks, the links are a great help. See what i can come up with.

Yes it is a cutting process. The tool is removing about 0.0001 every pass it makes. Once it has done a cycle of grooves say 4 it then gets raised by a mechanism that rotates a screw in the tool that pushes on a wedge that raises the cutter up. 

A rifling button doesn't cut it swages the metal to the right shape. Puts alot of stress in the metal so we heat treat it afterwards. 

below is a cut rifling tool we use on our machines. A piece of carbide is used to do the cutting work and ground up with correct angles for any given twist as such.