Feature Request: ramp piercing (aka moving piercing)

I suspect that the second run of pierce only should have a shorter pierce delay. I do remember at the time it was developed people saifd to use a time delay which I thought Odd but I never needed to use it.

James found wiggle pierce very easy to implement. He added a pin to plasmac.comp that prevented the motion from being disabled during piercing. This allowed XY movement during the pierce delay.

Ramp piercing is much more complex as he said the THC was disabled when Z axis movement is disabled. I was chatting to him today and we decided XY motion could be done the same way as wiggle pierce but the Z axis motion needs to be done using external offsets.
I thought this could make it a candidate for adaptive feed.

I found it interesting to see the Hypertherm doc shared earlier ran at a slow speed initially to gouge a trough for dross capture before ramping down.

I used it for 1/2 inch mild steel, and as you mention the old consumables makes a bigger holes, and then with the new material arc never transfer because the lack of material, what I did is a little (les tah 0.125 inch) offset in the X or Y axis, to "simulate side entry" it worked for me so may be it can be inplemented in this feature.

grijalvap wrote:

One thing i saw from pierce only, is that when you pierce the first time(with old consumables) it works good, but then you change the consumables an start again, but the arc never transfer because there is not metal but a hole, how do you deal with this?
 

I know what you're talking about. I use Pierce Only mode on 1" material always, and even on 5/8 sometimes. It's a great feature I wouldn't want to be without; a feature that's also found on industrial controllers.

Here are 2 workarounds you can use. (re-posted from my answer to a user's question over on plasma spider; if it's not against the rules, I can provide a link here)

1. Change the X or Y zero position ever so slightly (about 1/16") after doing pierces. This way each pierce in your cutting mode pass will still be able to blow down through the hole pierced formerly, eliminating the blowback, but the torch will be off to one side ever so slightly, meaning that it will be cutting material and the ARC OK signal will activate. You could also bump the plate a tiny bit to physically move it on the table, achieving the same result.

2. Toggle "Ignore OK" on at each pierce, then toggle off again after motion has started. (Based on my observation, IGNORE OK has to be disabled for torch height control to activate) Sort of a pain, and you have to stay there and babysit the whole cut. But with thicker materials you tend to need to hang around and watch for issues anyway.

Seems like on the second pass, the piercing should be modified to move a set distance (eg your 1/16") at the beginning of the pierce.
There are clearly several gotchas with qtplasmac's piercing when people use it in a more industrial setting. It would be nice to have the flexibility to implement industry standard piearcing algorithms which was really why I raised this thread.

Perhaps a better option would be to move over a small distance in the "pierce only" mode -- do it on the first pass rather than the second? Then pierce at the correct spot in normal cutting mode? That might be easier to reason about than trying to build in a pierce offset IF pierce only mode was used first, and keep track of whether or not that mode was used... just thinking out loud.

The trick of slightly moving the sheet by bumping it with a hammer after pierces are done is something I learned from a friend with a commercial table with a hi-def plasma head and some commercially available controller -- perhaps Hypertherm EDGE, not sure. I was surprised to learn that he relies on pierce-only mode for thick materials too, even if he has a much more capable plasma unit than mine. I figured out that with qtplasmac it's easier just to slightly change work zero position, especially if you have a full 5' x 10' plate of 5/8 or thicker material on the table; this doesn't slide easily!

So if functionality for offset piercing in pierce only mode would be implemented, once again QtPlasmaC would be ahead of the commercial offerings. Personally I don't mind this workaround though. I think I'd be more interested to experiment with circular-motion piercing.

Looking outer the thread thought a quick update since Rob had mentioned aspects of my tinkering. 

1. have basic wiggle pierce working.  This is leveraging sheetcam to generate the movement code.  I created a new pin in plasmac to notify it of the type of pierce being undertaken.  This allows internal logic to allow motion during the pierce delay.

2. I have the beginnings of a ramp pierce based on hypertherm’s description of a moving pierce.  I have the delays/pauses in place along with the extra concept of pierce end height.   Next is the different feed rates through gouging and crawl/intermediate. Crawl being the speed used before transition to cutting.

the idea is to use a new/remapped gcode in python to set the pins with the needed values and to enhance the sheetcam PP to know when to generate this needed gcode and associated parameters. 

Github link for anyone interested. 
GitHub plasmac-pierce



cheers - James.