The importance of "cut rules"

I know that a few people on the forum are more into cnc plasma because they like to tinker, or are in it for a hobby like 3d printing, so I figured that I'd share some fine tuning tips. While it is something that you may figure out on your own as you go, I figured I'd share my experience because it's an area where I've already been there, and done that. I may not be the most tech savvy person in the world, but I'd like to at least contribute what I can when I can. I've been using a cnc plasma table for my business for a few years now. My old machine had a Miller plasma cutter so I didn't get book speeds and thc readings like I do with my current hypertherm. I had to teach myself EVERYTHING by trial and error(this ended up teaching me a lot in the long run). I ended up with a new material process that allowed me to dial in a new material in less than 5 minutes with good results (this took me hours for each material when I first started out). I have spent countless hours studying and experimenting with different variables in the plasma cutting process to produce the best possible cut quality that my machine is capable of. One thing that makes a huge difference is the ability to write cut rules. My old machine ran on Mach 3 which cut rules would not work for since the thc wasn't integrated into the software. This was one of the main reasons that I built my new machine and went with Linux. I should add, that my experience with plasma isn't one of a hobby, but for my business. I have countless hours running the machine and I would like to think that I have a pretty good knowledge of the process. I was able to, very successfully, cut anything from .060 Al to 7/8" steel with my little 40 amp Miller on my old machine which was only capable of 170ipm maximum (ballscrew machine). I'm just getting started with my hypertherm so it will take me a little while to get the machine dialed in. I wanted to show a side by side of what cut rules can do. I'm still working on a couple of good sets of cut rules to use that I will share when I'm done. Everyone has different plasma cutters, but you should just be able to go in and change a few numbers if need be. On the left is a picture of 3/8" plate without cut rules on a tight corner. The arc doesn't have time to catch up, and results in a pointed corner. On the right is a picture using cut rules to slow down a small distance before the edge, turn off the thc so the torch doesn't dive, and turn it back to full speed and thc back on a short distance after the corner. This helps keep the edges square and true rather than pointy and ugly. This is where a lot of finer detail can be made. You can keep higher tolerances, especially in tab and slot designs where it's extremely important to be able to do so. If you have a good single set of cut rules, you can apply them very easily in sheetcam to almost all of your cuts, but I'm still going to try to make 3 sets. One for thick material, one for thinner material, and one for cutting stainless due to the long trailing arc that is caused when cutting that particular material. I still have some dialing in to do on the cuts, but this is the very first cut in 3/8" hr on my new machine, so it's off to a good start.

Good advice. You should also try this without disabling the THC because Plasmac has a corner lock (or velocity anti dive) so that once you slow the cut speed down, the corner lock will kick in and disable the THC as soon as cut velocity falls below the threshold (say 90%)

I haven't tried that yet due to the fact that I think that the thc will lower the z prior to locking out the thc since it's so responsive. I'm also unsure if plasmac, like other plasma software, is reading the feed from the material settings or from the gcode. For example, if I was cutting at 350 ipm and I put in a cut rule to change to 90% of the feed rate, is the thc off because it slowed to 90% of the feed rate, or is it still on because the machine is matching the feed that you wanted. I'm not sure if I'm being clear in the explanation of my reasoning there. I also find it interesting though that you were upset at the idea of adding .005 seconds to your pierce time the other day, but are ok with allowing the cut speed to drop by 10% before the thc is temporarily locked out. Cut speed and accurate thc corrections are probably two of the most important things out there. I have really high acceleration values on my machine, so chances are that it wouldn't correct too much, but I'd like to be able to control it the best that I can. I'm not saying that it wouldn't work, it may work just fine, but I'd like to have control over it as much as possible. (especially with how fast and responsive the mesa thc is!) Running plasmac has been a huge improvement in all aspects to how my previous machine ran, I look forward to using it more, and trying to dial in my process as best as possible!

If you have the feed rate in gcode set as F#<_hal[plasmac.cut-feed-rate]> then it uses the cut speed from the material as its target feed rate.

Thank you! I figured that's probably where it pulls it from, I just wasn't sure.

CBEAR wrote: I haven't tried that yet due to the fact that I think that the thc will lower the z prior to locking out the thc since it's so responsive. I'm also unsure if plasmac, like other plasma software, is reading the feed from the material settings or from the gcode. For example, if I was cutting at 350 ipm and I put in a cut rule to change to 90% of the feed rate, is the thc off because it slowed to 90% of the feed rate, or is it still on because the machine is matching the feed that you wanted. I'm not sure if I'm being clear in the explanation of my reasoning there. I also find it interesting though that you were upset at the idea of adding .005 seconds to your pierce time the other day, but are ok with allowing the cut speed to drop by 10% before the thc is temporarily locked out. Cut speed and accurate thc corrections are probably two of the most important things out there. I have really high acceleration values on my machine, so chances are that it wouldn't correct too much, but I'd like to be able to control it the best that I can. I'm not saying that it wouldn't work, it may work just fine, but I'd like to have control over it as much as possible. (especially with how fast and responsive the mesa thc is!) Running plasmac has been a huge improvement in all aspects to how my previous machine ran, I look forward to using it more, and trying to dial in my process as best as possible!

I guess pre-plasmac it was drummed into me not to filter any pid based process control variables. Plus people were wanting to debounce values without any evidence of signal bounce and I have always made evidence based decisions with plasma. You could set your corner lock velocity threshold to 99% if you wanted to.

But if your gcode reduces the velocity by say 60% the lockout will be pretty well instant with good acceleration.

I also have high acceleration now (0.5g) and should get my gantry ends stiffened up to get more out of my machine. I have done some work so that Linuxcnc has pins for the arc radius and I want to look at much more sophisticated cutting models. We should be able to calculate the minimum arc radius that can be negotiated by a machine with a given acceleration at the current cut velocity. If my ideas are right, there should be no need to add any gcode rules as they will all be applied in hal automagically.

I will forever say "automagically" form now on and claim it as my own.

rodw wrote: Plus people were wanting to debounce values without any evidence of signal bounce and I have always made evidence based decisions with plasma.

Sometimes you can be really helpful, and sometimes you insist on being intentionally obtuse and whingeing.

snowgoer540 wrote:

rodw wrote: Plus people were wanting to debounce values without any evidence of signal bounce and I have always made evidence based decisions with plasma.

Sometimes you can be really helpful, and sometimes you insist on being intentionally obtuse and whingeing.

Yeh, Its an Australian thing. Some of us call a spade a shovel interlaced with some occasional sarcasm...