Kerf compensation for cnc routers: take cutter diameter, divide by two. Offset by this amount. That is all Vcarve is doing in generating G-code for any cnc router, that is all the Shaper origin is doing. anything beyond that is manual.
Kerf compensation for laser: take beam width, factor in speed, material, power, material thickness, acceleration, deceleration, the terminal velocity of a sparrow…
You can’t insult me over grammar, spelling, or sentence structure. I’m dyslexic and well aware of that.
If you feel insulted or threatened by anything I have written perhaps you should reread your own post and reflect some?
Lasers use heat to vaporize materials and the nature of the process leaves an angled profile kerf cut, but also depending on variations within the material the kerf width is not as precisely predictable. Materials don’t always cut/melt/vaporize the same way within the same sheet, much less within the same production batches. Two cuts in maple hardwood for example, using identical settings, can vary from .004" to .008" (numbers I have measured from my own projects).
With a spinning tool, the profile shape of the cut takes on that of the cutting tool, and other than bearing/spindle runout within the tool itself, or tear-out of the material at the cut, the kerf is consistent.
Wrong, sorry. The kerf may vary based on speed, power, thickness, frequency, beam width, focus, temperature, ambient gasses, lens clarity, alignment, distance from laser emitter to material, moisture content, and any other nonsense you care to list. Compensating for that kerf doesn’t involve anything other than dividing the kerf by 2 and offsetting by the quotient.
Also wrong, sorry. Closer to right than @markevans36301, but still wrong. If I cared to perpetuate this discussion into the realm of pedantry that you’re introducing I might mention a couple other factors that can affect the kerf left by spinnies. But once you’ve settled on a kerf value that you can live with, compensating for it is as simple as offsetting by half the kerf width.
So, I guess what you are saying is that we’ll never have kerf correction because it’s too complicated and can’t be predicted?
I mean, a variation of .004"/2, how ever would we be able to stand such a thing?
I have no worries though, since this was an advertised feature I am confident Glowforge has already figured it out and just as soon as everyone has their units they’ll get right on getting it to us. (SUPER CHEERY FACE)
There is also tool wear, conventional vs climb cutting, work piece and tool deflection… So just as many issues as a laser has if you really want to break it down.
But that all goes into choosing your kerf; all the actual software has to do is take whatever diameter you say your kerf is and offset by that amount.
All correct, he did say run out and I count that the same as tool wear. All these mater to one degree or another but to my knowledge no one is even trying to automate this on a cnc router.
Go back in the thread and @Hirudin claims that Shaper pulled some kind of hat trick with kirf, and while these guys look to be super sharp they didn’t do any magic there. The small intangibles are left to the end user.
I employed a communication technique called sarcasm to convey one meaning while saying the opposite. I thought my overuse of superlatives would have been a clear indication of the intended meaning, but I neglected to consider that it’s still technically possible for some to miss sarcasm, even when it’s laid on very thick.
Yea no he didn’t, he just pointed out that they have kerf compensation in the machine and glowforge dont have it even though they are using an entire server to process it.
Precisely. I usually end in /s to indicate ending a statement of sarcasm. That said, I will forever think that his claim of sarcasm is post hoc as instead of coming back with “dude, I was being sarcastic” he came back like he’s… well I’ll stop right there. But yeah, even there, I’ll be a bit agnostic because, well, Poe.
@Hirudin, I haven’t been reading all of the forums lately. So not sure where the hostility between you and other long time posters is originating from, you do seem to be fighting on multiple fronts though, so I apologize for joining in against you.
But on this one, you are coming off as refusing to think things through, primarily because you are responding with points already countered.
You cannot just divide the kerf by two in the software. Because the software has no idea what the kerf is. And because the kerf will change within a single job, since it depends on speed and power (and not in a linear manner) and since the material may not be consistent across the entire workflow. Kerf also changes with focus height, which not many people change within a single job, but some do. And again, this is not in a simple formulaic manner.
IF the machine were granted a “test area” in which it could go make a cut, AND the camera were dialed in well enough to measure the resulting kerf THEN it could automatically compensate IF the material being used were completely comparable where you want to compensate and where you test the kerf. Now… materials MAY be sufficiently consistent that doing such a compensation would be better than doing none at all. But it would still require a dedicated testing area with enough room for every speed/power/focus combination, and well calibrated cameras to measure the result.
Once kerf is known, or approximated, then it is indeed just as easy to compensate here as it is with shaper. So, attacking Glowforge’s choice of not allowing us to approximate kerf and compensate for that approximation is a valid stance to have. But making it clear that you would be happy to have the ability to compensate for an approximated kerf helps them know you want a low hanging fruit, instead of a potentially impossible perfection.
Why is everyone being deliberately obtuse? He didn’t say he wanted super duper automatic kerf settings. Just kerf settings period. Like, how insanely useful would it be to be able to select a line and tell it inside cut, kerf 0.xxx. Then the piece that you want to snap into it you tell it outside cut, kerf 0.xxx. All it takes is a simple offset command that multitudes of programs have for glowforge to copy. Entering a “kerf” setting is SOP in cnc machining when shooting for tight tolerances, it is just called a tool wear setting there and it adjusts the automatic “kerfs”. In fact you do it right on the machine; many of which still take their programs from floppy drives.
Not having it makes inlay pieces a much bigger pain to program into the machine and goes against glowforg’s ease of use mantra.
I think the kicker is that it looks to me like Glowforge is trying to let people be successful making things without learning things like “what does kerf mean”? And in particular avoid making the UI intimidating by including things that people don’t dare to start.
If it were me, an engineer with some CNC experience, I’d prefer the insanely useful manual setting first and then automate later. However, I’d be willing to bet there are software designers out there (perhaps at Glowforge even) that will talk long and hard about conceptual load of the UI and that being a barrier to utility and sales. I find those conversations very frustrating though when the automatic option is still about 1 million hours of testing away from being so awesome as to obviate the need for the manual one.
But then again, maybe the manual kerf adjustment is just in the hopper and somebody will just put it in there.
I think though, that the essential tension in this discussion is one that will dog Glowforge forever. It looks like the market they are after really wants to make things without a significant education (in other than safety and perhaps not even that) and that pushes against people who already know how to do what they want.
The whole camera alignment thing is another example of this. It’ll only really replace the control experienced people want if it gets a whole lot more precise (and maybe not even then because it’s still pixel bound after all), but it’ll be good enough for all the neophytes (perhaps not yet, but soon?) and a lot easier to explain. Until the automatic option actually works though, it’s a lot more frustrating than the manual one that while hard to learn actually works (and in more situations).
It would not at all surprise me if glowforge held back features that work perfectly with minor amounts of user input that make their product much more useful because a 10 year old can’t understand what it does automatically.
I’m indeed asking for low hanging fruit, not the impossible. Thanks for acknowledging that as a possible reading of what I’ve been advocating.
The notion that you think a possible interpretation of what I’ve said is I want the Glowforge to calculate the kerf with micron accuracy, I want it to do it on the fly, and I want it to do it without any user interaction is astounding to me. If you were willing to find a quote from one of my many, many posts on this subject that suggests the kerf compensation should measure all covceivable variables, I would be grateful. It was not my intention to use language that could be misunderstood to such an extreme degree.
Is it because I’ve used the word “automatic”? Is that why you seem to think I’m expecting Glowforge to invent the singularity that possesses all atomic information which it would then use to calculate variances in kerf width in order to achieve perfect compensation? That’s not what I mean by “automatic”, what I mean is “the CAM software can offset lines, and then does offset lines”.
