[pp-TG]

What could be wrong if not getting true arc output? (g18,g19)

V25

[mactec54]

pp-TG

Were you wanting to work in G18= ZX Plane or G19=YZ Plane or G17=XY Plane this code does not define the arc, only the work plane that it would be used in

[mmoe]

First, I'm assuming you are talking about a 3d model since you are asking about G18/G19, and really there is rarely an instance where a 2d arc in a drawing would not generate an arc motion in the toolpath as well. There are a number of reasons why you would not get arcs. First, the toolpath strategy you are using needs to support "arc-fit". If that option is not available, the strategy is not capable of outputting arcs in G18/G19 (in some cases because the strategy is working in the G17 plane anyways). If it is available, you need to check the "arc-fit" box or it will output lines instead. If you are using a strategy such as Planar Slice, which does have the "arc-fit" option, you can't set the lace angle to anything other than 0 or 90 degrees (or 180 and 270, same thing). Arcs can only be produced if the motion is along the XY, XZ or YZ planes. If you have the lace angle set at 45 degrees, for example, you would not be running along any of those planes and arcs would not be generated.

V26 added arc-fit to strategies that didn't have it before, but IMHO the result was not really anything worth while. A strategy such as Equidistant Offset does not benefit much from arc-fit because it moves along in infinitely different planes and very rarely lands on the XY, XZ or YZ planes with 3d models. For those cases where it would, you usually would be using something else like Flatlands or Advanced Roughing.

[pp-TG]

Thanks mmoe,

The arc-fit mode is the problem then. We were using the Equidistant offset and I could not get it to post in G18/G19 for the life of me. The program was to big for the machine, G18/G19 mode really cuts down size, it needed to run in and we don't have that machine hooked to a drip feed. We ended up waiting for the computer with Gibbs on it and reprogrammed it and are good to go now.

Again thanks for your time.

[Forrest C]

V25 does not support arcs in equidistant toolpath and neither does BobCam V3. Its comforting that V26 does because I'm hoping the next release of BobCam will also.

[mmoe]

V25 does not support arcs in equidistant toolpath and neither does BobCam V3. Its comforting that V26 does because I'm hoping the next release of BobCam will also.

Just so you can't say no one warned you, Equidistant Offset is not a good strategy for using arc-fit anyways, so you'll likely see very little or no difference between V26 and V25 in that regard. I think it's an excellent strategy, so not a knock on Equidistant Offset (which is one of my go-to strategies). It will produce an excellent result without the arcs, IMHO. The issue, as I pointed out earlier, is that equidistant offset rarely is generating a toolpath that aligns with G17, G18, or G19 work planes, so it is not possible to fit an arc when moving off those planes. For most 3d parts, Equidistant Offset will produce mostly lines even with arc-fit enabled. The only exceptions I could find are when the shapes are forming a somewhat square shape overall. I did quite a lot of experimenting with this when V26 first came out, including the following test (blue lines are linear motion while pink lines are arc motion, ignore the odd full circles as they are a result of the backplotter producing an error, not the toolpath generated by Bobcad):



The above shows a couple of things and the part was designed as the best case scenario for an arc fit in Equidistant Offset. It's square, and the arcs are running along the XZ and YZ planes, so if you run the toolpath from the outside in, it should have the highest likelihood of producing XZ and YZ arcs. Back it up a little and look at the top two results. First, the main thing to remember is that you must be working along either the XZ or YZ planes in order to get an arc motion in the Z direction (vs. around the XY plane). The first image shows that if you set the lace angle to 0 degrees and run along the XZ plane, the slice planar will produce mostly arc motions (pink) on that plane. If you take the exact same setting, and change the lace angle to 30 degrees so that it is not running on the XZ or YZ plane, you get nothing but lines (blue). Equidistant Offset produce only a small portion of the job in arcs, while the vast majority remained in lines. Again, this part was designed with this toolpath in mind, and it has the best chance of any part of producing arc motions. Yet it produces very few.

Additionally, with Arc Fit enabled, the quality of the line portions of the toolpath actually dropped compared to the same Equidistant Offset strategy with the same setting other than Arc Fit. You can overcome that by increasing the tolerances to produce a finer toolpath, but you add a lot of processing time just to get back to the equal of the same toolpath without Arc Fit. The only times I could get decent arc fit results from E.O. was when it was a part that would be better done in something else like spiral.

All that said, the rest of V26 is more than a cosmetic update to Bobcad and I think the DMS system is really growing on me (and others who have given it a fair shake). It makes the whole process a bit faster, and makes edits a lot easier as well. In my opinion, that's the main reason to update to V26. I'm sure you'll find that Equidistant Offset is not a lot different with arc fit than it was without it, but I would also remind you that any toolpath strategy by any other vendor will also not produce arc moves that are not on the XY, XZ or YZ planes, so if you are moving in a meandering toolpath offset by equal amounts throughout (Equidistant Offset in a nutshell), you'll rarely find the motion to align with those planes.