Is Path workbench up to the task

[routalot]

From being a complete novice to actually modelling something so complex in just a few months is impressive.I find it frustrating that the things I would like to do can usually be done by Freecad but it often takes quite a lot of time to find the right workbench and then understand the way to go about it.Since I only got my machine running in November,I didn't spend too much time learning the ways of Freecad until then and I find the basic machining aspects quite easy to use.The sliptonic videos on youtube have been a huge help and I really hope that before too long one will feature 3D machining.I am getting 3D parts machined and it is a huge thrill to see them emerge from a block,I just have a strong feeling that there may be ways of doing the job more efficiently.

I went back to your uploaded file as I have a similar project planned and you have given me a very useful training project.I found that the huge number of surfaces and nodes was really slowing things down so I edited and simplified it a bit.What I did find is that the faceting is particularly severe at the tips and by using the texture mapping feature to apply the mirror surface I have as a saved image file this image shows what it revealed.I haven't enough experience with surfacing in Freecad to know whether it might be possible to adjust the U and V values to smooth the surface or if you are limited to the surface you have imported from you wing generation program.

[joaopcrodrigues]

I'm giving FreeCad one last chance....
Do you think PathWB could work tihs, by first milling the upper part, than in a second step with a clone of the piece up side down????
The .fcstd file is quite large 70mb, but I can upload it somewhere if needed.

Screenshot from 2019-03-27 16-29-51.png (266.14 KiB) Viewed 1791 times

Screenshot from 2019-03-27 16-29-14.png (249.51 KiB) Viewed 1791 times

[chrisb]

It's hard to see what this black stuff is, stairs? Those can be milled if you can use a mill with a diameter less than the width of the a single stair step.

[routalot]

Your latest model would probably defeat every system because it is impossible to get into all those sharp corners.If you are intending to machine an actual component,rather than a male pattern for moulding, the additional surfaces ought to be below the split line of the piece by an amount equal to the radius of the tool you are using as this will allow the tool to roll over the leading edge and trailing edge.If this is done I believe the outcome would be successful.As I said in an earlier post this topic is of interest to me as I have a vague intention to produce an improved tip foil for my rudder.I will try to find the time to have a try at something to verify my claim that it can be done and will report back when I have done something.

Are you familiar with the concept of draft on moulded components?It may be a concept worth visiting regardless of system.

[routalot]

I had time to experiment with the original model and to save some time I only worked with half of it.I projected lines through the mesh in a few places and created some surfaces that are a good approximation,but no more than that,just to establish the principles.At both the tip and the centre I added some extra material to provide a datum surface for the second operation and around the edges I added a helper surface which is below the modelled surface by an amount equal to the radius of a ball end cutter in my tool library.This allows the actual surface to be machined and there is no need for the cutter to drop down to the lowest point of the piece.This surface doesn't have to be there on the real thing-it is just to generate a toolpath.For some reason I don't fully understand one of the surfaces has been missed by the tool and I may have a normal reversed or a gap at the edges.I also used a large stepover as it illustrates the possibility without taking a huge time to create a toolpath.

It is important to be aware of the potential risk of the collet nut rubbing on the stock with some jobs,but I think the depth of this item shouldn't be a big problem.I hope the attached screenshot is useful.

[joaopcrodrigues]

I had time to experiment with the original model and to save some time I only worked with half of it.I projected lines through the mesh in a few places and created some surfaces that are a good approximation,but no more than that,just to establish the principles.At both the tip and the centre I added some extra material to provide a datum surface for the second operation and around the edges I added a helper surface which is below the modelled surface by an amount equal to the radius of a ball end cutter in my tool library.This allows the actual surface to be machined and there is no need for the cutter to drop down to the lowest point of the piece.This surface doesn't have to be there on the real thing-it is just to generate a toolpath.For some reason I don't fully understand one of the surfaces has been missed by the tool and I may have a normal reversed or a gap at the edges.I also used a large stepover as it illustrates the possibility without taking a huge time to create a toolpath.

It is important to be aware of the potential risk of the collet nut rubbing on the stock with some jobs,but I think the depth of this item shouldn't be a big problem.I hope the attached screenshot is useful.

It's more or less the same approach... but no good enough for me.
Thanks anyhow, not waiting any longer, as off tomorrow I'll be starting learning Fusion360, I hope it will be able to do the job.
Best regards.

Screenshot from 2019-03-28 23-28-38.png (200.72 KiB) Viewed 1710 times

[chrisb]

Thanks anyhow, not waiting any longer, as off tomorrow I'll be starting learning Fusion360, I hope it will be able to do the job.

It would be interesting to hear a short statement after you have some experience with Fusion360.

[joaopcrodrigues]

Thanks anyhow, not waiting any longer, as off tomorrow I'll be starting learning Fusion360, I hope it will be able to do the job.

It would be interesting to hear a short statement after you have some experience with Fusion360.

I hope I din't offend anyone, if I did, I'm sorry, it wasn't my intention.
But ok, I'll say something, just give me time, I have hours and hours of tutorials and learning.
Best regards.

[chrisb]

Don't worry, no offence detected here. I'm just curious.

[JulianTodd]

As the guy who wrote most of the CAM algorithms in Fusion (when it was done by a company called HSMWorks), and who has gone deeply into the path workbench code, like here https://forum.freecadweb.org/viewtopic. ... 8&p=280958, I've encountered some fundamental issues with the path workbench in that some things which should be simple have become super-complex both in implementation and the UI. This means that the more advanced 3-axis operations, like rest milling, pencil milling, constant stepover, and tool holder collision avoidance, are never going to be possible, because they require the kind of effort that is not humanly practical.

Toolpaths and geometry are quite simple things to advanced CAM algorithms, like you find in Fusion360; they're just sequences of line segments and meshes of triangles respectively. But it's nearly impossible to abstract down to this in the Path workbench code, as I've found out.