Slicer with built in CAM?

New Home Forum Software / Firmware Development Slicer with built in CAM?

This topic contains 8 replies, has 4 voices, and was last updated by  Aaryn 1 week ago.

Viewing 9 posts - 1 through 9 (of 9 total)
  • Author
    Posts
  • #92564

    Aaryn
    Participant

    I have seen videos of large commercial machines that will 3D print a layer then use a mill tool to smooth out the 3D printed layer walls.  Let’s assume it is possible to attach a Prind head and a router (or hot pen) to a machine at the same time.  Does anyone know of any slicer software that could generate the Gcode to use both tools on a single job?  Are there any similar projects in any pipeline?  Maybe a kickstarter?

    Because I think desktop 3D printing could jump to the next level of smooth, perfect to spec tolerance parts if each layer could be milled (or hot pen molded) right after it was printed.  It would eliminate most over extrusion ugly print jobs and parts that don’t fit due to over extrusion.

    #92682

    Ryan
    Keymaster

    I went to a show and they had two arms working at the same time. One was extruding 3/4″ (no joke or bigger) filament, and about 6″ below it the other one was 5axis milling it. It was printing half of a car full sized. Talk about mesmerizing!

     

    I don’t think it is as hard as it seems for a single head (dual head has to worry about collisions between them). For a single head you print the part, then use that as the base you build your cam on. Some fancy CAM packages let you define more than just a square blank.

    #92700

    Guffy
    Participant

    Honestly i have no idea how combining of fdm and milling (or fdm and laser) can be handy in desktop 3d printer. Imho, much actual is to use multi nozzle printing where the printer and slicer could use different size nozzles for outer perimeters and for inner and infill. Only a case  that i guess could be useful is when you need a lot if holes and it’s more faster to drill them then printing a lot of circles.

    Anyway indeed it’s easy can be done with two (or more) identical scara robots surrounded the work area.

    #92808

    Aaryn
    Participant

    I went to a show and they had two arms working at the same time. One was extruding 3/4″ (no joke or bigger) filament, and about 6″ below it the other one was 5axis milling it. It was printing half of a car full sized. Talk about mesmerizing!

    Oh!  That would be beautiful to see.  I can’t find it anymore but I saw a video a few months back about a laser metal printer.  5 Axis.  This thing was laying iron or steel beads about 1/4″ thick then it would switch tools and use a mill to give the part a perfect finish.  It was amazing to watch.

    Then later I saw in one of the threads here the video of an airplane part getting 3D printed then finished with a mill.

    For the last few months I have been noodling the idea of the same concept working on a desktop 3D printer.

    Lets take the 3D printable parts of the MPCNC for example.  Right now you (Ryan) have extra tolerances built into the design.  The feet for example are slightly larger than the conduit usually measures.  I understand this extra space is to compensate for the inaccuracy of a 3D printer.   Because when the common (FDM) 3D printer lays down a bead of filament it tends to ooze out the sides a little causing the bumpy layers.  Those bumps cause the parts to be a little tighter than the 3D model to they actually fit the conduit. In other words you had to compensate for the average tolerance problems with most 3D printers with the design.  And you did a great job from what I have experienced.

    But if desktop 3D printers could produce parts with tighter tolerances this kind of extra engineering wouldn’t be necessary.  So I was thinking about attaching either a low power router (or Hot knife) to a 3D printer (FDM).  The idea would be that it would lay down a layer of filament let’s say .2 mm thick.  Then it would switch to a cutting tool or hot pen like sculpting tool.  I imagine the tool would have a ball nose but a flat end might work better.  It could then reshape all the outer perimeters.  Maybe as low as 0.21 thick.  That way it would smooth out all the bumps and ridges.  In theory the parts would be perfect to spec.  You could easily print two parts that could fit together like puzzle pieces and they would fit without any sanding or post processing.

    I would love to build the machine myself but I don’t know of any software that could drive it.

    #92857

    Ryan
    Keymaster

    I see what you are getting at. Well currently you can, with the MPCNC, pretty easily do just that. Print a part, locate it (I would use a printed jig or a milled footprint), finish mill it. I see you are looking for an all in one software package but since that is not my specialty I recommend my actual specialty, making the best of what is easily available (conduit, skateboard bearings…)!

    #93154

    Jeffeb3
    Participant

    Ryan, always with the hardware solution 🙂

    You could do it in a post processor. Something that would read in the gcode, and collect extrusion moves into a layer, then on a layer change, it would know everywhere it had extrusions. Except I’m not sure how it would know what was empty space in infill and what was an actual space that needed to be milled. A post processor that also had the original STL would have the info it needed. It would probably be easier to do inside of slic3r, but who knows how to write perl? For a proof of concept, something that operated on the stl and only milled the perimeter and the top would be doable.

    I wonder if it’s worth the effort though. You’d end up with a lot of extra time, and a lot more to the cost and complexity of the machine. You’d also need to deal with the fact you’ve got a lot of mess to clean up in a printer, which I would consider a generally clean operation. Are you sure you could get a higher tolerance with a mill than you could with a well tuned printer? Jobs would also take a lot longer, especially if you milled the whole surface. A smaller nozzle and a higher gear ratio extruder might work a lot better, and then there’s the cost of SLA coming down (with it’s own problems). I suppose a much larger nozzle and a mill could recoup some of the time lost. The welding printer is a different story. That would be a good solution (although I think there are other challenges).

    #93170

    Aaryn
    Participant

    my actual specialty, making the best of what is easily available (conduit, skateboard bearings…)!

    I love this Forum!  I have been so focused on the idea of milling each layer I haven’t thought much about trying to use the MPCNC afterward to “Finish” or post process a 3D print.  Interesting idea.  I might actually give that a shot here soon just to see how a router bit will handle PLA and ABS.  If that doesn’t work then this whole idea is dead anyway.  But I could 3D print simple 3D part with a jig base.  Move it to the MPCNC and use the 3D carving in ESTLCAM to see if the finishing will work. I could also test the effect of using a hot pen instead of a router.

    You could do it in a post processor.

    Hadn’t thought of that either.  If the first test works I’ll look into it.

    I wonder if it’s worth the effort though

    You bring up some very good points.  Time I am not too worried about though.  I have learned to be patient in order to get better results.  If I could eliminate 90% of the post processing I have to do then it would be worth the time to wait for the machine.  I mean. Can you imagine a 3D printer that can be perfect to size specs?  That is why I bought my SLA printer.  It is really close but even the Form 2 is exact to spec. It is usually about 0.05 mm bigger than the 3D model.  So again that tolerance needs to be engineered into the models.  As you mentioned SLA has other issues.  Brittle, and cost being the largest ones I have to deal with.

    As for the mess.  Not to worried about it personally.  IF this idea ever went mainstream I imagine it will be used on machines that are more like a CNC first and a 3D printer second. By that I mean the belts, bearings, and rods will be located out of the way for the most part like the PMCNC.

    In any case.  Thank you all for the ideas and feedback.  I probably will not be able to take this idea very far on my own.  But I am going to make the prediction that this will be premium feature on Desktop 3D printer/CNC routers of the future.

    #93171

    Guffy
    Participant

    you, guys, probably forgot about shrinkage of plastic models when they have becoming colder. so if you want to just to make little milling to make outer surfaces slick it will be not easy to find proper coefficients to make dimensions correction. and not all shapes of a model could be acceptable for post-milling.
    milling during process of 3d printing will produce just a messing of extruding plastic and chips.

    ps. afaik, slic3r uses perl only for UI. core uses c++

    • This reply was modified 1 week ago by  Guffy.
    #93174

    Aaryn
    Participant

    probably forgot about shrinkage of plastic models

    That is also true.  There in lies probably the biggest reason why the entire idea could fail.  Hmmmm.  That would cause ABS and other materials that expand a lot to fail to meet spec more often.  PLA doesn’t expand and contract as much but I would guess that it would still end up with a similar tolerance issue as SLA then.  Better than no milling or sculpting but in that case it may not be worth the extra cost of the machine.

Viewing 9 posts - 1 through 9 (of 9 total)

You must be logged in to reply to this topic.