CNC hexapod cuts its first material!
CNC hexapod cuts its first material!
by winchy_matt
Fri Jun 06, 2008 9:14 am
Hi,
I haven't posted anything about my CNC hexapod project here before, but if you wish to see the origins of this project you can view them on my forum here: CNC hexapod
Ok, so its only polystyrene, but its a start
A couple of improvements were made to the router since my last post, most significantly I had a colleague make me up an dremel colet adaptor to fit directly onto the motor shaft, much better than my brass grub screw style colet.
For these two videos I'm using a 1/16" two flute slot cutter, the router is running at approx 35,000 Rpm, the additional battery on the to of the hex is a 12V NiMh 4500mah pack just to run the router. DXF files are loaded in to artcam where G-code tool paths are created. Due to the 10mm thickness of the poly and the cutter only being 6mm long, I chose to do two passes on the workpiece, this also shows that although the resolution is low, repeatability is good enough for multiple passes. Material is held down by the weight of the hexapod, which is current;y 4.3Kg with the router and additional battery pack.
CNC Cut Poly Square
CNC Cut Poly Femur - 2 pass cut
You may notice in the picture of the poly femur with the hole left in the poly, the hole has a cleaner cut edge than the cut piece. This is due to the fact that foam and poly cuts better in a reverse direction, however, I forgot and generated the tool paths for a normal cut direction.
My second attempt at the femur shape, I increased the size by 20% and reverse cut the piece.. much better!
Original and larger New part
I haven't posted anything about my CNC hexapod project here before, but if you wish to see the origins of this project you can view them on my forum here: CNC hexapod
Ok, so its only polystyrene, but its a start
A couple of improvements were made to the router since my last post, most significantly I had a colleague make me up an dremel colet adaptor to fit directly onto the motor shaft, much better than my brass grub screw style colet.
For these two videos I'm using a 1/16" two flute slot cutter, the router is running at approx 35,000 Rpm, the additional battery on the to of the hex is a 12V NiMh 4500mah pack just to run the router. DXF files are loaded in to artcam where G-code tool paths are created. Due to the 10mm thickness of the poly and the cutter only being 6mm long, I chose to do two passes on the workpiece, this also shows that although the resolution is low, repeatability is good enough for multiple passes. Material is held down by the weight of the hexapod, which is current;y 4.3Kg with the router and additional battery pack.
CNC Cut Poly Square
CNC Cut Poly Femur - 2 pass cut
You may notice in the picture of the poly femur with the hole left in the poly, the hole has a cleaner cut edge than the cut piece. This is due to the fact that foam and poly cuts better in a reverse direction, however, I forgot and generated the tool paths for a normal cut direction.
My second attempt at the femur shape, I increased the size by 20% and reverse cut the piece.. much better!
Original and larger New part
by winchy_matt
Fri Jun 06, 2008 12:46 pm
I had a go at 3D profiling today in high density foam. I have a few issues to sort out, mainly with my control software not having a puase or for that matter a stop button 
.. any who, I fed a 3D stl file into artcam and ran a couple of passes through the hexapod to get this:
I admit its not great, but you can kind of make out the eyes, nose and mouth.. kind of. The large 1/8" scar on the face was caused by operator error!
I will post a video of this process some time after the weekend, and try another test next week.
.. any who, I fed a 3D stl file into artcam and ran a couple of passes through the hexapod to get this:
I admit its not great, but you can kind of make out the eyes, nose and mouth.. kind of. The large 1/8" scar on the face was caused by operator error!
I will post a video of this process some time after the weekend, and try another test next week.
by BillB
Fri Jun 06, 2008 7:04 pm
Well done Matt. Very inspiring work.
Once again you have challenged the boundaries of what we think humanoid/hexapod robots can achieve with the current servo technology, and smashed through those boundaries.
It demonstrates what fantastic level of control and accuracy you have over your robot. I have trouble controlling the 2 legs on my humanoid, and am always amazed at how well you orchestrate all six legs on your robot.
You Rock
Once again you have challenged the boundaries of what we think humanoid/hexapod robots can achieve with the current servo technology, and smashed through those boundaries.
It demonstrates what fantastic level of control and accuracy you have over your robot. I have trouble controlling the 2 legs on my humanoid, and am always amazed at how well you orchestrate all six legs on your robot.
You Rock
by winchy_matt
Mon Jun 09, 2008 8:09 pm
Ok, attempt two at 3D cutting was much better.
I made a few improvements to the software that streams the data to the hexapod.. including a stop button! The cutter was the same, 1/8 ball nose. After a roughing cut, I did a raster pass with 1mm step, and a second raster pass in the opposite direction with 0.8mm step.
I could use a smaller ball nose cutter to get a finer finish, but I think I'm approaching the limitations of the robots resolution.
I will edit the video and upload when I have time.
I made a few improvements to the software that streams the data to the hexapod.. including a stop button! The cutter was the same, 1/8 ball nose. After a roughing cut, I did a raster pass with 1mm step, and a second raster pass in the opposite direction with 0.8mm step.
I could use a smaller ball nose cutter to get a finer finish, but I think I'm approaching the limitations of the robots resolution.
I will edit the video and upload when I have time.
by winchy_matt
Wed Jun 11, 2008 12:22 pm
I have just uploaded the 3D face video.
Cutting 3D face
To cut the face I did three passes:
1) roughing cut
2) X raster finish
3) Y raster finish.
The face took just under 30 minutes to produce.
Cutting 3D face
To cut the face I did three passes:
1) roughing cut
2) X raster finish
3) Y raster finish.
The face took just under 30 minutes to produce.
by StuartL
Wed Jun 11, 2008 4:55 pm
Are you finding that the robot is heavy enough to keep itself from drifting during the cut? It's an amazing thing to watch.
Next step is to get it fitted with a vacuum attachment so it can clear out its own CNC dust and as an added bonus can clear up the floor of all those dusty robot foot prints...
Next step is to get it fitted with a vacuum attachment so it can clear out its own CNC dust and as an added bonus can clear up the floor of all those dusty robot foot prints...
by winchy_matt
Wed Jun 11, 2008 10:56 pm
StuartL wrote:Are you finding that the robot is heavy enough to keep itself from drifting during the cut? It's an amazing thing to watch.
Next step is to get it fitted with a vacuum attachment so it can clear out its own CNC dust and as an added bonus can clear up the floor of all those dusty robot foot prints...
Hi Stuart,
Yes it seems plenty heavy enough for HD foam and probably some other harder materials. The feet do move a little, and so repeatability isn't great, but good enough for the purposes of the exercise.
I have thought of fitting a blower to clear the work surface, I also need a better heat sink for the spindle, as its currently getting a little hot after 30 mins use!
Matt.