Week 15

We are going to build the mechanical parts of a CNC machine this week. Below you can see a cardboard prototype as a milestone for this week. Further you can read about the process.

You can read the full story on the machine building group assignment page.

To be more speciffic, I did the design for the bottom part of the machine and 3D-printed bearing holders.

Individual Contribution

My contribution for the Mechanical Design week was to set up the group website and design the Y axis for CNC milling. I did the design part in Fusion 360 and it also included some parts that had to be 3D printed. I visited Joris at his lab in Perpignan where we did most of the CNC milling and 3D printing. You can read more about the process in the machine building group assignment page.

The Machine

For projects of mine I usually need to drill holes in aluminium profiles for connecting surfaces to them. It can be achieved with a CNC machine, but the limit is its bed size. What I would like to do is to have an infinite X axis.

To build the machine I teamed up with Joris Navarro and we quickly decided that Hattori (small format CNC mill and 3D printer) would be a good start. The following image is from the Hattori GitHub repository.

Hattori axes

Research

Find out a good-enough mechanism for moving the profile along the y axis. It should be able to push and pull the profile of an infinite length (theoretically)

Max profile height: 10cm Max profile width: 20cm Max profile lenght: 2m

The general idea is to take the Hattori Rhino project, use Grashopper to customize the machine, export it and use Fusion 360 to finish the rest of the mechanical parts.

For the wheels that move the profile forth and back we were considering to use rollerskate wheels.

We are going to use steel rails. Joris has a stock of 8mm ones. 40cm long.

On Friday morning we changed our minds as Joris found a design of another CNC machine that we can reuse. The only thing to be done is to flip the y axis as we want to use to clamp the profile to be drilled.

Process

The idea is to rotate the bottom part of the machine by 90deg in order to make it work as the infinite Y axis mechanism. Below you can see an photo of some of the first sketches of our mechanism.

One of the First Sketches

We got some parts together and decided to build a cardboard prototype first. That would let us see possible design challenges in a much more human way. For that we need cardboard and glue of course. One of the challenges here was to decide the size of the machine. If the previous day we thought 50x50cm would be the right size, today we went for 30x30cm as the right choice.

Glue

While Kris was preparing the documentation website and Andres doing research around gears, Joris was cutting cardboard and we ended up with having the following initial version of the frame.

Initial frame version

There were a number of challenges that we identified at this point. One of them was how to mount the motors. Another how to make it easier for the profile to be drilled to move through the machine.

Skech for rolling edges

One more issue is where to place the electronics. How to connect the spindle motor to the Z axis? What material to use to grab the aluminium profiles to be drilled?

We arrived to a point where designing custom parts out of cardboard seemed like the right solution. We designed rolling supports for the traveling profile to be drilled and connectors for the moving clamp wheels that would keep the profile in place, letting it roll at the same time.

Custom wooden lol
Cutting custom parts

Below you can see a video with a brief demonstration how the machine would work.

Cardboard machine intermediate demo

At this point we decided that the layout is OK and we could move forward with designing the actual frame and parts. While Kris and Joris would work with the part design and mechanics, Andres would start to test the electronics. The first step in terms of electronics would be to hook up the controller with an Arduino UNO, add the motor drivers and end-stops. Use the G28 GCode to try to home the electronics system, press the end-stop and arrive to a point when the motors stop when doing that.

On the mechanical side, one has to figure out where to put the end-stops.

7 May 2018. Kris is at the Fab Lab Poerpignan with Joris. Our task for today is to create CAD drawings of the missing parts and figure out the milling process.

Joris will prepare the Rhino design of the Jens Dyvik et al. Hattori machine. Kris will take care of designing the missing parts relative to the top part ot the Hattori machine in Fusion 360.

Joris and Kris working hard

It took some hours of work to get the Y axis modelled. Combining something existing made by someone else with custom design is hard. The naming of parameters and uderstanding of design software workflow varies. However there was progress and we got close to milling the parts here.

Initial CAD phase
Initial CAD phase

We switched from using 18mm plywood to 15mm and that is an interesting challenge always when you think that your design in Fusion 360 is truly parametric. Which it was. A bit more extra adjustments and we were able to start laying it out.

Initial CNC layout

Using the Fusion 360 Dogbone addin by Case Rogers was the new experience in terms of using Fusion. Below you can see an image that shows how the dogbones look like. It was very easy, the most important part is that you have to lay the parts out on a flat contruction plane before using the Dogbone addin.

Dogbones

And we were ready to mill. Below you can see an image of Joris checking the milling process. There were some fails since we were in a bit of a hurry. One of the fails was that we forgot to set the Z axis before using the 8mm tool. We noticed it shortly before it was about to drill into the frame of the machine. Another thing was that we forgot to add some of the inner pockets that were supposed to create holes in some of the parts.

Joris checking the milling process

All the parts with Joris

Meanwhile plastic bearing holders were being 3D printed.

Bearing holder printing

In the left image above you can see a custom cardboard support being placed since we forgot to configure supports in Cura software. In the image to the right you can see a bearing fitting into the 3D printed part.

Meanwhile 2x Andres was working on the electronics and could not manage to get the Arduino CNC SHIELD to work. He tried many things, our hope at this point was to find help from other groups using the same board.

Arduino CNC SHIELD failing

At this point we figured that another cycle got completed and we made a list with things that should be improved in the mechanical design part.

  • Add two more rolled connectors (the plastic parts with bearings) for extra support near the Z axis.
  • Add extra finger joints to connect the bottom plate of the Y axis.
  • Join Y axis and X/Z axis frames together.
  • Leave space all arount the clamp part.
  • Add linear bearings (LM8UU) to the clamp part so it moves smoothly.

The plan was also to put Y, X and Z axes in one Fusion 360 design file.

Notes

Stress and Strain. Stress is what happens when you press. Strain is pain or the force that the material holds when kept pressed down.

Gears are one of the most interesting parts to make since the issue is how to optimise the pressure applied from and to tooths of the gears.

Lead Screws are the screws that lead. Usually attached to a motor. There is a structure with a matching thread that moves when the screw is being turned.

Inspiration

Conclusions

It was a fun week, but at some points it felt like it would have been smarter to choose a less complex machine. Anyhow, I was lucky to have Joris in my team who had previous experience with building machines and he had many parts and solutions at hand.

I contributed with my Fusion 360 and web development skills mainly and learned a lot during the process. It was a good idea to build on top of something that is ready for a machine as complex as the one we built.