For our final project we had to make two chess pieces, one white and one black.
We had to use the CNC to carve out a positive wood piece, then make a negative mold of it, then cast the two positive final pieces.
After choosing the queen as our piece, the first step was selecting a design. We could have chosen the standard Staunton style, but since my partner and I enjoy designing and making things, we decided to have some fun with it and make it more interesting. We opted to produce a miniature model of a bust of Queen Nefertiti instead.
We found a mesh .stl file online and downloaded it, then opened it in Meshmixer. There we made it a solid, added a base for the piece, and split it in half. We had to do this last part because the CNC only works in 3 axes, so we had to make two halves and carve them individually.
I know it is possible to do most of the design work in Fusion 360, but neither me nor my partner were familiar with it, so we did some of the work in other programs such as TinkerCAD. After adding a base plate and a few indexing circle locations in Fusion, we opened it in TinkerCAD and mirrored the image to create two symmetrical pieces (We tried for hours but were unable to do this in Fusion). We then extruded the circles up or down to make the indexing keys or holes, saved the file and exported it to Fusion.
Once we finally had the final design in Fusion, we struggled for a while to get the axes correct, the setups correct, the G-code correct, etc.
For people doing this in the future:
Step 1: To open a .stl mesh file in Fusion 360, you have to go to the Insert menu and select Insert Mesh.
Step 2: You have to tell Fusion to not capture the Design History
Step 3: Then you have to convert the Mesh to BRep.
Step 4: We also had to download the Easel post processor for Fusion 360. To do this, go to Easel, click File, Import g-code, the select Autodesk Fusion 360 from the list. People who did not do this had problems getting the CNC machine to work with their g-code. Just click the link to download then follow the instructions to install.
Step 5: Go to the CAM page in Fusion and make a new SETUP. Our axes were wrong (Z should be up) so instead of using “Model Orientation” we chose to “Select Z axis/plane and X axis.” To select the Z plane, click on the horizontal plane.
After getting the axes correct, for the Origin we selected Stock Box Point and chose the top corner such that the X and Y arrows were pointed toward the piece.
Under the Stock tab, we chose a “Fixed size box” and input the dimensions of our piece of wood, but other teams used “Relative size box” and it worked for them.
Step 6: To start telling the machine what to do, go to 3D. We chose Adaptive Clearing.
First select what kind of cutter tool you will be using. Then disable the coolant. Then input the appropriate spindle speed.
In the Geometry tab we deselected everything, because one of the many tutorials we watched or read said to.
We left everything in the Heights tab as it was, but in the Passes tab we deselected “Stock to leave”. If you select “Smoothing” then that might help quite a bit. Also if you reduce Fine Step-down then the part will take longer but it will be smoother, which will reduce post processing (sanding or filing or using a Dremel). We did not change anything in Linking.
Step 7: After clicking OK, we waited for it to generate the G-code, then under the Actions tab we Simulated the cut to make sure it was doing what we wanted. The first time we did this our axes were wrong and we would have had a huge problem, so it is a good idea to Simulate the cut. You can also use third party programs, such as this one, to do this.
Step 8: Once everything looked good we again clicked on the Actions tab, and selected Post Process. We then chose the Easel post processor that we had installed earlier and clicked OK. We saved our new G-code (.nc) file, then opened it in a text editor to remove a few lines that our CNC machine was not happy about.
Step 9: After saving our file to a flash drive, we loaded it into the Machine shop computer and again checked that everything looked right. While one of us was opening and loading the file, the other was setting the origin for the cutter head at the appropriate spot. As you can see below, we should have offset the origin a few centimeters, but we did not figure that out until it was too late. We recommend to use as big a piece of wood as possible so that you do not have to worry about it as much. You can always use the bandsaw to remove excess afterward. We had a piece of 2×6 to use, but it would not fit into the small clamp.
As the machine was cutting, it did not cut to the outside of the material, so it left some wood that I had to remove manually so that it did not interfere with the machine. To do this I pressed the Feed Hold, then turned off the cutter and jogged the machine away from the piece so I could have room to work, then used some wire cutters to cut away the strips of wood left around the edges of the piece. After turning the cutter back on we just told the machine to start back where it was when I paused it.
As you can see from the photos above, the top of our queen’s headpiece extended past the end of the stock, so it did not get cut. Some of our indexing pins and holes also were off the material, which made aligning the mold difficult, as you will see soon.
Step 10: Post processing the wooden positive.
To fix the queen’s headpiece I used hot glue. Clay would probably have been a better option but I was using the hot glue gun already anyway. I put a layer of glue on top then let it harden, then added another layer and so on until it had built up enough material. Then I used sandpaper and files to shape it. I also used them along with a chisel to remove some of the cut marks and smooth out the piece. A Dremel would have been faster and easier, but as a woodworker I enjoyed the opportunity to do some detail work with hand tools.
Step 11: Making the mold.
In the past I have been interested in making cinderblock buildings, and I have experimented with different ways of laying the blocks to make them as strong as possible, so this part was fun for me. Apparently I thought I was making a pressure chamber, because I way overdid the Lego box. After making sure that each Lego piece was offset from the ones underneath it such that absolutely none of the seams aligned, and wrapping them in duct tape, and putting a line of hot glue around the inside along the cardboard bottom, and also along the outside, I then used a few dots of hot glue to secure the wood piece to the cardboard. Some of the other teams had problems with their wood pieces floating up in the mold material, so we made sure to secure ours down.
We mixed the mold material according to the directions on the buckets and then poured it into the Lego boxes.
Here you can see where the mold material seeped under the block of wood. Securing it was a good idea.
Step 12: Preparing the mold.
18 hours later we returned and deconstructed my miniature Fort Knox. We did not put any release agent on the wood before pouring the mold, but it turned out fine.
Because there were gaps between the wood and the Legos, we had some excess material that we had to remove to make the two halves fit together well. This was done easily with a box cutter, which we also used to cut a pour hole into one side of the mold.
Because of the complications with origins and axes and dimensions etc, and because our indexing pins and holes were not cut properly, when the two halves were aligned properly they looked like this:
To keep them aligned we used tooth picks, and to keep them together we used rubber bands.
Step 13: Casting the pieces.
We mixed the SmoothOn casting material according to the specifications on the containers. To make the black piece we added two drops of brown dye and one of violet. The bust of Queen Nefertiti is over three thousand years old, so we figured that the brown (for age) and the violet (purple for royalty) would be appropriate. Plus they were out of black dye…
We mixed the ingredients together thoroughly and quickly and then carefully poured them down the pour spout into the mold. 30 minutes later they were ready to be released from their tomb.
Step 14: Post processing.
As aforementioned, things did not really go to plan when CNC cutting the pieces. And although my hot glue shaping skills are pretty great, I still have some room for improvement. These imperfections made aligning the two halves of the mold difficult, and consequently the final castings did not come out circular or symmetrical. The pictures below show how the two sides did not match up perfectly.
To fix this and clean up the rest of the pieces, as well as remove the material left in the pour spout, we used box cutters, sand paper, files, a Dremel, and a lot of elbow grease.
Step 15: Finishing.
Eventually we got the pieces looking acceptable, so then all we did to finish them was douse them with a healthy coat of clear gloss sealant.
We were going for that “3,000 year old look,” and I think we achieved it. Queen Nef has been through a lot, she certainly looks her age, and has come half way around the world to visit us here at Rice. The clear gloss does give her skin a nice revitalized shine though. She looks ready for years reigning over knights and bishops and pawns.
Conclusion:
Our chess pieces did not turn out perfectly, but I think that for two undergrads (one an Econ major) using equipment and software and materials and techniques with which we were previously unfamiliar, things turned out pretty well. It is hard to tell from the photos, but if you look closely at the real thing you can see the small details like facial features. I think it is really cool that those could be preserved through all the steps we went through, from design reduction on the computer, to the CNC carving of the wood, to the mold making, then the casting. If I had to do it all over again there are some things I would do differently, but the fact that I can say that means that now I know the proper way to do them, which is an indication of the fact that I have learned a ton through this project and this course in general.
Now who wants to play me in chess?
The End.
Post Scripturum:
This was way harder than I expected. At times it was very frustrating, but then I remembered that I love this stuff and if I was not doing this assignment I would probably be using the time for designing and building something else anyway. Hopefully if students in the future do this project, this blog post will be helpful to them. I am grateful for the opportunity to push myself and learn new things and develop new skills. I have always loved designing and making things, and this class has made that even more true.
