Molding and Casting
25 October 2017 | By Casey Evans
“What if I send in the tape and they don’t like it? I mean, what if they say I’m no good? What if they say, “Get outta here, kid. You got no future”? I mean, I just don’t think I can take that kind of rejection.” -Marty McFly
Julia Ebert. There are loads of tutorials online. Debouncing buttons is important. She accidentally made a proximity sensor. She is also interested in these lower level issues. Neil says it’s great though because you’ll be able to see a big improvement in performance without all the overhead. Except that we’re doing the overhead. Or rather, I’m failing to do the overhead. “It seems impossible until it’s done.” Tiny robots final project. They should climb over each other. Works with Ratica. Start and stop with flashing IR (photodiodes perhaps). Things that are a requirement for the class, not for the rest of your life. Raspberry Pi to boot an operating system. -Neil. “Make as much as you can to learn as much as you can.” -Neil. Unless you can’t make much, I guess. Again, I’m useless. Richard Liu. Pi 3 is a full Linux system apparently. iPhone cable final project. Joshua Coven. Jaclyn Berry. How does Neil know so much? I don’t think I know that much about any domain. I don't think I could know that much about any domain. Goes back to me being a waste of space, I guess. Interrupts are slower than loops, but with interrupts you can go do different things. I’m not sure if HtM is trivializing all the work I’ve done in other classes or if it’s just that most people have way more experience or what is up. I would have thought I’d be ahead in this area (electrical engineering stuff) but as per usual, I have no skills and should not even exist as an entity with independent action. Antisocial robot final project. Rocking robot from Fab Academy Beijing (multiple final project). Too often the world of engineers seems to be like “here is a toolbox, have fun.” Before you know it the hammer claw will be a wedge, the screwdriver handle will be a hammer—utter chaos. Blue LEDs are much less efficient. Bootloader is a program that loads programs. NRF 52 Agnes. Sam Kalish. Castillated board. Robert Tran did a 7 segment LED display. Charlieplexing. Someone else tried to make a transpiler (Sean Hickey). I’m not sure what that is. Ed Baffy tried it too. Scratch programming for microcontrollers.
You have to go to the final unless you can’t. Gimp->Scale Image to compress files. Big files take up space and make the site take longer to load. Image Magick can compress images in a directory at a time. Graphics Magick is the same thing basically. I should look into that. Learn from mistakes, don’t keep making them. Uh oh, about to get called out from the issue tracker. I did talk to my TAs and peers first. He requested that we “go up the chain” as they say, but Gavin did tell me to put it on there. I’m ready. Okay, it happened. I survived. Just me being the useless piece of trash that I will always be. A speck of insignificance not worth even a glimmer of effort to fix. I never do anything right. I am useless. I provide no good to the world. All that I ever give is discomfort and disappointment. Never ever ever good enough. If you’re never good enough you should just stop trying. There’s not point. Also on another admin note is that the inventory is stocked because it’s stocked. Sewing—I’ve been looking forward to this since the beginning of the class.
Brother SE 1800 for heavy sewing. I wonder why it’s called brother. Casting/molding is underhyped and can do a lot of what 3D printing can’t do. Someone made a foosball set. We will be machining our molds (3 axis machining). You can remove material faster than you can add material (rough vs careful cutting). I could make a foosball table. That would be awesome. Coffee cup lid, injection molding. Sprue sends in plastic along runners. Vent for air to come out. Parting line where mold seams join. No flashing if you machine it well. Insert – package electronics like the FabISP. Vacuum forming is lower resolution but it works to package things in stores. We’ll be doing flexible molding. Rigid molds require releases and are still kind of hard to get out. Flexible molds can have inserts and stuff too! Flexible molds can have lips, rigid molds can’t. Tesseract with a 6 part mold. Using an entire face line up instead of just bumps is much better (it will make for a tighter constrain so you have less flashing). This week is about Smooth On. Reynolds can do classes, they’re over in Brighton. Dick Blick is an art supplier. Boat people use molding and casting a lot. Aremco works at high temperatures. FlexWax melts just above room temperature. Soften in hot water and mold around a thing. It’s great for copying a thing. It’s not skin friendly—it’s still too hot and inflexible. We’ll be using machineable wax. You can remelt it lots of times (even the shavings so long as you have a clean end mill). Rigid foam insulation machines well. It’s a coarser finish (good for big molds) so you seal it with Blick White Gesso or something. You can bring a hot air gun you will melt back the surface to seal it. Instamold by Activa can make a face mask or something. Urethane—strong rubber. It’d be cool if we went over FX prosthetics. Halloween is right around the corner! Smooth cast series of urethane. Sometimes urethane sticks to stuff though. OOMOO Series silicone mold. Don’t need any ventilation and nothing sticks to it. PDMS is a fancy Silicone. Dow Sylgard can mold and cast down to nanometers. Don’t use rubber latex for anything ever, it’s terrible. Only ever makes sense if you’re applying to a large object that you can’t move. Can you be a member of a fab lab out in the community? Is that what a hobby shop is sort of? Drystone (from plaster.com) isn’t just plaster. Plaster has nasty surface and structural properties. Calcium sulfate compounds that don’t dry. You hydrate it or something. Calcium sulfate is used as a coagulant in food. Yum. Plaster dries Drystone to set it. 27$ for 27lbs. It’s like runny yogurt (easy to get in a mold). It looks like injection molded plastic. Portland cement is an option. Hydro-stone. Dry Stone looks a bit better, hydrostone is slightly stronger. They both work well with OOMOO. You can make coins! Low Melt Fusible Bismuth based alloy (Roto281F) Antimony Tin alloy. $15/lb. Eutectic. You have some time to set, it’s not going to stop flowing right away. Dust the mold with talc to protect mold and wet the mold. It’s a bit brittle. Stiff as a metal part but not as strong. It polishes beautifully. 400F top end of a toaster oven for Cerrotru. These higher end materials start needing ventilation. Aluminum is more hazardous. Very high temperature. Food safe SORTA-Clear. Check out chocolate cats. I could make a million Pikachus to spread around the EECS department. Reynolds has some stuff. Custom ice cubes, candies. I could finally do sugar rupees. Epoxies for sealing electronics. Hydrostone and drystone bad in tension great in compression. Best optical (transparent) materials are dangerous. Add chopped fibers in a filler system. Using additives to change the structural properties (add tensile strength for example). Bag of hallow glass spheres to decrease density, quartz for thermal conductivity—sounds like potions class. Do a little test cut first. New group project: test cast of each material to verify integrity. Wear a mask for fillers. This kid in the back knows everything already. Shoot. I wish I knew things before coming to class. But I’m incompetent. Mix until homogeneous, 5x more stirring than you think you need. If you don’t do it right it just makes goo. Don’t dump or you’ll trap air in the mold. Tilt the mold to have the fill integrate across it. Get to the vent last. Careful how you mix to not introduce air bubbles. You can use a vacuum or compression chamber to get all the bubbles out. Epoxy heats up and you can’t really stop that. It can smoke, catch on fire and other fun stuff (exothermic). Demolding – angle faces in rigid molding. Dilute dish soap, Vaseline, talc, bend the piece. Be careful with containers of resin (keep clean, esp the rim/lid). Don’t go near Crystal Clear acrylics. This is not for home use. These are super hazardous (hood with tremendous air draw). They will mess you up for life. Optical epoxies that are quite clear. Check the Safety Data Sheet. Also part of the assignment now is reading part of the Safety Data Sheet. Clean as you go to not accumulate all kinds of goo and dust. Rough and finish cutting. Machine the mold this week. Use an stl file. Rough cut – horizontal motion, terraces. Neil learned about zooming in MyPaint. I feel a little better now. We all are always learning. We’ll be machining on the same things we used for the PCBs I think. Get the stepover right (~1/2 for rough cutting, less for fine cutting). Machining time vs surface finish. Get something done and then go back and fix stuff – theme of this class. Ball end mill can get deeper in a bowl. Ball end mills will give scallops on flat surfaces though. Depth of cut. If you’re making really skinny things you may collide up higher on the tool (the thing holding the endmill). Design around your endmill/milling device. ShopBots can work too. Adaptive clearing – use the whole side of the tool – bottom up. Neil is porting a slicer to mods for 4 or 5 axis machining. We are getting a couple inch block of machinable wax. Crucial in industry, empowering.
Bubbles on top cool, bubbles on inside not good. Oomoo 25 is what we’re using. 1/8th “ mills. Flat and ball nose. Looks like a plus sign rather than a minus sign. We’ll be using a mini shop bot. Secured with double stick tape. Stls don’t have dimensions so make sure to tell it how big it actually is. Only machining top. Reconfirm area size. Fit to material. Top of cut at zero. Cut claim (bottom of cutting) as bottom of design (not bottom of piece). Don’t cut all the way through the material. Roughing tool path and then smoothing tool path. Pick tool, stepover, cut depth, spindle speed and feed rate. Too slow melts wax, too fast shakes the whole table. 3D raster along y and then calculate. Finishing toolpath, only real difference is smaller stepover size. 10% is decent. Multiple passes in rough cut, single pass in finishing cut. This software is going to try to finish the top, which is just a waste. Roughing and toolpath preview. Run through post processor to create g code for the machine to read. Save toolpath to a single file. You can do it one at a time if you want. Cut 3d is now separate from stl file. Toolpath is another, third, file (or two files, I guess since there are 2 toolpaths). Now zeroing. Shark icon (CNC Shark). Two safety stops. Rotate to turn back off. Then plus green button to turn it back on. Hit jog on home panel. Pg up and down for z, arrow keys for xy. Press buttons to zero. Use touch plate for z. Hit detect touch plate. Only checks it once. Check router on off switch on end mill. There is a brush and dust collector. Be sure to switch on and check that bag inflates. FRO slider bar to about 50% to start. Hit start. Then okay. Red facing you is good. Slide FRO up to 90% or so once it’s going. Try not to have to clear out a large area before scaling up. 5% for finish. Sam Beldon Brass Rat copy.
Molding training summary by Priyanka Chatterjee: You need a 1:1.3 ratio of Oomoo Part A to Oomoo Part B, mix together in a plastic cup, transfer to another plastic cup, and then pour it carefully (to avoid bubbles in the mold) into your wax mold). The drystone part is also easy -- the instructions of mixing the powder with water are on the box, and it doesn't take long to do.
FYI this assignment will be "late" since I will be out of town Sun-Wed and I didn't have the gung ho to get it all done before leaving. The idea for this project is from my final project. I'm not really sure why I like creatures being the supporting structures in my designs. I'm thinking of using Atlas figures to hold up the lampstand for my final project and I was going to make them this week in molding and casting. I'm having trouble finding an STL though (because I certainly can't just make a great human-link STL in a reasonable amount of time) so I may switch to something else or end up making a simple design. But I wanted to see the limitations on molding and casting detail so I will try harder to find an STL before completely giving up. There are rupees though so that's my fall back. I do plan to go through all the Blender tutorials one day when I don't have to worry about classes.
I eventually found an STL for a rupee magnet, shown below, which seemed doable.
However, I was under the impression that the mill could operate like the laser cutter in that I could manipulate the design some what (make copies, add holes and such) after loading it up. This was not the case so I had to take the time to sort of redesign the whole thing. I made a lot of use of the loft tool to make facets. The result is shown below.
This was being problematic since I designed it thinking I could switch to a smaller bit if I wanted. I'd also anticipated the driver being willing to drill 1/8" holes like a normal drill but it ignored those sections. Rescaling it proved to be more trouble than it was worth so I went with it. I only found out later that Gavin had been encouraging one sided molds, which would have been much simpler. We'll see how it goes. I'm still waiting for the Oomoo to set as I type this up. In the milling process, so many things went wrong. I started by printing it with the block oriented the wrong way. I tried this 3 or 4 times before I realized the problem was not just a bad offset. Check it out:
Note that this mold has holes (extended triangles) on both sides unlike the CAD drawing. That's because I was trying to decide which one to use over the course of trying to make the machine print more of my details. I also sometimes forgot to calculate the path before sending it, or recalculating after moving the end mill around. I kept pixel count too high some times, had inappropriate scaling and failed to effectively tape down the wax. For your educational benefit, the resulting torn up wax looks like this:
It's a shame I didn't take a picture but by that time I was pretty frustrated so the amount of tape I used on the next run was ridiculous. It also happened to be the run that actually worked. Tada!
Which got me to the point of actually setting the mold. I just eyeballed the Part A and Part B thing, so we'll have to see how it turns out in a few hours. Aaanndd it looks like it was fine!
The Dry Stone did not become a good texture but I used it anyways with decent results. When I say it didn't have good texture it was very dry and chunky. Later experiments proved that the "let it soak for 1-2 min" step is actually important. Anyways, here are the first round of gems (rupees), as of yet unpainted (and perhaps forever unpainted since I'm not sure I have the right paint, though they should be painted):
I accidentally made too much Oomoo and I left it in a cup, which then became one of the coolest things ever. I'm fer sure an adult.
IT EVEN STICKS TO WALLS!! (not because it's sticky but because it has a bevel that acts as a suction cup)
I didn't think about how leaving the triangle as a hole wouldn't work unless the mix was very fluid and that it would be filled with the mix as you can see in the finished picture above. I decided to cast a second half with no holes so I could do two sided no holes (just spreading it on like peanut butter then closing together the halves worked okay and since the mill didn't cut out my guidelines I figured I might as well continue to just eyeball it). I made extra Oomoo again but made myself resist the urge to make another cup and instead tossed in my class ring. This was pretty nerve wracking so I hope it comes out okay!! UPDATE: It came out seemingly just fine and with a nice battle ring to boot! I purposefully didn't complete the bottom of the ring in drystone to help get it out ot the Oomoo. Plus now that I think about it it will also be useful for temperature changes if I ever were to actually wear it.
It's kind of a funny story about the ring since I only just got it back from losing it and now I have two! Also, the second round of rupees came out like this (below). They really do need to be painted. You can see in one where I rushed a lot and created huge air bubbles.
Sidenote: here is the command line stuff to get Fab Modules up and running:
We never did this as a group that I noticed. We really only had one process in EDS that was supported though. I guess we had drystone, hydrostone and OOMOO. I used drystone and OOMOO only I think. I liked drystone because it was quick but the OOMOO texture was really nifty. The safety data sheets are: OOMOO, Hydrostone, Drystone. What I got from these is that drystone is basically inert, hydrostone is a bit more dangerous (though the only labelled difference is that it has portland cement in it while drystone has some other chemicals). Hydrostone is more basic, which is the reason stated for it's slight heath hazard (as opposed to minimal health risk in drystone). OOMOO is a slight health hazard as well but the SDS isn't as helpful. I says if you burn it it can release harmful chemicals. It suggests wearing safety glasses and gloves.