Molding & Casting

I took these concrete plant pots for succulents as a reference and started working on my mold. I noticed that to make a plant pot, I would need my wax stock to be much more than 35mm in height. Sadly our machineable wax blocks were not tall enough. So I decided I'd make my own wax blocks by melting them.

Then I created a stock material to cut out the negative mold of my object from. This will basically be the rubber mold that I'll make after machining it. I placed my object inside the box, used combine to cut out it's dimension from the stock.

Now I had a stock material that had a hollow portion which could contain my 3D model.

Now that I have a working negative mold, I needed to create a positive mold that I could machine. First I created a new wax stock, now this stock will be the actual block of wax I'll mill.

Then I aligned everything on the wax block and used another combine operation.

With that, my positive mold was ready, but I had to add some small features before it was machinable. I added 1 degree drafts to all sides of the mold, added 1 big hole to pour material inside and 3 smaller holes for the gases to come out.

I decided to use recycled machinable wax for my mold. This was done for two reasons

• The wax stock was only 35mm in height, I wanted my mold to be a bit bigger than that in depth.
• I watched some candle making videos on youtube, and always wanted to try it out myself. This was a good opportunity to try.

First I tried to find a safe melting point for my wax which I could use in our convection oven. Since the container I was using was made out of Stainless Steel, I had to make sure the oven was running in convection mode not in microwave mode.

I found the safety datasheet for the specific machinable wax I was using which can be found here . The Melting point for this wax was at 115.6 degrees Celcius and the flash point for it was 301.7 degrees Celcius. After some trial and error, I used 250 degree celcius and it took about 50 minutes to completely melt and get liquid enough to get rid of the bubbles.

I made some molds out of waste acryllic sheets at the lab and assembled them with masking tape and a light coating of acryllic glue to make it water tight. The surface of the acryllic seemed smooth enough for making the wax ingot motds. After that I carefully poured the molten wax in it and let it cool down. As the machinable wax cools, I found that it shrinks a bit, which made removal of the wax amazingly easy. The Acryllic molds were reusable and didn't even need to be taken apart to take out the wax.

I tried exploring Fusion 360's Manufacture option in our CAM week, but I was overwhelmed with all the options. The usual workflow at Fablab CEPT for CAM tools is to use partworks for generating toolpaths. But due to some technical errors, the lab laptop which had partworks kept crashing. So I was left with no other option than to make fusion's CAM tool work.

The Best resource I found so far for CAM in fusion360 is this video by Ecan And Katelyn, I have been following this channel for a long time now and they make exceptionally good videos with the right amount of information with the sprinkle of humor.

---

---

I started with the ER-20 collet by physically measuring it and adding all the dimensions into Fusion's Library manager.

During our Machining week, I found a 3mm endmill which seemed excessively long to me. I didn't really think it would be appropriate to use this bit becase the lateral force on a long bit like this while cutting hard materials would make it bend.

Machineable wax and Foam are the perfect materials for this sort of endmill because it gives you a long clearance to work with.

So I measured the bit and prepared the tool library for it.

Feedrates:

I found this really nice article from Autodesk on milling machineable wax. Handnote had the feedrates for machining wax without melting it.

Now we need to create a Setup for our CAM process. This is where we will define what to cut, and what to cut from and the co ordinate system we will follow. From Evan and Katelyn's Tutorial, I started with making a Stock first. And machined my model out of it. Here are the settings for my setup.

For generating the Roughing Toolpath, I used the Adaptive Clearing option. It takes out large quantities of material effectively. It is unique in that it guarantees a maximum tool load at all stages of the machining cycle, and makes it possible to cut deep without the flank of the tool without risk of breakage.

The strategy first makes a series of constant Z- layers through the part, and then clears them in stages from he bottom upwards. Because it can cut so deeply, the fist step down at each stage should be the effective cutting length of the tool. Then Clearing of the intermediate layers proceeds into the shallower layers to maximuze the efficiency of the tool use.

These are the settings I used for the roughing toolpath. There are way too many options that need to be fine tuned to master these settings. Sadly I didn't get enough machine access to try out everything.
I depended on the simulation option to optimize as much as possible and get a workable toolpath.

For generating the Finishing Toolpath, I used the Countour Operation. This is the best strategy for finishing steep walls but can be used for semi-finish and finish machining on the more vertical areas of a part.

If a slope angle is specified, for example 30 to 90 degrees, the steeper areas are machined, leaving the shallower areas up to 30 degrees for more appropriate strategies.

These are the settings I used for the the contour toolpath. My selection of the contour process for finishing was probably not the best of ideas. The result from this finishing was worse than the roughing toolpath.
The Simulation did show me rough edges, but I didn't think they would vary so much.

The Best feature I found about Fusion's CAM tool, is the simulation. You can simulate and see exactly what will happen during the machining process. The simulate option can be found by right clicking on your setup.

Only when you're sure about your Toolpaths, you can go to machining. One downside is that calculating the toolpaths is a bit processor intensive for my low spec machine. It takes about 5 mins to generate and calculate a toolpath for the simulation.

Exporting to the shopbot is super easy from Fusion360. All you have to do is click on post process > select a post processor that's suitable for your machine and click on post.

My first Attempt was on Friday. I started machining the toolpath I made with Fusion. But at 33% progress, it was closing time already. And I was left with this partially done Wax block. The finishing looked quite good and I was sure that this would work. The only problem was that the simulation time wasn't matching the actual time.

I knew the feed rate settings were right as the wax chips were all fine powders. If the wax was melting they would have been thicker.

The Second Attempt was on the next day. I had trouble getting a slot for Machining as people from CEPT were using the CNC, but this was the result.

I knoew something was wrong, but had no idea what I did wrong in Fusion. There are way too many options where I could have gone wrong.

No one at my Lab Uses Fusion 360 for the CAM process, they use partworks. I Couldn't debug it in time so I went back to partworks.

First thing I did was to put the two wax blocks and the waste back into the oven. I was collecting all the wax dust that was coming out all through the process.

I made a new design, with both the top and bottom molds in one wax stock. In order to reduce workload. It's the design You can see on my design part of this week's documentation.

Generating toolpaths in Partworks is like a 'kid mode' to machining. You just follow through 7 steps and your toolpath is ready. It does not have many options but it gets the job done and it is the standard CAM tool used in Fab Lab CEPT.

The reason I could not use partworks earlier was because the lab laptop which had partworks kept crashing. So first I fixed it by removing that annoying windows watermark, cleaning junk files, cleaning temp cache, removing everything from the desktop and removing redundant bloatware.

Then I generated my toolpath in Partworks using the same feedrate and spindle speed I used in Fusion 360.

Someone had fiddled with the Shopbot Control Software and it kept getting an error that the limit switches on X and Y were on. None of our G-codes were working. It was a disaster! It was Tuesday and none of our Machining was done, let alone testing the casting materials.
Our Instructor was not here either and the only thing we could do was check our previous year's documentation on the default settings of the shopbot. And we found something on Aditya's Week 8 documentation. The limit switches 2 and 3 were always on, on his images as well. That's why even going back to the default settings were not working. So I disabled the limit switches and the machine started to work again.

In partworks, I used two bits for my toolpaths. And one of which was that insanely long bit. My Rouging toolpath went all well, but when I swapped the bit I had to re-home the Z axis. But this bit was not long enough and I didn't notice the clearance. The collet touched the outer wall and broke it. And this machine job was a failure as well. This happened earlier on Wednesday afternoon and now I was completely out of options on how to present for this week.

I wasn't using the modella because I wanted to explore Fusion's CAM tools. While I was struggling with the shopbot, other's were struggling with the Modela. For some reason, our stl files were not working on fab modules. It was loading when we select the mesh(.stl), but stopped responding as soon as we clicked on 'calculate height map'.

So we went back to 2016's documentation and saw that an older version of fab modules, which was completly offline, worked for Rudraprakash Solank. We located the laptop he used and found the local fab modules program, ran our files in that, and it worked perfectly.

Finally, I milled the new design using the old fab modules. First I used the Rough cut, with the default settings.

Then I used a Finish cut, with 0.85 overlap. The total milling took about 3 and a half hours. I was sititng infront of the machine with a vaccum the whole time, since I had no option for it to go wrong this time.

The first casting material I tried was Smooth-Cast 305. I mixed 30gm of part A with 27gm of part B, and poured it into my mold. During Curing, the smooth-cast 305 kept expanding. The pressure was causing my two part mold to come off.

I assumed this was happening because I was pouring in too much material. So I tried again while not filling the mold fully. I assumed the material was expanding as it cured. But as you can see, this is a recurring problem even when I reduced the material in the mold.

I googled a bit for probable reason for the bubbling, because it didn't happen for guys who did the casting on the day before. I found that bubbling might happen due to the following reason;

• Humidity: since the Smoothcast-305 was opened 2 days ago and was used by many people, it probably came across humidity.
• Pressure: In the product page for Smoothcast-305, it is mentioned that the best result can be obtained by curing the material in a pressure chamber while subjecting it to a 60 PSI pressure for the full curing time of the material. It would certainly explain why bubbles were forming in the cast on the inside while not on the mold surface. It would also explain the sudden expansion of material.

Next I tried casting wax, since we had alot of machineable wax chips everywhere, I just melted the and poured it into my mold. The Oomoo30 product page specifies that wax can be cast in it, so I just went along with it and I was really impressed with the output this time. There was no leakage from the mold, the edges came out sharp and the finishing was exactly as the mold.