Group Assingment: review the safety data sheets for each of your molding and casting materials, then make and compare tests with each of them. Individual Assignment: design a 3D mold around the stock and tooling that you'll be using, machine it, and use it to cast parts..


The materials we were using for our molding and casting week were SMOOTH-ON Mold Star 30 Silicon Rubber and Smooth-Cast 320 resin. I have never worked with this material so I was excited to learn.

---> Mold Star 30

Mold Star silicon rubber is supposed to have a good shelf life with little long-term shrinkage. It is ideal for making parts with with the following casting materials such as; casting wax, concrete, gypsum, and resins. It requires a room with good ventilation but is relatively harmless to human touch unless one is allergic to silicon. It takes aproximately 6 hours to cure. Unless it is poorly mixed or if it comes into contact with a contaminant, it is a relatively easy material to work with.

The safety data sheet can be viewed HERE. When reading, it is generally classfied as not being hazardous. For those of us working with kids, this is good news.

---> Smooth-Cast 320

Smooth-Cast resin is used to make prototypes, sculptures, props, and small objects. It is supposed to be durable, machineable, and can be painted. If exposed to temperatures over 30ºC they suggest using a foam backing for lightweight support. Temperatures over 94º C the resin is considered flammable. The results of this resin are off-white and take 10 minutes to cure. As it mixes it gets hot to the touch but nothing too crazy.

The safety data sheet can be viewed HERE. The hazards section is significantly larger than the silicon.

Resin requires much more care when handling. It will irritate the skin and can dammage lungs if inhaled. It is also important to wear safety glasses and of course gloves. Clearly, you should only use resin in a well ventilated area.


For my mold model, I decided to make a pixel ring holder. Something I might end up using in my final project. The idea is to hold a pixel ring around a custom circuit in the center. I went through a couple iterations until I finally settled on one that made the most sense for this weeks assignment.

---> version 1

My first model was going to make it possible for me to hold multiple neopixel rings. I would put the custom circuit on the back. The first step was to find the dimensions of the 16 & 24 LED Neopixel Ring from the Adafruit website HERE.

16 NeoPixel Ring
---> outer diameter 44.5mm
---> inner diameter 31.7mm
24 NeoPixel Ring
---> outer diameter 65.6mm
---> inner diameter 52.3mm

I added channels underneath the pass the wire and connect to the custom space circuit board. I began to think this model, simple enough for 3D printing would be pretty challenging for my first 3D milling, molding and casting. I realized it woudl require a two sided mold. So I simplified.

---> version 2

Instead of making a ring holder for multiple rings, I decided to make a holder for just a single ring. I could use this for the Color Clock idea which is a component of my final project. I could put the circuit in the middle facing out. This version still had channels and screw hole mounts

One of the things I needed to do was to create drafts on the edges of my model. This is so the models won't get stuck in a mold. I don't think this is so much an issue with the Silicon mold as you can bend it off but I decided to play it safe and add them anyways. I used Draft Analysis which can be found in Fusion 360's Inspect panel. I selected the x,y plane, and the side surfaces that I wanted to draft.

The goal is to get everything to be green. I used 3º draft angle. Anything under a 3º did not seem to work.

I still felt like the model was too complex. Drilling holes and small features was not something I was ready to tackle in terms of the time it would take to troubleshoot. So I simplified some more.

---> version 3

I finaly settled on a simple design. The center would house the circuit, the ring in the channel, and holes in the back to make sure I could pass wires if necessary. I added a surrounding plate to help with the making of the mold. I figured this would help make a base to pour the silicon in.

Here is the final model after applying the drafts.

Finally, I decided to 3d print a test to see if my measurements were correct. Pictured here is with a version of the ProLogo board I have been working on in the previous electronics weeks. Not pictured, the wires fit nicely behind the rings. Next step is to mill this, mold & cast this thing and see if it matches the fit of the 3D print.


A Post Processor is used to process the instructions from your CAM software into G-CODE. Because all machines are different, you need to choose post processor that matches the preferenece for your particular machine. I decided to try milling wax with the Carvey desktop mill. It uses a form of Grbl but you will need to use Inventable's Post Processor in order to properly communicate he with Easel, the cloud based software that communicates with the machine. There are other workarounds if you don't mind voiding your warranty but Easel is designed to communicate with the Machine's firmware.

There are two ways to install the Easel Post Processor in Fusion 360. You can download it to your computer or link to it through your Fusion360 drive in the cloud.

To download the Easel Post Processor to your computer go to: cam.autodesk.com/posts and search for "Easel". Download the post processor "easel.cps". Navigate to your home directory and move the easel.cps file into your Posts folder located in .../Autodesk/Fusion 360 CAM/Posts/. The Easel Post Processor will now be available in Fusion 360 as a Personal Post

To link to the Easel post processor through the cloud do the following: In Fusion 360, go to Preferences menu, under the General tab, select CAM and make sure that Enable Cloud Libraries is checked off.

Inkscape - Clone Tool

NOTE:The instructions from Inventables directed me to look for my Assets folder in my A360 drive but it was located in my Fusion drive.

Inside the Assets folder are three folders: CAMPosts, CAMTemplates, CAMTools. Upload the easel.cps post processor file to the CAMPosts folder. Note: There seems to be a few versions of this post processor out there. If you are following the prompts in Easel it may be called f360-easel.cps or something similar. The Post Processor will now be available in Fusion 360 through the My Cloud Posts option of the Post Processor selection window.


I was excited to dig into 3d milling and to learn more about Fusion 360's compatibility with the Carvey mill. Once I had my model created, I had to prepare the tool paths for milling. The clamp on the Carvey can take up a lot of space and everything is milled in relation to it. On other machines, like the Roland series, you can manually position your milling bit and set the origin of the x,y,z accordingly. With the Carvey, everything starts at the clamp. This proved to be a bit of a problem as there would never be enough space and material to mill my model near the smart clamp. Also, as I made mistakes I had to work with smaller pieces of wax. The smart clamp took up to much space and I had to find ways to "trick" the machine.

Note:If you are looking at this picture and wondering why I had the straws in there, it is because the Carvey does not come with bolts for stock thicker than 1". The material is secured using double-sided tape so I only needed the clamp for homing purposes. I had to use dowel, straw, and more double sided tape to secure it in place. I later received a gift from my Fab Academy colleague Mark which was the right sized M5 1.5" bolt. This "jerry-rigged" setup worked but I don't recommend it. Inventables sells larger bolts or pick up your own M5 bolts at the hardware store. They are not as common in generic hardware stores in Canada and I was unable to get one the day I worked on this assignment.

---> research

Before starting to program all my tool paths, I visited a variety of forums and sites to figure out what the optimal speeds and feeds were for milling wax and also for milling wax on the Carvey. Here are the links I visited:
---> Machineable Wax - Technical Info
---> Inventables - Machineable Wax
---> Autodesk - Feed Rates

---> quick test

Before milling my part, I decided to attempt a quick test to check the workflow and test the Carvey post processor. I modeled a small shape and sent it through. I won't go into detail but I was happy that I successfully milled the wax. However, when it completed I realized that my tool path was wrong.

As you can see, I milled through the pin that was in the model. If you look at the image above of the tool path, you can see it spiraling down through the portion of the pin. The boundary geometry for that small pin in the middle of the model was not clearly defined when I sent over the job.

---> setup

After the quick test, I started prepping for the milling of my model. First I set the dimensions of the stock I was going to use and the position of the model within that stock. I had a piece of milling wax that was 7" (177.78mm) wide, 3" (76.2mm) deep, 1.5" high. I left a 2 mm offset from the top of the stock as well as a 2 mm offset from the right side of the stock and I positioned the model in the center of y. I have very little room on the y axis so centering just ensured that the piece fit within the stock.

Next I made sure the orientation was set to Box Point and that the origin was at the top of the material in the bottom left corner. Because of the Carvey's smart clamp the origin is set at the surface of the stock that is under the clamp.

---> facing

Once the stock settings were set, I needed to program the first tool path which was to re-face the surface of the wax. Although I was using a new piece of milling wax with a smooth surface, this step ensures that the surface of the stock is precisely located where the gcode expects it to be. It will help reduce any discrepencies. The fact that my clamping mechanism was not as precise as it would be with the right bolts, this step was even more relevant. It meant I could mill the wax to the right depth before starting to cut the geometries of my model.

The first setting was to choose the right tool. I used only one tool for this entire milling process.

---> 2 Flute Straignt End Mill

To learn how to import a tool library into Fusion I visited the following:

---> Importing a tool library
---> Carvey - Fusion Tools

After the stock setup, I selected the feed and speed settings. Generally speaking, keeping the spindle at 10000 rpm for finishing or facing was fine.

I made my model with a square base thinking it would help to create a perimeter around it. I selected this base as a boundary for the geometry. This worked out well.

Next I selected the depth per pass by enabling Multiple Depths and selecting it to mill in increments of 1 mm.

Lastly, I selected the Bottom Height to equal "Model top" and the Top Height to be "Stock top". You will notice Bottom:-2 mm which is the model offset that I had entered in the initial stock setup.

After applying the simulation, the Facing operation looks good.

This process went well.

---> pocket clearing

Next process was to rough out the caveties of the the model through 2D pocket clearing. I selected what I thought were the right geometries but as it turned out, I was wrong. The tool path I had selected cut through the model so I stoped the process.

I removed a layer of wax using a saw and prepared to try the process again.

I then performed another facing operation after fixing the geometries so it wouldn't cut through the wall of the model as it did in my first try and I attempted to carve it again. The carving process was successful although it took a long time. The problem I was now facing was that my model takes up almost the entire width of the material leaving very little room. This resulted in a very thin perimeter wall of my model. The tooling was perhaps a bit agressive as well. Cutting through the pockets was a bit harsh and I had to pause the job a couple times to clear debris.

I decided I would do this again. I made some adjustments to my model by changing the wall thickness so that I would have a little more room to play with. Of course, I was limited by the fact that I was bound to the dimensions of the 24 LED NeoPixel ring I was using.


Things started getting interesting in my 3rd attempt to mill this model. I was now working with less than half of the amount of material I started with. The good news was that I could now use the largest bolts I had on hand to secure the smart clamp. The bad news, there is no way I can position my piece under the clamp because it would obstruct the milling process. Surprisingly, it is hard to find good examples of people using the Carvey for milling wax or circuit boards but I did find a good tutorial that inspired me to "trick" the machine.

---> Machinable Wax, Carvey, and Fusion 360

In this tutorial, the user uses Fusion to trick the Carvey into thinking there is a bigger piece of stock in the machine than there actually is. He then changes the G-Code start height by adding G0Z2 before the start command. While I chose a simpler way of cheating the machine, I now realize that I can bypass the smart clamp by using similar techniques.

First, I found a stack of materials that closely matched the height of my stock.

Next, I positioned the stack under the smart clamp and moved my material into a position where where it could be easily milled. I made the measurements so that I knew what size stock I would need to input in Fusion.

---> 216 mm x 102 mm x 28.5mm
---> Offset from the bottom of 25.8 mm

Next, in Fusion I set my stock accordingly.

To make this process easier, I installed the Carvey Wasteboard model into my Fusion Model. To get a Fusion model of the Carvey Wasteboard visit:

---> http://bit.ly/wbfusion360

---> facing

After the initial setup, I performed another facing operation.

Then another 2d pocket clearing for the main areas of the model.

After clearing the main areas I did another pass of the pockets at the back of the model which are going to make the holes to pass the wires of the NeoPixel it is designed for. I got a few warnings about the bit crashing into the stock. I watched the simulation and it seems to be a non-issue. In the end, it was not a problem although I am sure there is something I am not doing properly.

Finally, after milling the pockets, I did a contour pass. This helped to translate the 3º draft that is designed into the model. The RPM of teh spindle was increased to 12000 rpm to help smooth the surface but this did not help much because the 2 Flute Straignt End Mill I am using is not the ideal for this process and left a pattern on the side of the milled part. The To Slope Angle was set to 87º. This is the difference created from the draft.

Once the milling process was complete, I collected the shavings and investigated the results.

I was very happy with the results. A lot to learn here but I am I think it looks pretty good. You can see the texture along the side face of the model. A milling bit with a ball end would have been more subtle.

The fit is perfect!


Making the Silicon mold was relatively easy compared to the Milling process.

---> casing

First step was to make a case. I went to Makercase and entered the dimensions of my model and cut the case out of Acrylic which will not bind to the silicon rubber of the mold. I taped up the box and put the model inside.

---> silicon rubber

I used the 2 part silicon rubber that was described at the beginning of this page and measured equal parts A & B.

I measured the mass of my model on Fusion and used that to get an idea of the minimum amount of silicon I needed.

I poured A & B into seperate containers before mixing them into a third. I took the images on the box a bit too literally. The problem with this is I ended up ruining two containers. When you mix the materials in the same container you can peel the silicon off once it cures. As individual liquids, they are very sticky and hard to remove.

Once I mixed the materials and poured them over my model it looked like I made the perfect amount of silicon. It poured smoothly and I let it settle until my next visit to the lab.

The final results looked great. The side of the mold is very thin. I could have made a bigger box to help strenthen that section but other than that, I am very happy with the results.


I put the mold back in the acrylic box. I mixed 2 parts A & B of the csting resin mentioned at the start of this page.

---> 1st attempt

I was not in love with the off-white version of the resin we had. My instructor suggested I try adding die. I took some organic squid oil and poured in a couple drops and basically invented a good prop for moon rock. The resin got quite hot and started frothing. I was a bit surprised. Not at all what I wanted but it was kind of cool. Luckily, it did not affect my mold.

---> 2nd attempt

For my second attemp, I did not mix the material long enough. It never set properly and was essentially a greasy mess. Pouring is a bit more tricky especially around the holes. I think I would make a deeper model next time just so there is more room to make a thicker base.

---> final attempt

I tried one more time. I mixed the parts A & B longer before pouring. It looked a lot better. I tried to pour more carefully to cover the thin areas but it is still a little thin in one area. The end results looks really good. I was a bit blown away at how much detail transfers over.


Here is an example of the new casted piece next to my 3d print test. And finally, I check to see the fit with the NeoPixel ring. Pretty good I think!

Lastly, a view of all the trials.

Creative Commons License
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.