In this, my first instructible, I detail how I (with assistance) created a variation on a prop from the BBC show Torchwood. The prop in question is the "Vortex Manipulator" found on the wrist of Captain Jack Harkness. Below is a screenshot from the show of the item we are trying to replicate with some small changes. My intent with this instructible is to show how I went through the data collection, model manipulation, replica testing, mold making and pewter (yes, molten metal!) melting, pouring, and finish work.

Step 1: Getting the Details

The first step involves gathering as many details about the item as possible. Through screenshots of the actual show as well as a Google image search we were able to find quite a few good images that give details on the actual device under the leather strapping. I saw that there are plenty of people out there making this replica and the majority of them appear to be plastic or resin copies. In order to be a little different I decided that a pewter metal cast would be the ideal way to stand out from these other copies. This step is rather uninteresting but crucial to the process.

Step 2: Making a Digital Model

I gave the 3D model my best shot using various software packages like AutoDesk's 123D (http://www.123dapp.com/) but my lack of experience in modelling really showed itself at this stage. After about 9 hours of work, the model that I created simply had the basic outline of the device. At this point I called in some assistance. I have the incredible luck of being friends with someone who enjoys taking a collection of images and turning them into a 3D model. Using Rhino 3D (http://www.rhino3d.com/) he was able to create an incredible 3D model (as shown in the screenshot below). I made the decision to forego the hood shape found in the center and change it to a small "display" of sorts. The reason for this was that I will be cutting this model on my small CNC machine and want to avoid undercuts. I could possibly fashion a hood of sorts to sit in the small channels but that will be a finishing decision.

Step 3: Trial and Error (Foam)

First, I took the 3D model and used Vectric's Cut 3D software http://(http://www.vectric.com/WebSite/Vectric/cut3d/c3d_index.htm) to transform the 3D file output from the Rhino software into a set of g-codes files. These files are the instructions that the machining software uses to tell the cutting head of the CNC unit where to move and at what speed to accomplish our output. I decided to make a few test runs of the actual model (rather than the inverted one for the mold) to see the detail of the model. My first run is in the green insulation foam as shown below. The first image is about half way through the roughing pass, the second is about half way through the finishing pass, and the third picture shows after the finishing pass has been completed.

About my CNC setup: It is a 7"x7" Zen Toolworks DIY kit (http://www.zentoolworks.com/product_info.php?products_id=74) controlled by EMC2 Linux software (http://www.linuxcnc.org/) and it is fairly obvious from the below pictures that it is desperate need of a vacuum attachment for clean up.

The foam model was slightly disappointing. I think that the roughing pass was not close enough to the finishing pass so the finish bit had to remove too much to leave behind a smooth finish. This is readily apparent on the outside edge where the finish bit traveled down to the outside bottom and made contact with the edge. This will have to be remedied before the wood cuts are performed, otherwise broken bits and machine could result. (Lesson learned, watch your paths and use foam first)

Step 4: In the "woods"

To continue with the trial and error, I decided to create another of the positives in "wood". I wanted a soft consistent wood and thus decided to use MDF fiberboard. Not wanting to purchase a 2' by 4' piece of MDF when I need something less than 4", I discovered that your local Home Depot (and likely other similar stores) has free sample of flooring, most of which is a fairly thick MDF type of material. So, with a few such samples in hand. I performed another positive cut after tweaking the cutting parameters in the Vectric Cut3D software.
The results were much more impressive than the foam test, thought the pictures don't really tell the tale as well as the physical models.

Here is a video of the cut out pass, which didn't go deep enough as shown in the pictures:

Step 5: Finished! (Not Really)

It's at this point that the friend that I am making this prop for advises me that the wood master that I just made is "good enough" for his needs. He takes the wood master and paints it and creates the cool looking prop you see in the pictures below.

Next, we experiment with molds...

Step 6: Testing a Mold

As mentioned, my intent was to make a pewter cast of one of these masters to make something a little different than the other models out there. It will also get me more comfortable with casting pewter (something I wish to to more often).

This first trial involved taking the previously created foam model, placing it face up on a piece of wood and covering it up with a complete layer of drywall joint compound. Once that was dried, I puzzle over how to remove the foam model. After watching a few videos, I thought that a "lost foam" cast might be the best bet. With that in mind I performed a test pour. I placed a wedge of the foam in the sand of my sand fire pit (created based on the awesome instructable by jonsarriugarte https://www.instructables.com/id/Sand-Fire-Garden/) and poured some melted pewter into the foam wedge. The result was sadly disappointing. The metal is not hot enough to melt this type of foam in a quick and efficient manner, leaving behind a solid plastic-like residue that needs to be chipped off.

Here is a video of me pouring melted pewter into the foam wedge. The melting device is a Lyman Big Dipper and comes in a handy kit. (http://www.lymanproducts.com/lyman/bullet-casting/Big-Dipper-kit.php)

With lost-foam out of the question, the next test was to see how to effectively remove the foam. Using a heat-gun, the end result was again that same residue mentioned above. In the end I discovered that a blowtorch was the required item to completely burn away the foam. Surprisingly, the detail of the foam was definitely still there despite the scorching left over form the blowtorch's heat.

Here you can see the heat gun melting the foam in the mold. I did not film the blowtorch removing any remaining residue because I had both hands busy (and a fire extinguisher nearby).

Next. we pour some pewter into this mold.

Step 7: Mold Results

In this next step, I melted and poured the pewter directly into the mold created previously. As you can see from the pictures below the result is less than ideal. The texture was more coarse than I expected but the details (slightly enhanced in the foam master with pen tip) appear with decent detail. You can see the height changes from the pen tip that deepened the smallest lines (Lesson learned, should have left it alone).

After some work with a Dremel, the surface texture smoothed out somewhat but the line details are still rather poor.

After some research I discovered that some people have successfully cast pewter in MDF wood molds, so I inverted the digital model , created new g-code files and began the process of carving a mold in which to pour the pewter.

Step 8: Getting 'Mold'ier

Creating the inverted mold was fairly easy with input from friends. Below you can see the process and end results of creating this mold in MDF (again using a flooring sample).

Step 9: Hot Metal!

I wish I could say the mold created previously resulted in a perfect pour but that would be untrue. While the mold looks just fine, what I discovered (among other lessons learned) is that MDF is probably best suited to small quickly cooling molds. The size of this cast resulted in not only scorching of the wood but also smoking. The smoke has to escape somewhere and that will be through the cooling metal resulting in some loss of detail and even flaws.

As you can see in the below pictures, the process is fairly simple. Melt metal, pour metal, let cool.
This first video shows how I failed to let the ladle heat up enough to carry the metal to the mold. That was rather embarrassing, but even more embarrassing is my lack of full safety gear. I had appropriate eye protection on but neglected to put on the leather gloves appropriate for this operation. Fortunately I did not have to learn a difficult lesson.

After allowing the ladle to heat up, I poured again. This time I learned another lesson, the ladle should be of appropriate size to fill your mold. The result of this pour had great definition on one side but the other side was full of gaps around the model. This was likely due to both smoking and re-pouring metal on already cooling metal. Lesson: One smooth pour.

Step 10: End Result

As mentioned previously, the pour of molten pewter into the mold resulted in less-than-stellar casts. The final cast was the best and has the best detail. It could use a really detailed clean up and maybe I will come back and post pictures of that at a later date. (Update: ran a fine abrasive wheel over the last pour, pictures below)

I've learned lots of lessons during this experiment and I hope that they have been found informative. Not everything goes as expected but every chance to Learn and Do should be explored fully. Please enjoy the below pictures of the several pours of this cast.

Lessons:
Be Safer than you expect to need
The mold is fine, leave it alone
Heat your pouring container
Have enough material to fill your mold in one smooth pour
Don't expect perfection on your first try.
and
It's OK to fail at something when you learn something (and, boy, have I learned a LOT!)

Thanks for reading! Let me know if you have any questions of suggestions! I hope to be able to do more of these. I plan on using soapstone for 2 part molds in the near future.