The full-size AAVP7A1 has cast metal track links, with a hard rubber traction pad on each link. The links are 21 inches wide. Below is a close up of the full-size links:
My 1:9 scale links are cast aluminum with cast-in-place neoprene traction pads. Not all details can be reproduced by gravity casting aluminum (at least in this scale), so my model's links are missing a few details: The fine rib details on the bottoms of the links are missing and the alignment blade on the top side of each link is a little wider and shorter than scale. Also, there are no neoprene cushions on the top side of each link.
The process of making links began with the creation of a master link. Because I do not have CNC machining equipment, the master link was made up of several individually machined parts that had been glued together. After gluing the master link together with CA, epoxy putty was used to produce the final shape of the link. The master link has long 1/16" diameter pins where link pins would be on a real track link.
Plaster molds can only be used one time for aluminum casting - thermal stresses and the demolding process destroy the molds. Because not every attempt to cast a link is a success, I had to make over 200 sets of plaster molds:
1) The master link was used once to produce the top and bottom of a female polyester mold. I waxed the master link and sprayed it with silicone release agent before pouring the polyester. The polyester I used was from Tap Plastics. It worked, but set up very quickly so I had to repair air bubbles in the polyester mold. To ensure correct alignment of all subsequent molds, 1/4" holes were drilled thru the polyester mold before the master link was demolded.
2) Once the master link was demolded, 1/4" pins were placed in the polyester mold halves and four sets of male rtv molds were created. Once again, wax and silicone spray were used as the release agents.
3) The four sets of rtv molds allowed me to make four links worth of plaster molds at a time. When making top the mold halves (which would shape the bottom half of the link since I poured the links upside-down) I used a plexiglass plate and tube, and an aluminum pin to form a funnel to direct aluminum into the mold. I found from experience that a 3" high sprue works a lot better than a short sprue, since it allows more hydrostatic pressure to force aluminum into all parts of the mold.
 |
At left is the mold components to make the top and bottom
plaster molds for a single track link. The RTV mold furthest left is a positive of the bottom of a track link. Above it are a plexiglass plate with a plexiglass tube, along with a pin that is used to form the "funnel" into which aluminum can be poured. The RTV mold to the right is a positive of the top of a track link. The plastic mold box above it is used to contain plaster over the RTV. |
 |
Plaster has just been poured. The plexiglass tops are
clamped in place to keep plaster from leaking. The plaster is actually a 50%-50% mix of plaster of paris and talc. Supposedly the talc improves the heat resistance of the mold. TheApril 29, 2006r of paris. It is not mandatory to spray the RTV with silicone release agent, but the RTV molds last a lot longer if you do. I needed to use each mold 50 times, so I used the spray. You always have to use the spray on the plexiglass and aluminum mold parts. The plaster is only workable for 5-10 minutes so you have to work quickly. I didn't have a vacuum chamber to de-air the plaster so I had to tap on the molds to get air bubbles to float to the top. |
 |
The plaster can be de-molded in around 45 minutes. Several plaster mold halves after demolding from the RTV positive molds. Note the holes for 1/4" alignment pins. Also note that the mold halves have impressions of the 1/16" pins that will be used to form holes in the links for link pins. |
After the plaster has set, the molds must be baked to remove excess water. This is because when 1300 degree F aluminum hits a damp mold the instantaneous formation of steam can make the mold explode. I baked the mold halves at 200 degrees F for two hours, and then assembled the mold halves with 1/4" alignment pins and 1/16" graphite-coated pins to form holes for track link pins. After the molds were assembled, I wrapped them in foil and returned them to the oven for a minimum of one hour at 500 degrees F. I kept the molds in the oven until a minute or two before the aluminum pour.
Note that graphite sticks to used wire better than to new wire. For new wire pins, either apply a minute amount of light oil (WD-40) to the pins first, by rolling the pins on a lightly oiled paper towel. An alternative to graphite is to cover the pins in soot from a candle (sooty citronella candles are best).
 |
This photo was taken immediately after pouring the aluminum.
The assembled molds were placed in a sand-filled box. Small rocks were placed
on top of the molds to keep the hydrostatic pressure of the aluminum from making the top
half of the mold float. Aluminum must be poured all at once, don't dribble it in. It is also very finicky about pour temperature. I found only a 40 degree F range of temperatures that worked. I used a high temp thermocouple to measure the temperature of the aluminum before pouring. The first several attempts to pour links were a frustrating trial and error process. I think I know what I'm doing now. |
 |
Once the molds are cool enough to be touched, I open the molds and then break off the remaining plaster. |
Several time consuming steps are still required to finish the metal portion of each link:
1) Remove the 1/16" pins by grabbing them with vise-grips and twisting them out. Most of them came out fairly easily and were un-damaged so they could be re-used. Around 1/10 of them were stubborn and required significant application of oil, vise-grip wrangling and cursing to get out.
2) Saw off the sprue within 1/8" or so of the link. I didnt try to cut closer than that because I didnt want to scratch up the links with the saw blade.
3) Use fine pliers to break off large pieces of flash. I removed the rest of the flash and other minor defects with a 1" wide belt sander and a Dremel tool with carbide grinder bit.
4) Mill off the remainder of the sprue. I used a 1/4" end mill for speed.
5) Clean up the sprocket holes with a 1/8" end mill. Rather than trying to precisely measure, I just did this by eye.
The final step is the neoprene traction pad. I used Flexane 80 castable neoprene, with the metal primer and the intermediate primer. The use of both primers is recommended for greater bonding strenth and/or immersion service. In order to cast the neoprene in place, I made a template of the traction pad out of aluminum and then used that template to form modelling clay into the mold.
 |
I molded 13 links worth of traction pads at a time. I didnt pick this number for luck, it just happened to be the greatest number of links I could cast before the Flexane started to lose its workability. |
 |
Close-up of a few molded links. The blueish color of the metal is
the blue FL-10 metal primer. Note that the surface has been roughened up for better
adhesion (I used a miniature Dremel-mounted cutoff wheel). Note: Before I formed the modelling clay molds, I cleaned the links carefully with acetone to make sure I didnt have any oil or grease present. |
 |
This photo is taken immediately after pouring the neoprene. I stayed
around for the next 5 minutes or so with a pin so I could pop any air bubbles. It is possible to de-mold the links after the neoprene sets overnight. Full strength is achieved in a few days. The finished Flexane 80 is about as hard and tough as a truck tire. Harder than an ordinary automobile tire. |
 |
Flexane kit instructions advise that you use the entire kit at once, but
it is not practical to pour 168 links all at once. And a Flexane kit costs around
$30. Fortunately, if you can weigh out the resin and hardener within 2% accuracy,
you can mix small batches. The cheapest accurate balance scales that I am aware of are used for weighing gunpowder and are available at most gun stores. I modified the scale a bit so it would fit a 2 oz plastic mixing cup. It is impossible to get a precise amount of resin because it is as thick as mollasses. For 13 links, I add about 200 grains (unit for weighing gunpowder) of the resin, and then accurately weigh the resin. I then divide that weight by 0.77 to get the total weight. I dial the scale to that weight and drip in the hardener till it balances. |
The final assembly of the links required drilling the rear pin holes out to 5/64" diameter and the front pin holes to 5/64" + 0.003". At the front of each link, I tapped three turns of 3-48 threads on each side. I then cut pins from 5/64" stainless steel wire, assembled the links with the wire pins, and installed 3/32" long 3-48 grub screws to retain the pins.
CONCLUSION:
I am nearly done making tracks, but this task was much bigger than I originally anticipated, taking around one hour per link. It was a good experience, and I leaned a lot about metal casting, but if these tracks ever wear out, I think I will pay someone to professionally cast the replacements.