--- /dev/null
+# Test 3 #
+
+First attempt using Erythritol. Simulation suggested that putting 5
+long grains in a 38/720 case would yield a reasonable motor.
+
+Formula is a simple mix of 65% KNO3 and 35% Erythritol. Both used in
+granular form directly from stock. Because Scott Fintel's site
+indicated near infinite pot life in the multi-cooker, We decided to
+melt enough for 8 grains, or a 520 gram batch size.
+
+Before starting this batch, we cut our two delrin mandrels in half to
+get 4 single grain mandrels.
+
+[Tru-Core](http://www.rocketsaway.com/html/casting_sets.html)
+casting set with four bases each set up with 0.5 inch mandrels. Four
+pieces of casting tube cut to grain length of 2.250 inches.
+
+## Process ##
+
+The multi-cooker was preheated to about 250F with the lid on. Temp was
+increased during melting to 300F indicated.
+
+We poured/scooped one grain at a time, leaving the multi-cooker heating
+and putting the lid back on between grains to keep the propellant at
+temperature. The propellant was really hard to pour into the 38mm grains
+with mandrel already in place. Lots of drips over the side. We reduced
+heat to about 225F while waiting for the first 4 grains to solidify. It
+was easy to pull the coring rods after an hour or so. The caps were a
+bit hard to remove.
+
+Given how much trouble we had pouring into grains with the mandrel already
+in place, we paused to fabricate some new casting bases using the CNC
+router and left-over hickory flooring material. The resulting bases are
+used with aluminum foil to hold a casting tube each. The casting tubes
+are marked with a fill line on the inside ahead of time, then the mandrels
+are pushed into place after the propellant is poured, penetrating the foil
+bottom. A cap is then placed over the mandrel and casting tube to keep the
+mandrel centered. This worked a lot better! We called this test 3B.
+
+## Results ##
+
+* Test 3, Grain 1, 61.78g net
+* Test 3, Grain 2, 61.54g net
+* Test 3, Grain 3, 61.84g net
+* Test 3, Grain 4, 61.89g net
+
+* Test 3B, Grain 5, 60.36g net
+* Test 3B, Grain 6, 61.23g net
+* Test 3B, Grain 7, 60.40g net
+* Test 3B, Grain 8, 60.79g net
+
+## Observations ##
+
+The new casting bases worked very well, but we did notice a problem with
+the caps. When the propellant is carefully controlled to completely fill
+a grain, it touches the bottom side of the flat caps when still hot and
+wet. When it cools and contracts, this leaves a thin skin above a ring
+of mostly air. Cutting the "skin" off the top and scraping a bit to leave
+each grain with a slightly concave top so there'd be burn space between
+grains left us with remarkably consistent looking and weighing grains. In
+future, we might want to either fill the casting tubes slightly less, or
+perhaps switch to caps with an explicit air space above the poured
+propellant in hopes that all bubbles would "surface" leaving the propellant
+solid and with a bit of air gap at the top of each grain's casting tube.
+
projects / web / gag.com / blobdiff