I've been interested in lost PLA as I have some parts I'd like to make that would be easiest done casting. The trouble is that the casting isn't simple due to the inclusions, underhangs, etc.
So, I set out to slay the dragon and tonight I did my first pour, but the journey started about 2 weeks ago with revamping my electric furnace/kiln to adapt it for doing PLA burnouts (separate post on that).
The part I chose to make was a timing belt pulley for my CNC mill. The Y-axis didn't have a tensioner on it, and the belt was crazy tight, had to basically cut it off else worry about mushrooming the shafts.
The timing belt used as an HTD 8M profile, and I went for a 28tooth pulley w/14mm keyed shaft. I found some plans on line and borrowed an STL and mocked up a pulley. I added a 3.5 inch sprue, a gate and a riser. I modeled everything exact dimensions, with the intent of using Repetier's scaling feature to add the 3% expected in shrinkage. This was ok until I read up on the fact that the over casting will shrink, but the keyway will actually expand as the metal shrinks 'away' from the hole. I took the 14mm diameter and reduced it by 3% in Fusion, (all other dimensions left the same.
I used a tin can from some tomatoes as my container, and whipped up some standard DAP plaster of paris 50/50 with some sand blasting aggregate I had (came from industrial site that recycled their sand -- lots of metal flakes/paint in it). I setup the mold last night and after 10 hours of sitting, into the new heat-treat furnace it went (upside down, with sprues pointing down).
I baked with this recipe:
200 celcius for 6 hours
330 celcuius for 3 hours
500 celcius for 2 hours
Baking went well, in the 330 degree phase the pla was burning up fast (no forced air, just radiant heat), and there was actually fire for a bit as the PLA dripped down and started going. Overall, my biggest fear was the horror stories about PLA not burning out completely. From what I could see, it came out pretty good, good enough to showcase the keyway/shaft bore.
Since the furnace is already at temp, in went the steel crucible, some ingots and I went and mowed the grass. I set the furnace for 500C and when I came in the PID read 500, the aluminum was melted (crazy how much heat is trapped in the refractory. I added another ingot, waited 20 minutes, set it to 600C and it was ready.
For the pour, I put on the full face shield, gloves, and took all precautions -- didn't know if water content was still a factor. I normally use oil bonded sand and never vent my moulds, and I didn't vent this one either. The pour was uneventful. No smoke, no steam, no popping, no rolling. Nothing. Cool!
I cut of the can after 25 minutes and the result was not perfect, but damn, I'm hooked. If this is the result for a first melt, there is so much upside.
I put the models side by side with the actual. I cut the sprue wells off to see the keyway. Man, did that come out sweet. It is undersized on the shaft diameter (shrink factor was as much as expected), but this better than too big. Overall, this will be VERY viable for other such items. The ability to cast a very defined keyway opens up a lot of possibilities
The only part that was disappointing was the flange on the top part of the pulley. This would have been inverted in the furnace during the burnout and I likely had a little residue in there. Next time (oh, and there will be next time), I'll turn the can over and bake it for 20 minutes to let gravity help me.
Absolutely no complaints on the plaster of paris recipe. awesome. the sand provides some texture and there was zero cracking during the baking/burnout.
So, I set out to slay the dragon and tonight I did my first pour, but the journey started about 2 weeks ago with revamping my electric furnace/kiln to adapt it for doing PLA burnouts (separate post on that).
The part I chose to make was a timing belt pulley for my CNC mill. The Y-axis didn't have a tensioner on it, and the belt was crazy tight, had to basically cut it off else worry about mushrooming the shafts.
The timing belt used as an HTD 8M profile, and I went for a 28tooth pulley w/14mm keyed shaft. I found some plans on line and borrowed an STL and mocked up a pulley. I added a 3.5 inch sprue, a gate and a riser. I modeled everything exact dimensions, with the intent of using Repetier's scaling feature to add the 3% expected in shrinkage. This was ok until I read up on the fact that the over casting will shrink, but the keyway will actually expand as the metal shrinks 'away' from the hole. I took the 14mm diameter and reduced it by 3% in Fusion, (all other dimensions left the same.
I used a tin can from some tomatoes as my container, and whipped up some standard DAP plaster of paris 50/50 with some sand blasting aggregate I had (came from industrial site that recycled their sand -- lots of metal flakes/paint in it). I setup the mold last night and after 10 hours of sitting, into the new heat-treat furnace it went (upside down, with sprues pointing down).
I baked with this recipe:
200 celcius for 6 hours
330 celcuius for 3 hours
500 celcius for 2 hours
Baking went well, in the 330 degree phase the pla was burning up fast (no forced air, just radiant heat), and there was actually fire for a bit as the PLA dripped down and started going. Overall, my biggest fear was the horror stories about PLA not burning out completely. From what I could see, it came out pretty good, good enough to showcase the keyway/shaft bore.
Since the furnace is already at temp, in went the steel crucible, some ingots and I went and mowed the grass. I set the furnace for 500C and when I came in the PID read 500, the aluminum was melted (crazy how much heat is trapped in the refractory. I added another ingot, waited 20 minutes, set it to 600C and it was ready.
For the pour, I put on the full face shield, gloves, and took all precautions -- didn't know if water content was still a factor. I normally use oil bonded sand and never vent my moulds, and I didn't vent this one either. The pour was uneventful. No smoke, no steam, no popping, no rolling. Nothing. Cool!
I cut of the can after 25 minutes and the result was not perfect, but damn, I'm hooked. If this is the result for a first melt, there is so much upside.
I put the models side by side with the actual. I cut the sprue wells off to see the keyway. Man, did that come out sweet. It is undersized on the shaft diameter (shrink factor was as much as expected), but this better than too big. Overall, this will be VERY viable for other such items. The ability to cast a very defined keyway opens up a lot of possibilities
The only part that was disappointing was the flange on the top part of the pulley. This would have been inverted in the furnace during the burnout and I likely had a little residue in there. Next time (oh, and there will be next time), I'll turn the can over and bake it for 20 minutes to let gravity help me.
Absolutely no complaints on the plaster of paris recipe. awesome. the sand provides some texture and there was zero cracking during the baking/burnout.
