I'm getting a result that I don't understand. Curious if anyone has any insights. Because of downstream CNC processes, it is important that my castings be as dimensionally accurate as possible. I decided to do a calibration test. I put together this CAD model: 3D printed it, converted it into wax models using my standard RTV rubber process, and then cast two copies in bronze. The results surprised me. Shrinkage in the horizontal dimension was 4.19%. Shrinkage in the vertical dimension was 2.06% The average of just over 3% shrinkage seems about right, but I'm confused how I could have over twice the shrinkage in one direction than the other. The castings were made using solid block method, with the parts oriented inside the block just like in the pic above. Any thoughts?
https://maritime.org/doc/foundry/index.htm here is a link to the Navy foundry Manual, read page 1 chapter one " how metal solidifies ", then read it again and again, and it only gives the most basic answers, but your answer is there, basically thicker parts shrink more then thinner parts, but there are many many variables , I assume you are doing some sort of shell casting, that and die casting have way way different results then sand casting as the "mold' can be preheated and made to feed , much longer V/r HT1 P.S. repeat your test, and switch the Horizontal and vertical and see your results
What wax are you using? I suspect up to half the overall shrinkage and most of the uneven shrinkage you're observing is from the wax. Did you measure your wax before you invested and cast it? How was the part gated and positioned for the pour? Best, Kelly
The wax is "Limerick" from R&R. I should have thought to take some pics, but here's what the sprues look like: I attach the four tabs on the wax mold to the marked places on the sprues. I didn't think to measure the wax before doing the pour. I've still got the RTV mold, and will make another wax copy later and have a look. Cheers! Jeff Here's the update: Got the wax mold measured up. Surprised (and pleased) to say that it is pretty much unchanged from the 3D model. The measured variances are on the order of 0.05%, which is probably just imprecise measuring on my part.
