Summary

An aluminum casting of a turbine, using an ABS mold.

My understanding is the original turbine pattern is here. https://www.thingiverse.com/thing:590416, user by Engi-IRL.

I modified a possible remix, from another user, prk_mek. http://www.thingiverse.com/thing:680664

Instructions

A turbine file was considerably scaled down to about 1.25 inches in diameter.
The ABS molds were printed with 0.20mm layers at medium speed. 120 minutes... 22 grams ABS if my notes are correct...

The (Afinia default) support angle was reduced from 30 to just 10 degrees to minimize support at the curve of the blades' tips.
The fins are so thin (about 0.02 inches) that there would be no way to remove support from such skinny and curved crevasses.
And it looks like most of the fins came out in the casting.

Aluminum was poured at 1250F. It was a small amount for a small part, so there wasn't much smoke. The hot rotor did melt through the mold face, but it just didn't have sufficient BTUs to scorch the plastic... Cooled rather quickly.
The semi-finished casting weighs about 7.5 grams.

I made several modifications to this prototype centrifugal casting machine I've been working on, and I searched for something to test it on. This turbine was good looking and appeared challenging. I'd estimate the casting came out 90% good, and 10% defects.
A better gating system (sprue, runners, gates, vents, etc.) should fix most of the 10%.
ABS molds are new territory so there's much trial and error.
This project follows a few prior explorations into what casting metal in a 3D printed mold may have to offer.

http://www.thingiverse.com/thing:735587 aluminum heat sink
http://www.thingiverse.com/thing:733205 gravity cast tube

Aluminum turbine rotor cast in ABS mold