Bifilm Centrifugal Separator

I've seen the centrifugal separator in the runner system in some of the bifilm literature. Figured I'd try one out. As with most everything else I made it too big, but it was a start.

I turned a bell shape.



Then a disc and a tangential runner. Runner is 0.4" square, a lot larger than I want not that I've done the glass flask video. Here they are stacked up. Parting line is under the bell on top of the disc.



The disc is intended to be bound sand. It is 3" diameter and 1/4" thick. I want to use this to separate heavy parts at the lip of the bell and light bubbles at the top of the bell. My plan is to have a bottom exit so the bound sand disc is to seal the top of the gate, buried in the drag. A smaller bell of bound sand goes on the disc, inside the bell with a hole down the center to let metal out.



Then the gate, picture is with everything upside down.





Made a plaster core box for the disc and internal bell, then made a sand and sodium silicate core.

 

Proceeded to make a casting. Drag



Broke the core, core box needs some work.



But it will work. It's main function is to separate the separator chamber (big bell) from the gate so all flow out has to swirl around and rise in the chamber before exiting down. That also gets the gate 1/4" into the drag to assist in bottom feeding.

After ramming the cope and pulling the patterns the core is in the drag.



My gate was too short, the cut down parting line was too narrow and broke off.

Here's the cope ready to close.



It cast with a little shrinkage on one side of the pattern. I wanted to use the separator mass as a riser but think the 1/2"' center tube is too small to properly feed the pattern. It needs a separate riser.



That's a sand defect from closing the mold with loose sand trapped, on the pattern top. Sloppy molding work.

Another view



It is easy to see how the disc seals off the gate from the bell. You can see the shrinkage on the side of the pattern.

 

Section through the bell.



The square on the right is the last bit of the tangential runner. Once the metal flows into the bell it will swirl around the bell then begin rising. When it gets to the outlet it will start feeding the outlet tube but hopefully keep rising and fill the tube. I wanted slower flow in the gate but am now thinking the vertical tube is too large and the bell may have not filled until the pattern too a lot of metal.

The end of the pattern is pretty clean. Up is to the top, gate entered low, and I only milled the right two thirds. New cutter conventional cut.



Two large porosity points, undoubtedly lots of small ones. This is remelt with presumably lots of mixed in bifilm. I was real pleased by how clean this was.

Here's the top nozzle. Not cut deep enough to clean it up fully, but porosity shows up here.



These cuts were made with the side of an end mill. I'll use a fly cutter next time.

In summary I think the entire separator needs to be much smaller, with a smaller outlet tube to insure the outlet is flooded fast, like a sprue should be. Then a separate riser to feed the pattern.

It seems like a lot of trouble, but if it really does clean up the flow it would be worth it. A separator for a 0.2" runner would be a lot smaller, and with a separate riser it does not need to be very tall.

As always, comments and suggestions are solicited.

 

I missed the video but have seen them in diagrams, that was supposed to be the idea. The internal to try to get rid of entrained air added a lot of complexity and I'm rethinking that.

Really need to get set up to do specimen polishing to really see porosity but at least there aren't gaping holes.

 

I'm interested in your method for polishing specimens. I'd like to clean mine up a little better, but the file still leaves marks. Do all the powered methods smear? Sanding and buffing?

 

I think speed smears. 50 years ago in college we polished specimens for analysis. I remember a slow turntable with polishing compound and maybe water on it. The specimens were buried in Bakelite to hold them but they were small, like broken bolts. I can't remember much (it has been awhile and I've done some other things since then) but I suggest hand sanding on sandpaper on glass progressing finer grits until you can go to polishing compound on glass. My plan is to make or buy a slow speed polishing wheel. I also need a microscope and need to know how small a hole is still porosity.

 

For hobby microscope, you could do hand polishing without the turntable. Start with large grit sandpaper on a flat surface, pour a few drops of water and start up and down motion in the direction perpendicular to the marks of the cutting wheel. When the marks disappear you go to finer sandpaper and polish perpendicular to the marks made by previous sandpaper. That is how you can be certain that you have reduced roughness and will not have deep marks. I usually start at 80 or 100 grit paper and finish with 1000 grit (or 2000grit sometimes). For final polishing, you can use diamond paste and velvet cloth that you can buy at curtain shop /better than on glass/.
Steel specimens are easy since it is hard. Aluminum tends to smear, but it is doable and with some practice you will be able to see the microstructure
Good luck.

 

Thanks, CLR! That's how I polish by hand, but it's good to have validation. Velvet cloth I did not know about. I do have some diamond paste on the way, it is really cheap. I also have a really cheap microscope on the way, we'll see how it works.

 

Looking forward to seeing this process.. Thank you for sharing!