On making a long thin core using sodium silicate

Discussion in 'Sand Casting' started by Melterskelter, Jun 13, 2019.

  1. Melterskelter

    Melterskelter Silver Banner Member

    I am making a new pattern for a triangular in cross section 18” long cast iron prism for use as a scraping master to be used by precision machine restoration people.

    The angles of the triangle are 45, 60, and 75 degrees and the faces of the prism are roughly 2.25 inches. The target for the wall thickness is 1/4”. We’ll see how this all works out.

    But one essential component to the molding process is a long skinny core. That in itself is not such a big challenge but one concern I had is making a vent to run the length of the core as I understand such a vent is a necessity. Hmmmm, how to do that. I considered a bunch of options and decided to try (expecting a good chance of failure) using a piece of 1/4” O1 tool steel which is cheap and supplied fairly well ground and polished. I was pretty sure that using the rod bare with a release agent would result in a tightly stuck rod simply because the numerous tiny imperfections in the rod surface would provide “hand holds” for the core sand. Since the cured core is quite hard, I figured there would be no way to pull the rod once the sand had set and the sand has zero green strength. So I waxed the rod and wrapped it in wax paper (figuring the paper was soft, non-adherent and cushioning) and slipped it into the core box and then packed in the sand last evening. Rather than attempt to gas such a long thin core, I used propane carbonate as a catalyst to cure the silicate.

    With much trepidation, after packing the box last night, I gave the rod a test tug and much to my surprise it slipped right out! I am going to let the sand set up a while longer with the box in the sun for heating before opening it. I hope/think the core will be strong enough to span the 18” in the green sand mold I will pack today or tomorrow.

    103B53A9-BFD8-493F-B1E2-F6EE318F99C7.jpeg DDC7F788-C59E-4061-8859-8162B10ACCC7.jpeg 881B36B8-F872-406B-8949-876A602A9E30.jpeg

    The corebox is cobbled-together MDF.

    Whether this pattern and core setup provides a decent casting or not, this method is an option for venting a long thin core.

    Other methods could include using a heavy cotton cord stretched along the central axis of the core with the plan being to simply leave it in place.

    Tobho Mott likes this.
  2. myfordboy

    myfordboy Copper

    I would not bother with that vent. Better to put a steel rod through to give the core some handling strength.
    The core only needs venting at the end to the surface of the mould. Quite often I have forgotten to do this without it making any difference.
    This comment applies to SS/ CO2 cores. A baked one definitely needs the vent.
  3. Melterskelter

    Melterskelter Silver Banner Member

    Now you tell me! ;-). You might laugh at how much time I spent figuring just how to make a vent as I thought such a long narrow core would surely need one.

    Good to know if it does not need a vent in the core itself as packing the core will be a lot easier without the rod. Next one I will skip the rod if the core proves strong enough without the rod. I intend to have a void in the greensand at each end of the core to make insertion of the core easier.

  4. HT1

    HT1 Silver

    I probably should stay out, but I gotta ask a couple of questions out of curiosity :
    1) why the Triangular core if you are Making a triangle scraper , why not just a round core to keep your Cross sectional area and the thickness Minimized. the Triangle core will make for seriously hot spots and the potential for tearing /cracking in the parts.
    2) have you considered casting the part standing up, this will really help with so many problems, like flex in the center of the core, feeding... add a couple of inches to the Part, and you have a built in riser ... long parts really cast well standing up

    Lets see some finished Product

    Best of luck

    V/r HT1
  5. Melterskelter

    Melterskelter Silver Banner Member

    I could surely do a round core, but the main idea here is weight saving and aesthetics. A round core diminishes weight saving and just does not look as good to my eye. That is surely just a matter of my personal taste. I am copying an old design and figure if it was done before, it should be possible. Let’s try...

    I have considered vertical casting. I have flasks all ready to go from another pattern for horizontal casting and am inclined to give it a go that way. If that does not work, then go for vertical. The triangular core shape should be less saggy than a round one I would think. It will be oriented with the pointy 45 deg angle pointed down.

    You’ll see pics of the results just as soon as I can get there.

  6. Melterskelter

    Melterskelter Silver Banner Member

    Well, I opened my corebox tonight. The results were “instructive” but that is far different than “optimal.”


    That mess photographed is the rather sad result of a rather casual pass at making a core box. But not all effort is wasted as there are some observations that I find useful.

    1) you can see that the sand/silicate stuck selectively in areas that were not as slick as the well separated areas. The scrap MDF I used had a very slick black plastic-like surface on one side and a raw surface on the other side. You can see that I cut the box so two of the three surfaces were the slick finish and one side was the raw surface hit with a couple coats of sprayed Bullseye brand shellac. You can also see I glued in half round and on the ends somewhat flat bars intending to make recesses in the core so that ribs would form in the prism. Those half rounds and flats were sprayed with a couple coats of lacquer after being glued into place and were then half-heartedly sanded and final coated. Not much sanding at all was done on the adjacent over-sprayed areas that extend maybe three quarter inch on either side of the bands.. Then the whole works was sprayed with silicone spray.

    Lesson demonstrated: The value of a slick finish is well shown. The plain black ares separated cleanly. The white and raw areas stuck pretty tightly. So, for future reference I need to not be casual about finish and should sand things quite smooth. Secondly, I am pretty sure a liberal application of paste wax would have prevented sticking. I did not do that to learn if spray silicon was adequate. It looks like it is not. I need to try graphite.

    2) I did not mix the sand, silicate, propanol carbonate as well as usual as I usually do. I thought I could get away with hand mixing the small 1 pound quantity of sand. If I had been diligent, I am sure I could have mixed well by hand. Buuuut, I was satisfied with pretty good. The end result was varying strength of the sand. Some areas were really strong. Others were soft enough to get a fingernail to dent them. Not good enough.

    Summary: poor technique coupled with poor technique results in poor core. Surprise! Not really. I’ll try again tomorrow. This time, no rod since Myfordboy tells me I don’t need a vent. I really think well- mixed sand and good packing will provide adequate strength ( or not). We’ll find out.

    Tobho Mott likes this.
  7. DavidF

    DavidF Administrator Staff Member Banner Member

    I just have to ask.... why ramming it on the end instead of just having a v slot and ram it on the horizontal and striking off the top??
  8. Melterskelter

    Melterskelter Silver Banner Member

    You get to ask. Hard to make the ribs if I strike off the top. End ramming will be easy without the 1/4" bar. I will probably get the sand rammed tighter that way than striking open vee as well. Without the bars open vee would be OK. Do I need the bars? Weeeellll, I WANT the bars. ;)

  9. Melterskelter

    Melterskelter Silver Banner Member

    There may be a bit more to the sticking sand problem:

    I got to thinking about the areas to which the sand stuck as outlined above. A factor that I did not initially consider is that the packed corebox was baked at 220F in the oven for a few hours to ensure completion of the silicate cure process. That temperature likely resulted in softening of the sprayed areas but not hard slick black areas and so caused the sand to embed in the paint and shellac. Adding to that theory is that I popped off the shellacked side prior to baking and noted only slight sticking near one or two of the ribs but no where near the amount seen post-baking.

    In the past I have often finished silicate cores by baking them just for good measure and to be certain moisture was minimized in them. But I think in all cases the core had been stripped from painted boxes or the boxes were metal tubes not subject to heat softening and sticking.

    So, maybe I attributed to much of the difficulty to poor surface prep when softening of the paint and shellac were significant factors.

    So, going forward on this core I will not be baking it in the box. I will continue to use propanyl carbonate as a catalyst and pay closer attention to thorough mixing using a drill-driven paddle. I may drill gassing ports in the box since trying to gas it thoroughly from the ends is impractical I think. I will be using Johnson’s paste wax on the box to improve stripping. I will pay close attention to good surface prep.

  10. oldironfarmer

    oldironfarmer Silver Banner Member

    Very interesting thread. Much more information than if the first one was perfect.

    A couple of observations:

    I think you exceeded the working limit of shellac at 220F.

    I'm going to be surprised if the core is hard enough to support it's own weight at 18" long. I hope so and need to copy you if so.

    If you follow David's advice can you press ribs into the struck surface after striking?

    How much does SS sand need to be rammed to make a solid core? I'm struggling with that, I think I have made some real strong cores with light ramming but not sure if all else was equal.
  11. Melterskelter

    Melterskelter Silver Banner Member

    I think 220 is definitely too hot for shellac and similarly too hot for the lacquer. Lesson learned.

    I am not certain how tight is tight enough. I am starting to think firm ramming results in greater structural strength than looser ramming. A well constructed silicate core can be quite strong. The design of this core is somewhat segmented which promotes breaking however so careful handling will be needed.

    It would not be hard to test this by making a box to form maybe

    Yes, I could carve in bars in an open vee. I may move to that, but intuition moves me in the direction of a closed box. But I definitely will not experiment with melting sand into shellac or lacquer. ;-). DOH!


Share This Page