Help in pattern separation from the sand

Discussion in 'Sand Casting' started by my_account_todoist, Apr 15, 2021.

  1. Hi! I'm dealing with quite a gnarly pattern - really it is just a plate which has been hollowed out and had webbing added. The webs have plenty of draft, but are quite thin and the "cores" between them end up being pretty small. To get past the obvious: yeah, redesigning it would be best. But this is no longer about the functional object for me, not for now. This is a challenge :).

    The base issue I have been facing is the pattern dragging the sand pieces between the webs up with it. A primitive solution is to try and tap them out so they drop on the pattern and then use them as little cores of sorts, but really, this is very non-ideal.
    Moving from an FDM print to an SLA print helped, and sanding that SLA print down to 2000 grit has gotten me my first bit of success.

    However, I'm really wondering if there's anything one could add to the pattern to help "lube it up" a bit, and thereby aid in pulling it out. I'm very close - all I need now is just a little boost in separability from somewhere so I can push this over to the finish line.
  2. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    It'd be easier to make some suggestions if you posted a picture of the pattern and/or mold. Given your reference to 2000 grit, I'll assume your pattern has a nice smooth and polished surface.

    Talc, Paste Wax, Graphite. What are you using for parting compound?

    Imbedding small wire chaplets in the fragile sections is a common technique for stabilizing the offending portions of the mold.

  3. On the pattern itself, I'm using nothing. Otherwise I'm using just some calcium carbonate which seems to have the effect of gluing one layer of sand to itself very tightly. This does allow it to establish a parting line of... sorts... but I must say I'm not 100% confident it is working as intended.
    Why didn't I think of adding some wire reinforcements. You are a genius.
  4. crazybillybob

    crazybillybob Silver Banner Member

    paste wax that thing up. Once it's hard and buffed add talc or Graphite.
  5. crazybillybob

    crazybillybob Silver Banner Member

    Graphite will last a few pulls before you must add more. Talc is one and done.
  6. Petee716

    Petee716 Gold Banner Member

    Along with the above suggestions, go easier on the ramming if possible.

  7. Melterskelter

    Melterskelter Gold Banner Member

    Pictures? Wax, Graphite, talc. Plenty of tapping---just enough, not too much.

  8. Melterskelter

    Melterskelter Gold Banner Member

    Oh, and how are you drawing the pattern out? Are you using draw spikes and tapping it up for the first 1/8" or just pulling it?

  9. DavidF

    DavidF Administrator Staff Member Banner Member

    Let's see it.....
  10. Alright, alright! By popular demand, some photos attached :D

    1oz silver maple for scale - it really is a wee little one. But it does have fairly nice draft and some radiuses just about everywhere, so it shouldn't be a nightmare. Pictured also my most successful cast so far - I've cast every mold that has made some significant progress, since this is my first really "challenging" piece and I thought it might make for a nice keepsake, that row of improving attempts culminating in an actually successful cast. Also because of the way I am set up, casting is very very quick and easy compared to mold making, so its partly a case of "might as well". Good job too, because along the way I have discovered the importance of venting those little webs, among other things.

    For pulling it out, I am screwing a woodscrew into the hole that I cut threads into using that same screw. Then I tap it with a very light hammer - so far it has seemed to be the best - though I have tried a lot of objects for tapping.
    Adding copper barbs worked once perfectly!! Too bad I then messed up the gating it seems, and got no cast out of it. Tried it a few times after that and never saw quite the same success..... might be that I was ramming too hard those other times. Today I intend to try going a bit lighter on the ramming as well as perhaps try using a bit of steel wool as reinforcement. I was slightly worried it would contaminate my sand, but my experiments with the barbs suggest it is actually pretty easy to clean them out by just remembering to deal with the reinforced area of the sand separately. 20210416_112519.jpg 20210416_112505.jpg 20210416_112527.jpg 20210415_202435.jpg
  11. Looking at how olfoundryman uses parting powder, I think I am now picking up that he does in fact want it on the pattern - just in moderation, with the excess brushed off. Something to try, I think. So far I've tried to keep it off the pattern altogether, figuring its purpose was exclusively to keep sand from sticking to itself.
  12. Tobho Mott

    Tobho Mott Gold Banner Member

    Give it a try, those shallow greensand cores ought to mold up fine and not break off if they have draft, likely without needing those wires. I like to sprinkle lots of dust on new patterns then rub it into the surface of the pattern by hand before brushing off the excess. You definitely don't want any clumps of parting dust on the pattern but a fine film can be helpful IMO.

  13. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    The points in the acute angles of the triangular reliefs will always be very weak and hard to pull. The slightest error extracting that small pattern will damage the points and/or dislodge that portion of the mold. Not sure how you managed 2000grit finish on those included angles but sometimes a dipped coating of finish will do the trick with no handwork. If you haven't been using any parting compound on the pattern that's the first thing to correct. Adding a radius to the points of the triangles would help. What kind of sand/binder are you using? Good green strength? Next will be getting the mold to fill. Is that aluminum your pouring? Pour temp?

  14. A good bit of time sitting outside with sandpaper and a tiny little metal spatula of sorts that could get right in there. SLA resin sands like a dream, so that slightly offset these features being a PIA to sand.
    Visual perfection isn't necessarily needed here, a bit of extra material is okay. The purpose of these reliefs is to reduce material consumed while maintaining stiffness. However, if I reprint this guy ever again, I'll be sure to radius all the corners there much more generously. Not sure why I didn't to begin with - I guess this is the kind of thing you pick up with experience.

    Using a greensand - not sure how to describe its strength, having little reference point. It's a natural sand, from my parent's yard; the area has sand on top and plain clay further down. In between there's a bunch of clay-infused sand that as a kid I used to dig tunnels in - shudder to think how sketchy that really was, but they held up. I was at the time very impressed by how this earth material could effectively be carved without it collapsing, so all these years later when I needed sand for this I was reminded of it. I dug out a whole bunch and painstakingly sifted it down to 50 grit (0.35mm), got about 250 litres dry volume. Probably ten hours of sifting alone.... man, if this thing doesn't actually work for this application, it'll take me a couple of days to get past the denial :D

    I'm pouring Zinc (ZL12), which is very fluid and should be good for these kinds of thin sections. That's the hope, anyhow! However, I'll worry more about filling the mold once I have one :D
    Headed out to try applying parting compound onto the pattern, per Tobho's suggested method.
  15. rocco

    rocco Silver

    One thing jumps out at me from your pics, you molded it with the webbing facing up in the mold, try flipping the pattern so that the flat surface is facing up in the mold. Let gravity work with you instead of against you.
    Last edited: Apr 16, 2021
    3Dcasting likes this.
  16. I've done it this way too for the very first attempts. I flipped it around so I could feed the metal in from the bottom and provide better venting opportunity. So far I have not had issues with the mold breaking after I have managed to form it, certainly not due to gravity related issues. And in the one pictured, I had good success getting the metal to rise up to the top of the pattern everywhere except that one corner and in the middle.... why not there, not sure. Maybe messed up venting, maybe my sprue solidified on me and there simply was no more metal to go around. Don't fully understand what happened.
  17. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Should be very doable in Zamak and with decent sand and parting compound, shouldn't need chaplets. I agree casting webbing down will likely fill better. Also, sand is considerably lighter than zinc so any loose mold media should float. Get the pour temp right and you'll get there.

  18. Melterskelter

    Melterskelter Gold Banner Member

    Some thoughts:

    Looking at the photos of your 2nd from the top photo I see a lot of point highlights in the angle between the base and the vertical fins. That suggests there is some roughness in those angles and most likely the verticals as well. That roughness, if present, will often inhibit sand release. Painting your pattern with glossy white spray lacquer would help you see bumps which, unless painted, often escape notice.

    One effective way to smoothly fillet those inside corners is to simply mix up a few cc’s of laminating epoxy resin and syringe a tiny bead resin into the corner while the pattern is resting at an incline that makes the various intersecting vertical faces be at about 45 degrees left and right and with pattern leaning away from you maybe 5 to ten degrees for the vertical fillets. For the intersection of a vertical face and the back the pattern will just need to lean away from you 45 deg. The resin will settle into the corner making a perfect fillet and will set up like that. Bubbles may be seen in the fillet just after you inject the resin. A very quick puff of flame from a propane torch will pop them. You can do the fillets in sets where you resin all the similarly-oriented fillets in a batch. If I figure it right this pattern would require 8 sessions. I’ve done a lot of fussy patterns this way. It is much quicker and less tiring than using Bondo and sand paper.

    When you do attempt to draw the pattern, having some sort of crossbar on the spike to allow upward tapping on the spike is more effective than lifting while tapping down on the pattern. Up-tapping on the spike takes advantage of the sand’s inertia to stay in place. It’s a subtle difference, but matters. Here are some spikes I made and use often. I thumb-finger gently tug on the washer while up-tapping on the crossbar.


    Use a very soft 1/2” diameter (about) water-color brush to help apply talc and graphite to the pattern. You have to get powder into every nook and cranny. Gently blow off the excess with a bulb-syringe.

  19. Tobho Mott

    Tobho Mott Gold Banner Member

    +1. I think it might fill easier that way as well.

    I didn't notice how sharp those inside corners were when I commented earlier. You could add some fillets there using ie. Bondo.

    Not sure what effect your natural molding sand might be having. Hopefully you lucked out and it's just right, but it's possible for too much or too little clay to have detrimental effects.

  20. Denis, those things look fantastic! I've got to build something similar. Just need to figure out the how, what with my lack of a welder. I suppose even using those little loop things with a thread at the bottom could give a fair bit of the benefit!

    Using the calcium carbonate on the pattern, I managed a pull-out roughly equal in quality to some of the ones with the barbs - and far better than anything I had accomplished without them.

    Looking at where the pullout failed, I think I'll go back to CAD and add some more generousness to the fillets and desharpify everything per your suggestions here. This does mean I might be waiting until tomorrow to do my next attempt, but it's 95% passive effort. I might get away without doing any, or at least so much, sanding - SLA layer heights are 0.05mm, and therefore the grabby gaps are less of a thing and might be reasonably filled in with the chalk. If not, then I'll go do some more sanding.


Share This Page