Surface finish help

Discussion in 'Investment casting Ceramic shell method' started by ka_6772, Apr 11, 2020.

  1. ka_6772

    ka_6772 Copper

    I've painted my hot rod and have been trying to get this casting worked out so I can cast the intake for it. That will be a different thread. It's a bit more than this LS one. Pic below. There will be a plenum with plexi sides and top over the bell mouths. Thanks to you guys, the casting is working out. Now I need to machine the castings.
     

    Attached Files:

    Tobho Mott likes this.
  2. DavidF

    DavidF Administrator Staff Member Banner Member

    Are you planning on staying with the lost pla casting for the possible production run, or switch to a different process??
     
  3. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Do I see 4 snails on that? Is that plexi plenum blow through? How much boost?

    Best,
    Kelly
     
  4. ka_6772

    ka_6772 Copper

    IF there is demand I may switch to something else if casting + polishing is less labor intensive.
     
  5. ka_6772

    ka_6772 Copper

    Yes four. 20 psi. The most I've had in it is 8. It's fast enough with 8. It makes 520 N/A. I never put any boost in it on the dyno. It's going to get a redesign on the intercooler. I was blowing through one of my billet 4 bbl throttle bodies but no one ever noticed the throttle body.
     
  6. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    20 psi is pretty healthy. What fuel? Besides intercooling, any other means of predetonation control?

    Best,
    Kelly
     
  7. ka_6772

    ka_6772 Copper

    It's 7.8:1. I run 93. I'm going to have water injection. I going to build for 20 psi, but I have too much time in the paint to put 20 psi in it.
     
    DavidF likes this.
  8. DavidF

    DavidF Administrator Staff Member Banner Member

    I ran 25 lbs of boost on my 2.2l
    It wasn't much fun picking the rings and ring lands out of the oil pan after trying to run on junk fuel....
     
  9. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Initial CR is sensible but I don't blame you about not wanting to throw 1.5 ATMs of boost at it willy nilly......1000Hp street engines are common these days. Even 2000Hp on occasion but so are catastrophic failures of same$$$$$........it aint child's play.......having that kin of power on tap is fun although being in close proximity to an engine at that power level grenading under load......NOT FUN!

    Best,
    Kelly
     
  10. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    I noticed a huge difference when I started using them too compared to conical pouring cups.

    Unless it has bosses and complex 3D machining, I'd CNC it without hesitation. But some thoughts for you to consider. When I have awkward pieces like that in lost foam (which that flange would be an easy part), I sometimes dip coat the feed system features as separate parts from the pattern and glue them together afterward. You could do the same thing and adhere them after being initiallly fired.
    1. My feed system description has some foundry jargon, but an offset pouring basin with weir.......fancy jargon simple object that functions superior.
    2. Tapered sprue.......doesn't have to be round. In fact, only has to taper in one dimension and is easier to make and better functioning square than round in cross section to prevent aspiration.
    3. A vent..........it's a coated piece of square foam or soda straw.
    4. A spin trap........it's a small cylinder of foam that the runner enters tangentially at the bottom on the outer diameter.......thats' it....and they really do work to prevent jetting and turbulence where metal is introduced to a hollow mold.
    As a car guy, and someone who developed my casting methods with the primary motivation being automotive parts, and among them, specifically intake manifolds and induction systems, there are reasons I chose lost foam.

    Lost PLA and Lost Foam are both evaporative pattern processes. The big difference in the two processes comes with how much labor it takes to get from pattern to casting. In lost foam you do one dip coat like your initial shell coat and let it dry. You are then ready to mold which literally takes a few minutes with loose sand and vibration. You are then ready to pour. No burn out, no cracked molds. No cleaning the debris from the mold interior. Molten metal management is the same for either process. Demolding for lost foam is literally dumping the flask (ie bucket/barrel) on the ground, throwing the casting in a bucket of water where the steam blows off 90%+ of the coating and a compressed air nozzle removes the rest in a couple minutes. No chipping away shell...nada.....just a finished raw casting in your hands. Virtually no cost of mold materials at all. The sand is literally dirt cheap and even so completely resuable and requires no conditioning other than sifting out chunks and debris as you load it back into the bucket. I've used the same 300lbs for three years and hundreds of castings. I treat it with complete disregard. The only cost to molding is the thin layer of dip coated slurry and that is pennies per part.

    Now, as far as making the pattern, foam isn't going to be quite as capable when it comes to fine detail as printed PLA but if you look at some of my patterns in the lost foam subforum I think you would conclude plenty good enough for automotive parts. If you make the foam patterns with CNC Router, it's fairly analogous to printing. Your runner would have to be done in two halves in foam and glued together. But, you could design the intake so all 8 runners were CNC'd in halves as a single piece and then glue the two halves together for a 1-pc manifold of any shape and save machine time. That entire pattern in foam would weigh a fraction of a pound. The only thing that can't be managed without lifting equipment is a 50lb charge of molten metal. When you make larger castings everything I discussed above becomes amplified. I could cast a V-8 cylinder head or a 4-cylinder engine block with my current set up. I could do a V-8 engine block with a larger flask and crucible. I lack the desire to do either but could do it. You have a CNC Mill. Making a foam pattern would be old hat to you. My CNC router arrives today!

    You're set up to do shell and it's I fine process. I'd use it for smaller intricate net shape parts. I have 5 gallons of shell slurry in my shop now. If you want to see the lost foam processes I've described, in your spare time check out these three videos. The resulting casting is a trinket but the intent was process demonstration. You might also take a look at some of the automotive parts I've made in the lost foam subforum. The process makes a lot of sense for custom car parts.

    Lost Foam Saucer Cup and Spoon

    Just food for thought....there is a method to my madness. -Really looking forward to your project coming together.

    Best,
    Kelly
     
    Tobho Mott likes this.
  11. DavidF

    DavidF Administrator Staff Member Banner Member

    Lost foam might be a good alternative over lost pla for the flange adaptor, but I do question how flat you would be able to get the casting in lost foam. Pure speculation on my behalf as the only lost foam I have done was a piece Kelly sent me to try, which was a holy shit that actually worked moment...
     
  12. ka_6772

    ka_6772 Copper

    N2 membranes are out! I put three in series and ran through a sample loop for a few minutes. Mass spec shows 11% O2.

    I read the article on the 10 rules. That explains a lot of the issues I was having.
     
  13. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Not sure how much better PLA fares as it cures or under heavy shell weights. For almost all of my castings it's been a non-issue.

    The only exception are those intake manifold lids I make. These were 1/4" wall castings approximately 24"L x 12"W x 2"T. Foam patterns with more shape and structure hold their shape well. You just need to keep the sand depth/pressure the same on each side of the pattern as you fill the flask. Sometimes it's helpful to provide support features for larger more fragile patterns to rest on as the weight of dip coat evaporates and dries.

    The original intake manifold lids were sand cast and not a single one of them I've ever had was flat when free standing.....most are warped about 1/4" or more corner to corner and that's after they have been machined! They were sand cast and either potato chipped as they froze, aged, or machined. They were probably just deflected flat by the machining fixture for machining. They are secured by so many fasteners to seal a flat gasket it doesn't make a hill of beans difference in use.

    At least my lost foam versions lay flat when free standing LoL.....even if half of them did need to be tweaked....sometimes just by hand. I've gone as far as using wooden and sheet metal forms/chaplets to rigidly hold a foam pattern in place and even so, as when it freezes it can still warp/deflect even if the pattern was absolutely straight. I stopped even worrying about because it was about a 50/50 crap shoot and it was far easier just to straight the casting and it's pretty common place in commercial foundries too.

    A foam pattern for a typical one piece V8 manifold would be a problem as far as pattern deflection but can always moved during freezing, aging, or heat treating. -Leave'em in the mold.

    Best,
    Kelly
     
  14. ka_6772

    ka_6772 Copper

    I'd like to get back into lost foam. It's been ten years since I played with it. It may be more suitable for making my plenum. I built a 4' cubed CNC out of 80/20 and used Mach3. I made servos out of treadmill motors. The trashcan I'm casting in is fluidized. I used it for the foam. My big machine can mill out some pretty smooth pieces in foamular. Thinner sections will warp, but I could rib them and just machine the ribs off.
     
  15. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Mill, Big Router....now I'm jealous.

    My first casting rig had a fluidizing coil in the bottom. I never used it because I sift the sand between uses and use the sand shower method and vibe to fill and pack every mold....

    How you support and machine away large foam structures does take some thought. For the most part I've found Foamular to be free of residual stresses, but, hotwiring thin planks can make one side shrink and the board curl.

    Did you ever mill big blocks that required laminating sheets? If so what did you use as glue to laminate and did it machine well or load up cutters? I can buy big billets of EPS or higher density XPS (2lb-8lb/ft3 marine dock billets for example) but not 1-1.5lb/lb3 Extruded Polystyrene (XPS) like Foamular 150 & 250. Do you have a source for thick XPS stock?

    Best,
    Kelly
     
    Last edited: Apr 20, 2020
  16. ka_6772

    ka_6772 Copper

    I always used 3m 77 to glue Foamular together. It can gum up cutters. When I used the 80/20 machine, I had a router with 1/8" and 1/4" mills. If I kept a jet of compressed air blowing on it, it wasn't an issue. Otherwise the end mill would get hot and it was all over. My big machine only turns 8k. Melting wasn't an issue. The foam chips set up in "orbit" around the bigger mills. It's fascinating to watch. I would periodically open the door and blast it with air. I would give it a periodic wipe with acetone to remove the glue. It liked to gum up in the flutes IIRC. I got a quote for XPS intended for lost foam. I thought the price was ridiculous at the time. We have plenty of boating around here. Does the higher density work for lost foam?
     
  17. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    You had better luck than me. I tried 3M 77. Had to use it very sparingly or it would attack the foam. Even so, it would instantly load the cutter and then leave an ugly scar everywhere the cutter tracked after it became loaded. -Horrible. It will instantly load a sanding belt or spindle and will not sand evenly across a joint. I tried 3M 78 which is specifically formulated for polystyrene foam.....no better for cutter loading except it didn't attack the foam.

    I've never had a problem with cutters melting foam and I'm 15krpm-22krpm. I bought several 1/4" double flute spiral bits that were supposedly cut and raked specifically for cutting Styrofoam. They work very well but can't say really any better than same style wood working bit. I use wood working bits for all kinds of profiles, mostly round over and ball nose for fillets, chamfers, etc. They can be much larger than 1/2" diameter and higher surface speeds....still no burning though. The open gullet of a wood working bit is good at clearing chips. I do get rougher finishes if I plunge and cut on both sides of even the up cut, double flute, spiral cutters because they cant clear chips so I usually make roughing cut slight over size, then finish cut .030-060" material cutting only on one side. Then it doesn't seem to matter even if you climb cut........produces very nice finish.

    Work? Yes, up to a certain point. I'd say above 2lbs/ft3, forget it. Higher density foam is stronger, machines, and finishes better. At nearly double the density of Foamular 150, you have a lot more foam to evaporate. It will reduce molten metal advancement speeds (typically 1-2"/sec for Foamular 150) which will in turn reduce the total distance your feed system can travel into/along a pattern for a given pour temperature. Higher density foam is also more prone to fold defects. High surface area to volume (SA/V) parts will fare better, Low SA/V parts worse.

    Foamular 150 = 1.30lb/ft3.
    Foamular 250 = 1.55lb/ft3

    Best,
    Kelly
     
  18. rocco

    rocco Silver

  19. Al2O3

    Al2O3 Administrator Staff Member Banner Member

  20. ka_6772

    ka_6772 Copper

    Do you guys think dipping the print in latex is a bad idea? The issues from last week's casting seem to have been caused by slurry getting into the print. I had repaired some spots with hot glue and coated the whole thing with shellac. It looks like the shellac has been dissolved by the slurry. I'm cooling and cleaning out the molds so ash isn't an issue.

    Thanks,
    Kevin
     

Share This Page