Turbine Engine Mid Frame

Discussion in 'Lost PLA casting' started by Monty, Dec 2, 2021.

  1. Monty

    Monty Silver

    Thought I'd start a thread for my latest casting project. It will be lost PLA in block POP silica sand investment. This is a part from a turbine engine I'm building. I've been working on this for a while. This is the second design. I made two castings previously, the first didn't work out, the second one did. I machined the second casting up to a point, but decided to change the design because of what I learned during the process. The raw casting is about 15 lbs in 356 Aluminum. I will be using a direct pour riser with a ceramic filter in it. I've had good luck in the past using these filters to cast pistons.

    Last edited: Dec 2, 2021
    BattyZ likes this.
  2. Monty

    Monty Silver

    Here is an example of the results I've gotten using the casting filter and direct pour riser approach in the past. Note the inclusions in the commercial piston! I melted these pistons down and used them to cast mine (after making degassed, fluxed ingots!). But the direct pour with a filter really gets rid of the inclusions.

  3. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Do you have a little more detail on the "Direct Pour" feed system?

  4. Monty

    Monty Silver


    Basically you just pour directly into the main riser. The filter sits at the bottom of the riser and controls the flow of metal into the mold. There is a certain pressure head the filter needs to start flowing. Usually the riser is about 3/4 full when the metal starts to flow. The filter keeps any oxides from entering the mold and ensures a nice laminar controlled flow of metal. Since the last metal poured is in the riser it's the hottest, and the last to solidify. You can see the casting taking up metal as it solidifies. I usually put a propane torch on the riser until the casting stops drawing metal. I read about this system online, but I haven't ever been able to find any technical info on it. I just tried it once and I've been learning as I go. You can get pre-made pour cups that the filter fits in the bottom of. I'm thinking of making one like you use to see if I can re-use it.


  5. Monty

    Monty Silver

    I made a little progress today. I finished the flask for the casting. Haven't made the investment yet. First I coated the PLA form with a mixture of water and propylene glycol. This is to act as a surfactant and prevent bubbles from attaching to the surface. The PLA form was screwed to an acrylic disk which forms the bottom of the flask.

    I used aluminum sheet to form the walls. The whole thing is held together with packing tape.

    Hardware cloth was made into a form that will contain the POP investment. This maintains the structural integrity of the investment even if it cracks (it happens...).


    Here you can see the casting filter that will fit in the bottom of the riser cup.


    I may pour the investment tomorrow if I have time.
  6. BattyZ

    BattyZ Silver Banner Member

    Definitely an interesting feeder system. I think Olfoundryman did something similar when pouring his four barrel carbs. A solid stream to the bottom then all of the bifilms float to the top riser as the system fills? Can't really knock it if you are getting proven results.

    The mold prep looks great! Will be fun to watch this thread.
  7. Monty

    Monty Silver

    I'm going to try purging the cavity with Argon before I pour. Should cut down on any oxides. But in the past I didn't have much trouble with it. Usually gets machined off one end or the other of the casting.
  8. John Gaertner

    John Gaertner Silver

    Question, what PLA are you using? I am doing some slip investment casting on parts and using PolyCast. Am trying to get as much information as I can on the burn out, time and temperture. I have been in touch with PolyCast people and they want a really hot temp as posted by me, earlier. Looks like a very interesting project.
  9. Monty

    Monty Silver

    I use natural PLA from Push Plastics. Prints really nice, doesn't stink, burns out clean.

    It is a bit hard to see imperfections when prepping for investment because it's translucent. Light gray is pretty good too, but natural leaves no ash. I'm going to try a little green food coloring in the PVA seal coat next time to try and make it easier to see where work is still needed.
  10. DavidF

    DavidF Administrator Staff Member Banner Member

  11. Monty

    Monty Silver

    A word on investment. The commercial stuff is the best no doubt. But it's expensive. If you are working with large parts like I am, it's prohibitive $$$. There is a recipe floating around here. It's 1 part plaster to 4 parts sand. WAY too lean IMHO. You have to keep in mind what this was for. BIG castings. The art people first sling a thin mix with silica flour and plaster that is like thin drywall mud all over the wax. This captures the detail. The heavy investment is then poured around that coated wax. In my experience this is going to be way too stiff and brittle. It needs a chicken wire flask just to hold it together, and needs to have foundry sand rammed around it to keep the metal in during pour. It won't flow. It will be lumpy and the sand will all sink to the bottom. The water will rise to the top. DON'T use this for small to medium sized PLA parts like I'm doing here.

    For this kind of part, you want a somewhat soupy investment. It needs to be viscous enough to keep the sand suspended. Even mixed using the peaking method where the dry mix is added to water until islands form it will be a little too thick. You need to pour it slowly and let it rise all around the part. This keeps air bubbles from sticking to the part, and captures the detail. The lumpy thick stuff is going to have air bubbles and all kinds of defects.

    I've used USG Hydroperm. The mixing instructions are very specific. You need warm water, weigh the plaster and water, and mix for a specific time. It's so soupy you think it won't ever cure.....but it does. No sand or anything added. Works well, but the shipping will kill you. Seems like styrofoam after oven cure. Really weird stuff. Great for RPM process though.

    I like a 2 part plaster, 1 part silica sand, 1 part clay mix. Mix a little soupy, kind of like thin pudding. The clay keeps the silica in suspension and adds strength. Use clean sharp white silica sand, not play sand, or some other locally sourced crap. River sand like that has organic impurities that cause gas inclusions......just don't waste your time. Silica is cheap. Start by putting the clay and plaster powder in a bucket. Put the sand on top. Then mix it up. If you don't put the sand on top you are going to make a lot of dust. It mixes really well if the sand is sinking into the finer clay/plaster mix. Add this mix to water til it peaks. Cold for longer working time, warm for faster. You can let it slake for 10 min or so. Then mix with a drill and mortar mixer for about 3 min. Adjust consistency with water if needed. Pour into the investment. FWIW...

    Last edited: Dec 11, 2021
  12. Monty

    Monty Silver

    Maybe my eyes are just bad, but there is no contrast with the natural stuff....One coat of primer and all the defects show up. I'm hoping a little color in the glue I use to seal it will help.
  13. DavidF

    DavidF Administrator Staff Member Banner Member

    I used to do alot of lost pla casting for a 3d printing company. They supplied the prints, and I would cast them. They loved to use the natural pla printed with a single layer and 10% infill. I chuckled to myself when I read your post. It is the exact experience I had myself and was good hearing that I wasn't the only one out there with this experience.
    I've tried alot of different filaments over the years with mixed results. Either they don't print well or belched black smoke during burn out.
    I even burned out abs. Man that stuff stinks up a city block!!
    I've switched to using the pla silk filaments for casting . They print very well, low warp, great finish. This filament he's some unusual properties, you'll notice when it it extrudes out the nozzle that it puffs up rather than forming a thin wire and that it is softer than regular pla.
    This allows me to print with multiple shells and top layers yet still be able to burn out without cracking the molds.
    The only downside is the stuff in the filament that gives it its sparkling appearance does not burn out and leaves a bit of ash. So far it has had little effect on smaller castings, but I am aware that it's is there and could cause inclusions on larger castings.
    But over all its what works best for me ;)
  14. Monty

    Monty Silver

    I had a few SNAFUs along the way. Moved my burnout oven location which required new wiring. Got all that done. Couldn't get any 356 ingots....don't know what's up with that.

    Soooooo...Melted down and degassed a bunch of old failed castings, sprues and such. Made some nice fresh hot alumuffins, and put a mold in the oven for Christmas. Hope to cast tomorrow.

    IMG_20211222_094727_075.jpg IMG_20211222_141422_328.jpg

    While I was at it I melted down a bunch of failed piston castings, and pistons to use for the diffuser. 2XXX Al piston casting alloy has about 2X the strength of 6061 at 400F. Used in air cooled aircraft and diesel engines.
    Tobho Mott and DavidF like this.
  15. Monty

    Monty Silver


    While looking through here I realized I only partially answered your question. For burnout there are two considerations. The PLA and the plaster. I try to keep the burnout temp fairly low around 450-500 F. Sometimes I take it as high as 600 F if the filament is stubborn about burning out.

    Natural PLA from Push Plastics starts melting and falling out of the mold around 170 F. It starts to sublimate around 350 F or so. I heat my molds slowly, and put a thermocouple in the investment so I can keep track of the progress. Heating too rapidly causes mold cracking. I usually use a drill to make some 3/8 in holes into the riser and vents to help the hot air get into the print and start melting it. I start with the oven on 170 and ramp up to 250. Then I watch the thermocouple. When the mold gets past 212 F I know the bulk water has all evaporated. Then I ramp up to 350-400 F on the oven. For some reason the molds will usually stick around 230 F. Must be some other phase change/chemical reaction that happens there. Once the mold gets past that I'll ramp up to 450-500 for final burnout. I know it's done when there is no more steam or smoke coming from the oven, and the investment is at the same temp as the oven. For a mold this size it takes 24-36 hrs to complete the burnout. When using other PLA that needs higher temp, I'll go up to 600. I usually pour metal with the mold at 300 F or so. Before pouring I use compressed air to blow out any ash or residue.

    Here is a good resource on mixing and curing metal casting plaster.

    Last edited: Dec 23, 2021
  16. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    That's great info Monty. Just wanted to offer a couple of comments and perhaps you could clarify several other details since we have international members and often times generalizations can cause some confusion. Also, there have been a number of threads about investment recipes, and some differing views on how effective they are, but you seem to have a process that is working well for you so thought I'd invite some more input.

    Plaster can be a very generic term. There are many kinds of plaster. You attach the USG document and they of course have many engineered products fit for purpose. Usually when someone says plaster or more specifically PoP (Plaster of Paris), it's gypsum based, but there are others. You mention PoP in the opening post so I assume you mean it, and further, IIRC, you are using garden variety PoP from the big box stores and modifying it as you describe in post 11. Is that correct? Also clay can be a fairly broad term. What kind of clay are you adding to the investment mixture described in post 11?

    I've seen that USG publication before but your post prompted me to take another spin through it. Several things that caught my eye was the large variation in burn out temps and mold shrinkage (or even expansion through certain regions). I mention this because if one thought "plaster is plaster", and used a burn out schedule that never got higher than 400-500F, you could have a very serious problem pouring metal into some investment materials. Not doubting your process for your materials, just saying some materials will require much higher burn out temps.

    The USG documents comments on the affect of both mixing time and burn out temp/time have on the dimensional change of the mold. Though most are ne numbers are in the .3-.9% range as high as 1.7%. is cited. Wow! Typical aluminum shrinkage is in the range of 1.3%. I use that for lost foam casting and with vibe packed sand media molds and that is dead on. If you are a technical caster, mold shrinkage of the amounts above is non-trivial. What do you see with your mix and what total net shrinkage do you target with your pattern? Inquiring minds want to know!

  17. Monty

    Monty Silver


    I've used both USG hydroperm and my "budget" mix. If I had a local source for hydroperm that is what I'd always use. It's a little fussy with the ratios and the mixing process, but yields a very nice porous investment. Unfortunately I have to order it from Vermont, and the shipping costs more than the product. If anyone knows of a good supplier in the center of the country, I'm all ears!

    For my mix the clay I'm using is just powdered ball clay from the local pottery supply. The POP is big box garden variety POP. The sand is 70 mesh silica sand from the local concrete supply house. I'm always playing with the recipe. I started using clay when I couldn't find silica flour. Silica flour can also be used, and actually makes for a very good investment. Helps with surface finish and detail capture. The clay seems to make the mix more viscous. It helps the sand stay in suspension until the plaster hardens. The investment does seem to be a bit softer though. I wouldn't use the clay on small detail type parts. I'd use silica flour instead.

    I've never had any problems using this burnout temp with aluminum. I think for bronze and brass you need to heat the plaster to 1300F. There is another reaction that happens at higher pour temps (IIRC). since I only work with aluminum, I've not had that problem, even with my ghetto big box mix. For hydroperm, I don't think you need to burn out to higher temps for bronze etc, but for regular POP I believe you do. One word of caution is not to use generic river sand. What we have here is not suitable. Lots of fines and other things in there that cause out gassing in my experience. I only use clean silica fillers for that reason.

    The shrink factor I use is 1.3%. So far as near as I can tell the dimensional error of the castings is coming out less than. .5mm!! I was amazed at how close the dimensions wound up. It's hard to even measure, because there's that much surface imperfection from the 3D print.

    Last edited: Dec 23, 2021
  18. Monty

    Monty Silver

    Merry Christmas Everyone!

    I must have been a bad boy, the casting gods sent me a lump of coal. Not a success, but not a total failure. I learned something about ModPodge. It's mostly polyvinyl acetate, but they must add some acrylic to it. The investment appeared to have burnt out cleanly at first glance. There was almost no residue in the cavity when I blew it out with compressed air. The pour seemed to go well at first. Riser filled up, then started taking metal. Partway into the pour though it stopped taking metal, and then started again. (uh-oh!) What was that all about? Unfortunately I soon found out as the dreaded gas bubbles started in the main riser. It smelled like burning acrylic. I had zero hope of any success as the bubbling continued for some time. I was surprised to find a mostly good casting when I cleaned the investment off. There was char in the investment at the bottom of the cavity. My theory about what went wrong: there was not enough air circulation to completely burn out the acrylic in the top of the mold (upside down in the oven). It appears to have burned out cleanly everywhere that air could get to. ModPodge doesn't look to be ideal for this purpose. I'm going to use straight Elmer's glue next time. ModPodge did do a good job sealing the print and filling in layer lines. I didn't have any investment intrusions, and most of the layer lines went away. You can see the brush strokes in the casting surface! Elmer's should do just as well, but should just be all polyvinyl acetate. I'm going to try thinning the glue with water so I can apply it with an air brush. I'll also add some food coloring so I can see the coat as it goes on. Should result in a very nice finish. It's not a total loss. Despite the gas defects, I can use this casting to prove out my cnc code before putting a good one under the knife. I'll add some vents here and there to help with burnout, and to ensure no gas traps. The direct pour riser is probably the only reason I got anything this good out of this disaster. It allowed the gas to vent while continuing to supply metal to the casting. I'm sure the center portion is full of porosity though. Oh well, better luck next time!


    Attached Files:

    BattyZ likes this.
  19. Al2O3

    Al2O3 Administrator Staff Member Banner Member

    Fails are a bummer Monty but as you say, the part looks pretty good for a fail, and the mold must have worked well.

    Might be too low of viscosity but ever try Poly Vinyl Alcohol? ....as in the barrier coating for parting compound? I've used it to glue lost foam together because the alcohol doesn't attack the polystyrene and it seemed to burn out fairly cleanly. It can also be sprayed and it dries very fast.

  20. Monty

    Monty Silver


    I was actually thinking about that the other day. I have some I use for mold release. When I looked into the physical properties, the melt temp was quite a bit higher than Polyvinyl acetate and above my 500 F burnout temp. The reason I went with Elmer's is the temperature compatibility with PLA for burnout, and the low/no toxicity of the thermal decomposition compounds.

    I'm also wondering about dipping the part in the same shell mix you use before pouring the investment around it. To try and eliminate surface bubbles. I had a few goobers on this casting. They aren't a big deal though. I just grind them off with a die grinder. It might cause other problems though, like a thin unsupported shell with a bubble attached, leaving a pocket behind it.


Share This Page