Hi, I’m new to the forum, just joined today in fact and I left an intro in that section. I’m trying to cast an oil pump in aluminum for a model aircraft engine using the lost foam process. For a pattern I’m using the pink higher density foam which I cut using CNC in sections and glue together with PVA. From the photo below you can see that it’s a fairly strange shape. It’s serves as a cover to the cam spur-gears, supports for the magnetos as well as an oil pump housing. So there is a mixture of thin webs 0.125 in thick as well as a solid body. There are no internal features. I have had a go at casting this already. I coated the pattern with a thin coat of drywall compound, brushed on. The first try resulted in an incomplete casting, mostly because I didn’t realize you needed to heat the aluminum much more with LF in comparison to sand casting. The metal cooled before it got to the ends of the magneto supports. On the next try I also gated the sprue to the ends of the support arms, and heated the aluminum to red heat. Here is the pattern after coating with drywall. I got a complete casting but the surface finish was very pitted and cratered. Here it is held up to the crankcase casting as it eventually will be connected. Not a pretty finish, particularly in comparison to the commercial crankcase castings. On the third try I coated the pattern with a casting plaster made by US Gypsum called Hydroperm. This is basically plaster of Paris but with talc and Portland cement and is probably meant for lost wax. They claim permeability from the talc I would think, but I don’t know how it compares to drywall compound in this respect. This concerned me so I added a drinking straw vent …. which of course just filled with metal almost instantly and didn’t really do anything ( no more vents). I ended up with a thicker layer of plaster than I really wanted just due to too heavy of a mix. So , the outcome of this was a complete casting with a very nice finish identical to the foam. But the casting was unusable. When I machined the surface, (the face on the underside of the photo but the upper surface as cast) I discovered the metal was full of voids of various sizes from tiny to several mm, from just below the upper (where the sprue/ gate connected to the main body) surface to about 0.1 inch down. There was no sign of an issue on the casting surface at all. I think this was either air entrained in the metal as it was poured or foam gas that didn’t make it out due to the plaster’s permeability. I should mention that I am using a plaster pouring cup on the sprue and the pour was hot and fast. The cup was embedded into the sand about 4 inches. I’m wondering if I should go back to drywall compound and try a slightly thicker layer, or try a thinner layer of hydroperm. Since the last attempt, I have experimented with applying more watered down layers of hydroperm, and it can brush on very nicely, better than drywall compound I think. Has anyone used Hydroperm successfully in LF casting. I’ve seen a few YT videos where folks have tried plaster of Paris as a coating and had good results as far as surface finish goes but they do not report any subsequent machining aspect of their experiment. I think it’s all down to permeability of the coating and how that is controlled through thickness of application. Without any means of measuring the permeability it’s all speculation of course, and maybe the bubbles are nothing to do with the coating material. I also think I should cast the part positioned more vertically so bubbles rise away from machined (lower) portions of the casting, although of course no bubbles at all would be ideal. I expect this is going to simply need experimentation as far as surface finish vs. bubbles goes but anyway, apologies for this long tale of woe. Any thoughts and suggestions would be welcome. Regards, Alasdair.
Hello Alasdair, I would suspect porosity is the problem since there was no surface problems. First thing is to make sure your burner is tuned to the lean side. Second is not leaving the metal in the furnace too long after it gets to the molten state. Red heat is way to hot for aluminium and it's picking up hydrogen the longer it's in the furnace with the burner on. In the first fail, a 50*F increase in temp would probably have been sufficient.
Mine 5 cents.. I do get more consistent result by using water based paint and powdered chalk in 1 to 3-4 proportion plus some water and soap/dishwasher. Coating must be thin. And you really need a thermocouple, and normal, ceramic/graphite crucible. On lostfoam working temperatures steel crucibles are leaky.
Thanks. The hydrogen absorption I think is always going to be an issue and the point is taken to keep the molten metal exposure to a minimum. I have just received a high temp SS thermocouple so I can at least spot check the metal temp in future. The metal is heated in a graphite crucible, the metal flask is just for knocking the dross into. So by water based paint would that be latex. The latex holds the chalk together until it melts at around 320F, leaving a lining of chalk between the metal and the sand. That sounds like it would be very permeable when the latex melts into the sand. Thanks for the tip. The trick definitely seems to be in getting the coating thickness just right and uniform. Alasdair.
If you search youtube for the username Luciano, you can find several example of the latex/chalk coating in action on his channel.
Not so. The visible (gross) porosity is controlled by the furnace atmosphere. Tune your burner to the lean side and it will go away. I suffered through it for a long time until, by accident, I found the sweet spot. I was producing baby swiss cheese castings and was so frustrated because I needed polished pieces. Trying to polish visible porosity castings is useless. I don't have that problem now.
Ah I think I understand, so the furnace atmosphere is mostly carbon dioxide and water vapor and so by adjusting the air to propane mixture it is possible to change the water content in the combustion. Do I have that correct or is something else being done. How do you lean the combustion, on my simple burner I only control the gas and it pretty much pulls as much air as it wants. Are you forcing air through the burner with a blower.
And if possible - add some argon bubles for few minutes. They drive off any hydrogen and work as safe flux. But this is more nedded for thicker castings. If you porosity is same on thin/thick parts - it is more due coating , than hydrogen from too dirty aluminium / water absorption. And yes my recipy come from Lucianio. Except ftom chalk.
That's it. If you just open the valve fully, most likely it's running rich. Start tweaking the valve closed and listen to the roar of the furnace. When it's rich it will have a dull roar, as you get nearer to lean the sound will change to a sharper, higher pitch. It doesn't take much tweaking and it's a fine line between rich and lean. When you fire up the furnace, play with the valve and listen closely. You'll find the sweet spot. After thought, You can run the burner rich right up to hot short. When the metal starts to melt you can turn the burner down and then it will be on the lean side. It will not add any appreciable time to the melt, maybe a minute longer but you will get really good, clear aluminium.
Bonz is right about the furnace tune and its affect on melt quality but furnace tune wont change the amount of water vapor present in the combustion air, it will always be along for the ride in a fuel fired furnace. You can actually observe the impact from weather conditions and this is well known to commercial foundry men. I use a resistive electric furnace and that helps a bunch. I think the variability you are seeing is more due to your melt quality, your feed system, and possibly the coating. For pour temp, LF typically requires about 100-150F higher than conventional sand casting, but on smaller parts you'll lose heat fast so I'd suggest 1400-1450F. How high of density? If it's Pink, it's probably Owens Corning. That comes in different densities. The most common is what they call Foamular 100, 150, and 250 (but there are four higher densities too not commonly stocked in the retail stores). The common ones ranges from about 1-1.5 lbs/ft3. Higher densities will produce stronger patterns that machine and finish better at the expense of having to evaporate more polystyrene. I mostly use Foamular 150 as the compromise. For me, that is a small part for lost foam. The only reason I mention that is because as parts get smaller, the Area/Volume ratio gets smaller and that means less coated surface area through which to evaporate foam. Still I think your part should be easily cast even with 1/8" webs. I've cast larger parts down to .062 wall with vacuum assist. See here..... http://forums.thehomefoundry.org/index.php?threads/saucer-cup-spoon-–-the-lost-foam-edition.1020/ .......but I don't think vacuum is necessary in this case......read on. This caught my attention. Did you bake that plaster cup in a kiln prior to use? If not, it will be a source of water vapor and problems with your melt at the worst possible time. The shape of the cup matters too. There are related links in this thread of mine. http://forums.thehomefoundry.org/in...-lost-foam-caster-5-years-on.1650/#post-37480 I don't think your part requires the elaborate gating. Just position it at angle and gate into the most massive portion of the part as close to the top of the pattern as possible. The other problem with small parts is the initial contact of molten metal with foam and turbulence in the cup can create oxides and entrained air. Sometimes it's useful to have a straight sprue to absorb this initial part of the pour and gate off the side of that sprue. In your case, you don't mention the sprue size or height, but increasing the sprue height will increase the hydrostatic pressure of the molten metal (some say metallastatic pressure). It's helpful in filling small thin parts and crates more pressure to drive gas through the coating. For coating, I honestly don't think there is any problem with (non-setting) drywall joint compound. You just need to sufficiently thin it for your intended use and the addition of a surfactant like propylene glycol (Dawn dishwashing soap) is helpful. It's just a matter of getting it properly thinned. There is discussion and links to dip coating in the "Confessions" link. Dip coating will produce thin and very uniform coatings. The reason I favor the non-setting drywall joint compound is they are intended to air dry and have less chemically bound water so when properly dried, they will also be drier than any coating that is self setting because, like PoP. The other advantage is you can store it indefinitely after you have thinned to your desired consistency. Just mix it before next use. By the way, that looks like a good size 4cyl model engine. What is it's displacement and did you cast the crankcase and sump? Happy (Lost Foam) casting. Best, Kelly
I newer have any luck with setting coumpound - gypsum or cement have too many water. Ready mix unsetting - i will give them new try. But still - paint / chalk mix is as free from water as possible and you can tailor you composition. And as i try to find more info - latex paint have no latex. Only acryl/vinyl resin. Named latex have more of them. I think i will try actylic primer with chalk next time.
All the acrylic paints I'm aware of have solvent carriers that will attack polystyrene. There's nothing new about using drywall mud as a lost foam coating. There were threads on Alloy Avenue forum 30+ years old discussing it for hobby casters. I had one fellow tell me he was using it in the 1960s. Best, Kelly
With regards to the foam. Yes it is the Owen Corning brand. I’m not entirely certain but I think it’s the 150 material. The sprue and gates are the regular expanded polystyrene. In my attempts so far the sprue has been quite short, about 2 inches in height and 0.65 in in dia. I was worried about the metal cooling so kept it short. The plaster pouring cup is about 4 inches deep and mostly in the sand so the pattern is about 5 inches below the surface. The cup itself was not baked but left to dry in a low humidity warm atmosphere for about a week. The cup is 1.5 in dia and funnels down to 0.65 in at the bottom where it joins the sprue. The intention of its use was to create a head of metal to increase the pressure in the mold. The original cups used were plaster of Paris. I am now making the cups out of Hydroperm plaster as it holds up to the heat far better, so I will bake them prior to use from now on. I think you are correct that there was likely water left in the plaster that flashed to steam during the pour. The first casting that used the drywall compound (the ready mix non setting stuff) but resulted in the cratered finish did have a thin brushed layer on it . It dried to white opaque that just hid the color of the foam. I think the drywall coating must have been damaged during the pour. Possibly by the steam and resulting turbulence ? There was evidence of turbulence and gas in the metal remaining in the cup. The smoking gun perhaps ! This would explain why the plaster coated version appeared to have a better finish, it was stronger and was applied thicker than the drywall so it held up better. Also when comparing the two coatings on some scrap pieces of foam I notice that the drywall reverted to a paste condition as soon as moisture was introduced, whereas the plaster did not. So if there was superheated water vapor in the poured metal then the drywall compound would not have stood a chance of staying intact. The drywall compound seems softer and more permeable than the plaster so I think it’s the better bet. I will dip coat the pattern in thinned drywall compound again the next time. Hopefully with a properly dried pouring cup things will go better. I will lengthen the sprue a little more and also experiment with tuning the burner as suggested. The engine is 1/4 scale DH Gypsy mk1. The first engine DeHaviland built themselves after their supply of WW 1 Renault engines began to falter after the war. It’s 57 cc displacement. The model engine was originally supplied as a casting kit by a model engineer in England, who passed away some years ago. My first engine (below), a V twin, was also a design of his, the drawings and a casting kit are available from a company in UK. As it turned out I ended up sand casting the crankcase of this one after a machining mistake, the heads and cam housing castings are also supplied. The designer’s son makes his Dad’s other engine drawings available such as the Gypsy but does not produce castings. These Gypsy crankcase castings were professionally cast by a foundry in Canada. Someone commissioned the work and they have some stock of them still available. They seem to be good quality. The drawings call for some other cast parts in addition to the crankcase halves, which the foundry does not produce, the part I am working on being one of them. Thank you so much for all the helpful suggestions. I feel renewed vigor to keep going. Best regards, Alasdair.
I think you will enjoy a better result. Here's the pouring cup I evolved into using. http://forums.thehomefoundry.org/index.php?threads/reuasbale-offset-pouring-basin.688/ But, if you bake the hydroperm on a similar schedule to what is recommended for use as investment, I see no reason why that wouldn't be suitable. Do/are you trying to get more than one use from them? If they are expendable, they could be sodium silicate bound sand or maybe even green sand. I have 50-100 pours through the moldable fiber cups and really like them but only good for non-ferrous work. Good luck on your next casting session. Best, Kelly
Thanks. The next try resulted in an improvement on previous attempts and I think I have the coating technique correct now. I sprayed the pattern with water / Dawn mixture and dipped it in drywall compound , it covered the wax fillets fully and I can’t complain about the surface finish, it was pretty much the finish on the CNC cut foam. I still have porosity but not as bad as last time. It shows up on the polished surface here. I still need to dial in the propane setting to keep the humidity a low as possible and I think I could lower the propane some more without extending the melt time significantly.. For the next attempt I have ordered some moldable ceramic fiber and will emulate your pouring cup as I think it will definitely lower the air entrapment considerably. Rgds, Alasdair.
If you lightly sand the pattern, the finish can be improved significantly over the cut finish. I usually just use 220 grit with several very light pressure strokes on the exposed surfaces, but the foam finish can be improved to 320 grit. You might be surprised how little effort it takes to nicely polish up a foam pattern, especially small ones. The smoother finish will also be less inclined to retain bubbles when coated. Best, Kelly
I still think you could try thinner coating. Write some text with black marker, and coat. If you can see text via dry coat - you are here.
Yes, agreed . I’ve watered down the drywall compound some more and only dip it once. There is just a faint trace of the pink foam along edges. There are sometimes areas on wax fillet that need a flick of a brush to cover but I’m definitely using less now. I just have to say the pouring cup made from inswool moldable ceramic fiber works great. What a difference ! , the pour is now a very much less dramatic event, and it so much easier to keep a steady flow without adding air into the metal. I found the inswool a bit tricky to use, I got the 10 oz caulking tube and squeezed it all over a foam pattern and squished it into shape but it tends to stick to gloves better than foam. I left it a couple of days then put it in the furnace. I let the foam core burn out as it still wasn’t very firm to remove it . After a 1600F soak it was glowing red and came out like a piece of porcelain. A badly made piece that looks like a failed second grade art project.. but it does the job. I wouldn’t consider doing a pour without this now. Rgds Alasdair.
