Intake Manifold Casting
 

This is a long term thing, but what the hell.

Long story short, my dad and I built a home foundry setup and I want to try to do something a bit interesting to get better at the whole process. Since we also have a Lotus 7 clone we're building and there's a nice straight shot out the side, I decided to try to do this.

Whole idea is a manifold to hook an Arctic Cat M1000 TB setup to an EProd ported 13B. Why that combination? The Arctic Cat TB because it's the cheapest 50mm twin throttle I've yet found, and the EProd ported 13B because we have the parts. Runner length to the end of the throttle body bellmouth is something like 7 inches, which is pretty short but driven by packaging.

Production process is intended to be lost PLA for the first one at least; we have a home 3d printer, so can go directly from CAD files to physical parts in plastic... and PLA plastic can be invested in plaster and then burned out like a lost wax casting. Over the long term, may redesign everything to be effectively sandcast, and maybe to fit into a normal car, not one that can tolerate having the whole induction system going straight out the passenger side.

Right now it's winter and molten aluminum hitting snow is probably pretty bad (you know, steam explosions spitting thousand degree metal around) so that part's waiting until it warms up but... hell, thought people might be interested.

https://www.rx7club.com/attachment.p...1&d=1425437749

Heck, I'd stick with plastic. Lighter, thermally nonconductive, you don't have to jigger around with deducing how much larger to make the mold so when it cools and shrinks it's the right size and shape, and the best part is you'd be starting with it anyway :)

I wonder how well the printed plastics would take to being given a cursory wrap or two of glass or carbon fiber mat/ribbon/whichever.

In theory yes! In practice though, two basic issues. One is that the plastics that my printer does aren't really suited to engine bay temperatures, at least not for structural stuff, probably anyway. I don't think anyway. And there's issues with getting to full theoretical strength. Wrapping it in carbon or glass could work though, to be perfectly honest.

Number two though is that I want to cast it so I can get better at doing dimensionally accurate (or close enough for finish machining) engineering-grade castings rather than stuff that just looks neat. I could likely do a plastic intake manifold, but I can't do a plastic transmission casing, or cylinder head, or so on and so forth. I have a lot of stuff I want to try to make.

Watched a few home casting vids on YouTube a while back. Very cool.

the FDM stuff isn't strong enough but works great for when people want use it for molds or carbon fiber wrap then just use a solvent or acid or something to melt the plastic away.

funny, we were doing leg work on this a last week. the engine in question is a PP though, so the intake is much simpler.

i forget which material we looked at, i want to say PLA, but i could be wrong, anyways it has enough margin for heat, but it wasn't compatible with gasoline.

we almost built it out of acrylic, which would have been really easy, and pretty cool, although i think it would have withered eventually.

the Rx8 intakes are a nylon, btw.

we settled on just building an aluminum intake, which i know is boring.

but yes do keep working! my friend 3d printed the casting mold for a ring, and it turned out really good

which 3d printer do you have?

Lessee...

Nylon is actually an option for 3d printable materials now, but it takes a lot of different stuff than what I have available. In theory I think maybe it would be possible to do an FDM printed nylon intake manifold. Might try it in a few years. Like I said before, a lot of the queue of stuff I want to make wouldn't be suitable for plastic anyway so I really really want to build the skill to do the whole process from CAD to aluminum.

3d printer we have is a BfB Rapman which is a price point they don't make anymore... various business rethinkings and they decided to go upscale. For its time it was pretty good bang/buck (tested design with tech support available as a package with a CAD system) but that time was before things like heated beds were commonplace. The build area is a lot smaller than I'd like and there's various other niggling details; next evolution is going to be something like a bigger Ordbot Hadron. I want to be able to do stuff the size of a bellhousing (for one example) and it'd be nice to not have to glue together a whole bunch of smaller pieces to do something.

there is certainly no harm in trying an intake, what is the worst case? it melts?

i do agree 3d printing is close right now, its not good enough for interior parts, but it can't quite do intakes either, unless you're willing to buy the $$$ one.

I wish my company would look into SLM... Skip the plastic step altogether. Looks like a killer project, do keep us up to date on how it goes!

SLS Ti becomes more and more affordable. so it will be a matter of time before one prints a rotor weighing 2.5 kg. only a bit difficult thing for a rotor is the side seal groove. one could make it in a lathe but then the grooves would extend till the rotor face, much like the smaller sachs and aixro engine are made. My idea would be to weld the very end with a TIG and then mill the hole for the corner seal afterwards.

Anyone has an idea how mazda makes this groove? Using a .7mm mill would take ages

Plunge EDM maybe? Dunno, that really is an interesting manufacturing question.

I'm actually looking to metal matrix composites more than titanium - aluminum with SiC particles is castable, and Honda was using it as a piston material in F1 MMCs got banned. The strength, stiffness and thermal expansion characteristics are interesting at some blend levels.

I think casting an intake manifold is ambitious enough for me for Q1 2015, though.

Yeah, building a sand mold to cast a rotor is really something over my head. drawing it and sending it per email to print in Ti is a bit more in my confort zone :lol: Probaly it will take a few iterations to get the clearances and eventual coatings right to prevent the seals from galling.

Plunge EDM could be but I cannot see it could be remotely affordable on mass production scale. its really slow and electrode wear would be also be huge a factor
Maybe they have a kind of lathe that only turns a few degrees and lifts the tool, turns back and this at a high speed sequence?

Please keep up updaed on the casting. really intresting project. Do you need to cast it in inert gas or can you just do it in normal air?

Way I figure, at the scale I'm operating at I can afford to experiment to figure out how to do it. Given as how I'm using a bunch of parts I had sitting around to build the foundry, I only have something like a hundred bucks into that side of things. I don't think I can as easily afford to have other people do one-off castings for testing purposes for me, much less (at current rates) afford to do laser sintered bits.

Aluminum is relatively easy; you can do it in air, at least for my purposes. Oxides form on the outside, but you just skim that off before you pour. Magnesium you need to do special things to though - unlike aluminum, the oxides magnesium form are porous, so they don't protect the rest of the melt and the whole thing apparently just starts burning. Kind of exciting sounding; I think I'll pass.

I'm pretty sure they milled them with an end mill. No big deal really except you need a lot of RPM. We use .020" diameter cutters all the time a work but we have very high spindle speeds available.

One could possible plane them in, like broaching or like how a shaper works. Incremental scraping along the path, with a tool of proper width, at , say, .001" deeper each pass. Whatever you could get away with. Maybe rough and finish tools.

we built a PP intake this week, and the first run was to 3d print the flanges, to check for fit. we considered a plastic intake, but suitable materials, and the PITA it would have been to draw up, meant we made it in AL. plywood is our mockup engine....

and a gratuitous picture of NODE in a 787, just to prove that anything is possible

lol node

i remember him. and the siezatar.

if you don't know NODE, we think you're a n00b, so good job!

What's kind of worst about this is between this and the other threads in this section, now I'm pondering redoing my peripheral motor in the style of the MFR housings - that is to say, flanges on the inlets instead of tubes with hoses. Basically cast the tubes and flanges then epoxy them in like they were just normal homebrew PP tubes. Sure I could just weld aluminum tubes to aluminum flanges, but I don't have an AC TIG box. And heck, after that it's just a "simple" job to make a slide valve throttle body...

... well, you sorta see why it gets real attractive to have the ability to cast stuff. One of the more pipe dream projects is a new(ish) design trans case to have an in-tunnel gearbox that takes Hewland FT gears and has an H-pattern shifter. Seems to me it'd be nice to have a dogbox with readily available gears.

Now there's a rabbit hole I wouldn't want to go down :)

Yeah. I'm not sure if it's "crazy like a fox" or "crazy like a crazy person". I have it conceptually figured out, and I think it can be done with only a couple shafts being manufactured (on top of the castings obviously) and I think there's a hole in the market that it fills... but even the first stage of getting the design done is a hell of a big step. Ideally I'd start by finding an FT200 to crib dimensions from, but I seem to not have any of those kicking around, and who'd have thought that purpose built racing transmissions were expensive?

Ah well; way I figure, what's the point of having un-ambitious goals?

I picked up a lincoln 185 ac tig for my pp projects for $1000.. criagslist score.. it was 1750 new but the guy had never even plugged it in. Surely you can find one. My next step is to buy a 3d printer. I love this formum.

Well, I do have to prioritize spending. Fixing DD, various race stuff, casting... a grand into an AC TIG box is a grand not into other, more interesting stuff. And casting stuff is kinda fun.

If it was viable I would tig your stuff for you.

. What's your age if you don't mind? Im 29 and my experiments are self funded. What's the cost of your home foundry stuff? Can we see some pic$? Want to make a forge in the next month.

the welder is a good idea, but there are already a ton of 3s printers floating around, i'm not sure i would buy one, just borrow one

What was said about ambitious goals?
I want to buy a $400-600 ebay assemble yourself 3d printer and from there... build my own starting with a 2x2x2 foot build area... using a mig gun for metal printing... and then on to spening around $6000-8000 building a 10 by 15 by 10 metal printer( or multi material)... what's the point of un ambitious goals? I'm just concerned with getting it done before I'm too old.

as far as the home forge goes... the one i saw online that used a hairdryer to blow the air in one port, and an old fire extinguisher used for the crucible.. i plan to use 4 ports that come into base at an angle to get a swirl effect and provide lots of even heat so i that i can go beyond melting alum... how ever im not sure what to use as crucible if i am melting iron/steel. i feel like i should go find some books on metallurgy...

This process should work good for an intake manifold.

Want do try out some casting my self, but my goals are less ambitious, and mainly just to have tried it.

lastphaseofthis, I'm 32 - all my stuff is self funded but I have a lot of stuff sitting around at my parent's place to cannabalize. The foundry was pretty much an example of that.

I attached a pic of the foundry at one point in its evolution... wife has better pics but she's busy and I just wanted to get something up before I start the day. It's a hot water heater shell filled with homemade castable refractory - recipe from the backyard metal casting site, basically a mix of cement, fireclay (well, oildry), perlite and sand. Inside opening is about 12" diameter and 18" tall, so plenty of room for a big crucible. Heat source is an oil burner from when my folks upgraded to a gas furnace. It's a great little thing - self contained igniter, blower and high pressure pump. Just feed it diesel and it goes. Since added an insulated lid for it on a hinge made out of a VW rabbit clutch pedal... because it was laying around. I think we have all of about $100 into it.

I have designs for a better, less... cobbled together version in mind if need be; this was essentially all to see if doing it at all was practical. So far so good.

edit: And the picture is rotated. Sorry, no time to sort that out right now.

This is a great thread!!! Very interesting and encouraging for DIYer. At this rate with increasing interest we can finally make those hard to find parts or aftermarket ones. Found another trick.
http://gizmodo.com/turning-styrofoam...asy-1694298734

there is some way to 3d print a material like that too, so potentially you can cast side irons, and rotor housings and stuff.

This is know as "Lost foam process". Look at an old 1st gen Saturn cylinder head and you can see the casting actually looks the same as a Styrofoam cooler because this is the method they used. The foam pattern has to be packed into the sand like the pattern in Pettersen's video. I would just use the wooden dowel to make the riser.

One more note on Pettersen's video for people interested in this. In the vid he puts a pre-made solid sand core in the sand impression made by the pattern. Where did that come from? It comes from a core box. One uses a different mold to make the sand core. A different sand, using a different binder, is packed into the core box and then baked to solidify the core. The core box has to have provisions for supporting the core in the cavity and so does the pattern. It shows this in the video.

Even with this it is still not all that complicated. Patterns can be like the ones in the video or split and placed on each side of a flat piece of wood big enough to cover the surface of the cope and the drag. This allows preparing of both halves before closing the mold and insures that the cavity is in the proper position between both halves.

Actually worked on this today; had some issues with the print. No big deal, just that wide flat items (ala flanges) tend to warp. Trying something different; we'll see.

One nice thing that can be done too is making core boxes with the printer too; investment casting from the plastic directly is nice for one-offs but sand casting is probably going to be necessary for any sort of production. In theory it should all work just as well with cast iron, if I could melt cast iron, which I can't yet, heh. Maybe someday.

someone should try this for 20b thick center iron :lol::icon_tup:

Are you saying you made a casting ?.............and the flanges warped?

Im a pattern maker if you need any help let me know.

Nah, the 3d printed pattern warped.

You will get warping and cracks due to stress in the materials when they are cooling down, but there are some methods to fix this.
You could try to bake them in an oven while you put pressure on them, wich should get them close where they need to be.

A better alternative, would be to build a box around the 3d printer and heat it up, so the newly laid plastic cools slower, wich should help. Just find out maximum enviroment temperature for your 3D printer, and heat it a little under that, just make shure the heat from the printer itself doesn't push it above the limit.

In my experience pls is a better choice than abs(not shure what you are using?), cause it requires less heat, so there is less stress created in the material.

Printing several smaller parts and gluing them together could also work.

+ Are you printing a solid part, or do you use support structure with a thin solid surface?

It's a shared machine that's not at my apartment so I don't get as much of a chance to finetune the settings as I might like. Looking to fix that. Right now it's PLA onto unheated Kapton tape. The intent was already to do it in two halves (printer won't fit the whole thing) but the flange where it bolts to the engine is a certain size and really has to be a certain thickness.

Next try is going to be turning it the part 90 degrees - instead of the big flange sitting flat on the bed, it'll be sitting edge-on. That should make it easier to do as a thinner shell too instead of trying to make a 6mmish thick PLA plate. Been a little while since I messed with this stuff.

I'm aware it has to be a certain thickness, but if you design it hollow, and use honeycomb support structure, it will be much faster to print and less likely to warp. You can still have it 10mm thick, but only 2mm solid shell filled with support structure.

I think we're on the same page... it was hollow, but when it was more than a few layers in, it pulled off the bed. I could probably fine tune settings to get it to work at the same orientation, but like I said, I think printing from a different angle will fix it in one shot.

Okay, new printer of my own... much more rigid, heated PEI build surface, etc etc etc. Half scale test more or less worked! It's literally my second print on this machine and I need to fine tune extrusion so it's thin in spots, but it's there. Pictures tomorrow when I can actually get decent lighting on it.

Okay, now I have to remember how to get pictures into the body of the post... ugh. This will probably get edited a few times as I screw with it. https://www.rx7club.com/attachment.p...1&d=1430353411

Well, anyway, a couple days late, but here's a half scale printed version to check it for ability to be printed. Next step is to do a bit of known dimensions and cast it, then measure it afterwards to be able to quantify shrinkage and so forth. Yes, I know that all of this is theoretically known, but there's no substitute for testing. After that, full scale!

Edit: Hey hey, got it!

Oh yeah, while I'm thinking of it. The printer I built can actually do pretty much anything including nylon. In theory. There's still issues with FDM stuff being kinda porous but I could maybe do a nylon intake manifold. Except the lower intake on the RX-8 is still aluminum and I can't help but think there's a reason for that.

Second off, this manifold is intended for Arctic Cat snowmobile throttle bodies, but I was thinking of doing an IPRA style manifold for GSXR throttle bodies too. And for that matter, curved manifolds that will actually fit into normal cars rather than just the specific Lotus 7 clone I'm building.

i've discovered with the 3d print, there is actually a big tolerance, so you wanna do some careful measuring to make sure it came out like you intended

Yeah, working on that too. Last calibration piece I measured out what about a a percent under CAD size, but at least it was nice and square. Fine tuning a bunch of little things - what would be the fun if there weren't still things to learn how to do?

Okay, cast a calibration cube today. Theoretical size was 40mm cubed - actual printed size was about 39.6x39.6x41mm or so... I have it written down. Size after casting... will be TBD. I want to use the same equipment to measure it that I did beforehand which means the sutff at work.

https://www.rx7club.com/attachment.p...1&d=1430708026

Here's the foundry. Yes, it's very very hacked together. Shell is a hot water heater tank, and it's lined with homemade castable refractory. Lid has the same plus some rebar. Heat source is the burner on the lower right side; it's a fuel oil burner we had leftover from converting to gas heat. Hinge is a VW Rabbit clutch pedal with some stuff welded to it - hey, it was handy. Inside size is something like 12" ID by 18" tall. I haven't taken a precise measurement, but burning diesel, getting from room temperature to pouring temperature for a couple pounds of scrap aluminum takes all of about half an hour.

https://www.rx7club.com/attachment.p...1&d=1430708026

This is the result. A few casting defects, probably because I didn't put any vents in or possibly because the aluminum temperature may have been a bit high. Or maybe I didn't let it burn out long enough. One way or another though, they're just surface defects and it should be close enough to let me establish a scaling factor.

Venting and thermal conductivity is where your problems are. Plus, fill. Fill at the bottom and use a stouter runner. Yeah, more waste but it's sand casting...

Yeah, I was taking far from every precaution I could think of; the sprue was literally a heated up candle that I jammed into one side of the cube before investing it in plaster. I was thinking for the intake manifold attempt, something like a 1 1/2-2" sprue between the throttle bodies, maybe half a dozen 1/2-3/4" gates going to various parts of the manifold and a bunch of 1/4" or so risers coming up.

At this rate maybe I should do something smaller with fine details before I try to do the intake manifold. Hm.

maybe do the intake in 2 halves?