Custom V6 EFI intake manifold

[ericjon262]

Howdy folks, I thought I would share the current progress on my custom V6 EFI intake that I'm working on, so far, I've mapped out most of the flanges, and the base of the manifold, the next step will be to start on the runners.

The big plan, the intake is going to be somewhat modular, so that I can change runner length and plenum volume, and maybe even allow for ITB installation. the hardest part is going to be keeping it low, as the engine has a 60 degree bank angle, making it taller, and then trying to cram the injectors in there in a way that doesn't make them impossible to get to.

Now that I'm starting on the runners, I'm going to start reading a little more and try and develop a more optimal performance setup than stock, which was designed around hood clearance.

here is what I have so far, I'll transition the ports to round, and slowly curve them to a vertical to complete the lower portion, and end up with something reminiscent of a f**d GT40 lower intake. the Lower will probably be cast aluminum, depending on if I can keep the mass low enough for my backyard casting setup, and then the upper portions 3d printed if I can find material capable of handling the stresses imposed by a running engine.

I'm very open to advice or tips by those of you who have built one before, I'm currently reading "Scientific design of exhaust and intake systems" but the publication seems a bit dated, is more devoted towards exhaust, and doesn't cover EFI intakes.

here are a couple of pictures of where I'm at now.

[digger]

how long are the ports in the head? will you you require manifold runner lengths to have curves to achieve the longest length you want to be able to run?

[ericjon262]

how long are the ports in the head? will you you require manifold runner lengths to have curves to achieve the longest length you want to be able to run?

the runners will leave the heads straight, and then after some amount of straight running, they'll curve upwards. this will be less severe than the stock manifold, which leave the head and goes nose down about 45*, and then pulls a 90* upturn, but the stock intake was designed around fitting under a hood, this manifold may or may not fit without added clearance.

The average between the SSR and LSR in the heads is about 4.875", I'd like to get the total runner length up around 12-13" depending on space constraints.

I started working on the runners, and the water crossover, the runners are a giant PITA to model, and right now, the only one I have modeled isn't even hollow...


I spent several hours to figure out I was trying to do it the really hard way though. I can probably do the other 5 in an hour or so, but I still need to figure out how to make them hollow. Now that I have a visual for what the runners look like, I'll try make them a little bit longer.



Edit:

And like that, I figured out how to make the runner hollow, but my program doesn't want to do it with the runner in place... I may need to produce the runner as a separate part, and then bring it to the rest of the manifold.

[ericjon262]

I've modeled 3 different runners so far. before I start working too extensively on a specific one, I'm going to get the engine back into the car so I can measure hood clearance. I haven't gone as far as shelling the runners yet, so yes, in the pictures they are solid.

[The Dark Side of Will]

You'll want area taper and to bring them to a common plane for manufacturability.

The runners should intersect the common plane perpendicular. Caddy Northstar runners intersect the manifold mounting plane at an angle, yet people have still tried to mount induction perpendicular to the plane... with lackluster results.

Also, you'll want water take-offs at all four water ports on the cylinder heads

[ericjon262]

You'll want area taper and to bring them to a common plane for manufacturability.

The runners should intersect the common plane perpendicular. Caddy Northstar runners intersect the manifold mounting plane at an angle, yet people have still tried to mount induction perpendicular to the plane... with lackluster results.

Also, you'll want water take-offs at all four water ports on the cylinder heads

yep, two of the runners in the pictures taper as the get further away from the port. the completed design will use 6 identical runners, all meeting on a common plane, with the intent of adding a level of adjustablity to the intake. the plan is to have all of the runners meet on a single plane that can then have longer runners bolted on, or a plenum directly to, an intercooler, ITB's ect.

the runners appear to leave the head perpendicular to their plane of travel, although, I havens cut a head to see the cross section directly.

I hadn't given four corner water flow a whole lot of thought other than more even temps is there another advantage to four corner that I'm missing?

[ericjon262]

another runner. this runner would require opposing cylinder's runners to share a common wall, but it should work fairly well as long as no one tries to hog it out too extreme. the inital curve downward appears to be less drastic than the stock intake. I am thinking about cutting a stock intake and head so that I can get a cross sectional view of the ports.

[ericjon262]

I'm really bad at time management, when there's two dozen other things I can work on, I do something like redesign my intake again. The runners are a bit longer, with an overall length from valveseat to mouth of about 11" not including gasket thickness (about 3/32") I also added in a rear crossover, I'll draw up a tube to connect it to the front, and then a recirculation pipe as well. 

[Caprimaniac]

That's a cool Project.

Did you calculate runner length and CSA needed in telation to RPM, req, flow etc? If you need longer runners, you could make it a crossram; as runners form LH side bank going to RH side plenum. Although Space might be a factor.....

Rear water crossover? maybe if OE manifold use this.

[The Dark Side of Will]

You went taller than you previous revs.

In theory, you can make your runners tapered and then put the throttle wherever you can package it. Any throttle per cylinder system is going to have better response than any throttled plenum manifold, but the best response comes from keeping the throttles as close as possible to the manifold.

Also, by making the runners taller, you've made the bounding box significantly larger, which will be more demanding on your production methods.

Given those two criteria, I'd move the turn to/from vertical closer to the head surface, with the result being that the two rows get further apart laterally and the plane where they stop comes down.

However, I do understand if you moved the vertical turn away from the head to make room for injectors

Have you seen that circle track intake that Shaun used to have?

Rear water crossover? maybe if OE manifold use this.

He's going to make 3x the power of the stock engine, so he's going to need more cooling. Four point coolant return is basically required for high powered LS and small block Chevy builds

[ericjon262]

You went taller than you previous revs.

In theory, you can make your runners tapered and then put the throttle wherever you can package it. Any throttle per cylinder system is going to have better response than any throttled plenum manifold, but the best response comes from keeping the throttles as close as possible to the manifold.

Also, by making the runners taller, you've made the bounding box significantly larger, which will be more demanding on your production methods.

Given those two criteria, I'd move the turn to/from vertical closer to the head surface, with the result being that the two rows get further apart laterally and the plane where they stop comes down.

However, I do understand if you moved the vertical turn away from the head to make room for injectors

Have you seen that circle track intake that Shaun used to have?

Rear water crossover? maybe if OE manifold use this.

He's going to make 3x the power of the stock engine, so he's going to need more cooling. Four point coolant return is basically required for high powered LS and small block Chevy builds

Bingo on the injectors. I'd like to place them outboard of the runners. That being said, I'm starting to try and think of ways to make the intake more modular so that runners can be changed for different shapes or lengths, maybe even allowing for 3d printing of the runners, with the injectors remaining in the base. at this point, I'm working towards seeing how deep I can get a set of injectors, the area just below the valve cover flange (where the mouths of the stock runners are) is probably going to have some fairly deep pits in them to allow the injectors to get down in there.

I'm probably using bad terminology, when I say re circulation, bypass is probably the correct term. The stock 3x00 engines (and LX9) use a ~3/4" tube to recirculate coolant from the thermostat housing, back to the water pump, preventing the water pump from being dead headed when the thermostat is closed. the current design will require something to allow this re circulation to happen. the front coolant crossover will go to the rear crossover(red line), then they'll both go to the thermostat(green box), where, is the thermostat is closed, they will go through the re circulation pipe (blue line) back to the waterpump.



I did see Shaun's setup, it was pretty cool, but I don't see any current pictures of it...

That's a cool Project.

Did you calculate runner length and CSA needed in telation to RPM, req, flow etc? If you need longer runners, you could make it a crossram; as runners form LH side bank going to RH side plenum. Although Space might be a factor.....

Rear water crossover? maybe if OE manifold use this.

Thanks!

in this case, it would actually be a front water crossover, as the rear crossover handles the thermostat, because the engine is transverse, both nomenclatures get a bit confusing... lol!

I've done some basic calculations, to compare curtain area to mouth diameter, and some runner length calculations, but for now, I'm going to try and keep the runner math out of the picture and focus on the modular design described above so that in the off chance I make more than one, they can be used in different applications.

[LoganD]

Why have such a sharp turn? Why transition to round? If you're 3D printing parts there's no reason to go to round.

Personally, I'd leave the narrow shape of the runner at the head intact and remove or eliminate the bend letting the runners cross. They do not need to terminate at the plenum perpendicularly, or you could have them stick into the plenum box.

Getting the runners the right length, shape, and with the least number of bends is WAY more important than getting the plenum a nice shape or the exact right size in theory for EFI. It's dry so just worry about fitting the runners in the car that work right, the plenum will be dictated by packaging. You will likely be extremely surprised how long they need to be to work best.

As others have said a taper is nice here, but one thing that works extremely well is having an elliptical shape at the runner entry that protrudes into the plenum box. So you don't just want a rounded edge transitioning into the runner, you want a "bell mouth" shape where the cross section is an ellipse. Again, you'll be surprised how big of an ellipse ends up being optimal, you'll basically want the largest you can fit.

[hoffman900]

Why have such a sharp turn? Why transition to round? If you're 3D printing parts there's no reason to go to round.

Personally, I'd leave the narrow shape of the runner at the head intact and remove or eliminate the bend letting the runners cross. They do not need to terminate at the plenum perpendicularly, or you could have them stick into the plenum box.

Getting the runners the right length, shape, and with the least number of bends is WAY more important than getting the plenum a nice shape or the exact right size in theory for EFI. It's dry so just worry about fitting the runners in the car that work right, the plenum will be dictated by packaging. You will likely be extremely surprised how long they need to be to work best.

As others have said a taper is nice here, but one thing that works extremely well is having an elliptical shape at the runner entry that protrudes into the plenum box. So you don't just want a rounded edge transitioning into the runner, you want a "bell mouth" shape where the cross section is an ellipse. Again, you'll be surprised how big of an ellipse ends up being optimal, you'll basically want the largest you can fit.

Logan,

Can you provide examples? I've seen the Corvette C7R manifolds with similar features...

[LoganD]

Cut apart a 2018+ Coyote plastic intake manifold. You're talking about a 7500 RPM engine making 84 lb-ft/L (SAE) through a full exhaust with cats and passing CA emissions. You'd be AMAZED how long the runners in these things are, and how ugly the plenum looks. They taper and have an elliptical runner entry. Another good example is the '00 Cobra R, this was a very advanced manifold for the time.

What the elliptical shape does is prevent a boundary layer from forming at the runner inlet, in effect making the runner entry seem smaller than it is. Air at high speed needs a surprisingly large and smooth transition into the runner to prevent this. Often times physical testing will make the engine builder/intake manifold designer think the runner wants a larger taper than it actually does, when in reality it just needs a more smooth transition from low speed plenum air to higher speed runner air. I'd take a runner with no taper and the proper runner inlet transition over one with just a simple radius and taper.

This is where you have to trust the CFD. As intake manifold designers have done this, specific torque has gone through the roof for NA engines in OEM applications despite ever more stringent emissions regs.

As for transitioning shape over the runner length, unless the transition is guided by CFD it will likely cause a boundary layer. These things are often very counter intuitive.

EDIT: I also want to add, when you do the CFD a surprising amount of the flow into the runner comes from the sides and even behind the runner entry. This is why you generally see the runner stick into the plenum if it can be done. Having the runner entry flush with the plenum floor is left over from wet manifold thinking.

[ericjon262]

Why have such a sharp turn? Why transition to round? If you're 3D printing parts there's no reason to go to round.

Personally, I'd leave the narrow shape of the runner at the head intact and remove or eliminate the bend letting the runners cross. They do not need to terminate at the plenum perpendicularly, or you could have them stick into the plenum box.

Getting the runners the right length, shape, and with the least number of bends is WAY more important than getting the plenum a nice shape or the exact right size in theory for EFI. It's dry so just worry about fitting the runners in the car that work right, the plenum will be dictated by packaging. You will likely be extremely surprised how long they need to be to work best.

As others have said a taper is nice here, but one thing that works extremely well is having an elliptical shape at the runner entry that protrudes into the plenum box. So you don't just want a rounded edge transitioning into the runner, you want a "bell mouth" shape where the cross section is an ellipse. Again, you'll be surprised how big of an ellipse ends up being optimal, you'll basically want the largest you can fit.

The runners would definitely be bell mouthed, I've drawn several that way already. the runners pictured here are probably not going to be anything like what goes in the car. at this point, they're more just a stand in as I brainstorm the next step.

Why transition to round? round will have the maximum volume/surface area therefore reducing losses, is it that significant? probably not, but, to be honest, it's also just plain easier to work with than odd shapes. at this point, I'm working on getting the lower surfaces mapped out so that runners can bolt on in any number of configurations. in other words, the "runners" in the lower manifold will most likely end up about an inch long, allowing just enough room for the injector to penetrate the port.

The "sharp bend" is way smoother than the stock manifold, which also has two bends vise the one I have here. but like I said, those runners have pretty much just become placeholders at this point, as I work towards a modular design.