How do I build a sheet metal intake?

I have some ideas that I want to put into play and I have a fabricator who can build anything that I can articulate, and he'll work for beer.


Thing is, I don't know if I need to start with nothing. Are there pre-made flanges I can buy?
Perhaps a valley cover with cutouts where the runners have to go so that I can have runners with a sort of "mini-flange" at the end.
For that matter, is there premade tubing that can be cut to the lengths I want, or do I need to make them from scratch?

Also, how do I calculate the proper cross-section so that I can maintain velocity without getting into too much frictional loss because I went too small?

And a plenum...
If I were to accept that I should have a plenum space roughly equal to 1/2 my engine displacement, do I count the runners as part of the plenum space or do I count only the common area?

Length tuning is another question.
Is there a reference table or rule of thumb to guide me in terms of the runner length and cross section?

What about a venturi?

I had a thought about putting a mild hourglass shape into the runners just before the intake port on the head to give the airflow a last minute "kick" into the head and "tighten up" the airflow a bit so that the main "stream" of air going into the head is trying to be "narrower" than the passage it has to move through.

Finally, for a throttle body is a "two throat" or "monoblade" design better, and why? Or does it even matter?

IMO, you are in way over your head here.

No flame intended, but successful manifold design takes a great deal of knowledge of engine design. There have been a few threads about this with recommendations of books to read.

By the way, what engine you are talking about?

Originally posted by OldSStroker
IMO, you are in way over your head here.

That's never stopped me before... ;)


BTW, it's for an LT1.


I had been thinking about taking an LT1 manifold, slathering the inside with grease and doing a wax cast of the runners to see what they "look" like, but I figured it couldn't hurt to ask.

I'm really not completely ignorant of fluid dynamics, but I wanted to keep my questions simple at first, figuring that I could always complicate them later.

Well, it's for a street engine, under 7000 rpm.

I had in mind individual runners in a sort of a hockey stick shape as the approach the ports, meeting at a cylinder in the center, but drawing from something close to the opposite side.

I'm picturing a square tube coming almost straight up from the intake port, with a slight narrowing in two opposing tube walls shortly before the port, in a sort of hourglass shape.
My thought is that this narrowing would have a venturi effect, "narrowing" and accelerating the incoming air charge just before it enters the head.

Above the "hourglass", the tubing bends gently toward the longitudinal centerline of the engine, where it goes above a center tube, curving slightly around and down and drawing air from the opposite side of the tube.

So if you were standing in front of the engine and looking at a runner coming from the passenger side bank, the tube would come up, make a right turn, pass over the center tube, touching at the "12:00" position, continuing and wrapping around the tube from the 12:00 to the 2:00 or 3:00 position, with that "window" (from 12:00 - 3:00 for instance), being the opening to the tube. There would also be a widening of the runner tubing in such a way that the widest cross sectional area would be right at the point where it joined the center tube, with a gradual and smooth narrowing down to the size of the actual runner.

The center tube would be a simple aluminum cylinder, possibly 3.5" x 18", which would yield a volume of 173 cubic inches, which would be 1/2 of a 346 cubic inch engine, or about 45% of the volume of a 383cid engine.

I would mount a TB at the front end of the tube, probably a monoblade, so that the bulk of the incoming air flow can flow down the centerline of the tube, instead of towards the perimeter of the tube as it would if I used a "two-hole" TB.


In fairness, I am not giving any consideration to hood clearence, assuming that if i build this I will design it to work as well as I can and worry about modifying my hood as needed after the fact.


So the main questions that I have in mind are these:

1) Will the narrowing in the tube just above the intake port have the effect I think it will?

2) The "diameter" of the tubing is fairly easy if I just follow the lead given by the size & shape of the intake port, but the proper length is an open question.

3) How do I address the "turns" in the tubing so as to maintain flow & velocity. For instance, lets say I have 1x2 tubing and I have to make a 135 degree turn. Do I maintain the size of the tubing, or do I increase the cross section to allow the airflow to make the turn with a minimum of frictional losses?

4) Relating to #3, do I expand the cross-section of the tubing on the short side or the long side?

5) Is a center tube of 3.5" x 18" a suitable size plenum for a 383cid engine when it's only 45% of the engine's displacement?

I found the best way to do hard projects is to copy a preexciting design do a search in yahoo and take a look how other companys are making them, that way your design is down with shapes, and how the runners are formed then all yuo have to do is find out how long you need to make them, and what size plentum.

Originally posted by LameRandomName
Well, it's for a street engine, under 7000 rpm.


So the main questions that I have in mind are these:

1) Will the narrowing in the tube just above the intake port have the effect I think it will?

Generally you want the runner to have a constant taper of 1 or 2 or ? degrees toward the head. IMO, somewhat longer (than LT1) runners help tuning at a usable rpm. This tuning causes a positive pressure, sometimes of several psi at the inlet valve as it is closing. That's part of how VE of 100%+ is possible NA. Remember that the venturi or hourglass shape accelerates the flow which drops the pressure. That sounds like the opposite effect desired to me.

2) The "diameter" of the tubing is fairly easy if I just follow the lead given by the size & shape of the intake port, but the proper length is an open question.

Yep, length is critical for the rpm range you are tuning. It gets back to what you want the engine to do, as aggie said. What shape torque curve do you want? what's head flow? exhaust plumbing? what trans, rear gears, tires, vehicle weight, etc? Those are the things you should consider in any engine design. The manifold is maybe the second last thing you need to specify. And the last would be....?

3) How do I address the "turns" in the tubing so as to maintain flow & velocity. For instance, lets say I have 1x2 tubing and I have to make a 135 degree turn. Do I maintain the size of the tubing, or do I increase the cross section to allow the airflow to make the turn with a minimum of frictional losses?

If you are designing a manifold from scratch, IMO, you absolutely don't want a 135 degree turn anywhere, especially in a runner which is probable in the 5 to 7 inch length. One of the reasons single plane intakes like a SuperVic works is the almost straight shot from the plenum to the port. Sure there's a 70 degree or more turn, but you can see thru the ports from above. Air struggles to turn, especially short, sharp turns. Long sweeping turns like in an LS1 intake are much better. Most dual plane manifolds have short sharp turns, but nowhere near your 135*.


4) Relating to #3, do I expand the cross-section of the tubing on the short side or the long side?

As you know,air has considerable mass so when it tries to turn it's inertia wants to take it straight, so it packs up against the long side. My guess is that locally there is a more dense charge at the outside of the turn and less dense on the short side. Again, look at a SuperVic runner. It's fluid dynamics.

5) Is a center tube of 3.5" x 18" a suitable size plenum for a 383cid engine when it's only 45% of the engine's displacement?

Depends on lots of things, IMO.




My $.02

Generally you want the runner to have a constant taper of 1 or 2 or ? degrees toward the head.

So, sort of like a long Tulip vase then.
I would assume that the end that meets the head should be the same size as the intake port, or very slightly smaller.
I'm guessing that the very slight taper you refer to is meant to help accelerate the air as it moves toward the head. Would that be correct?
I also notice that in the calculator section there is an intake runner calculator and it refers to 2nd, 3rd & 4th harmonics.
Would you be kind enough to try to explain to me what that means in practical terms?
Also, am I totally off the wall on my venturi idea?


BTW, as to the operating range, I figure that a relatively low torque peak would serve me best. Probably somewhere in the 4000 - 4500rpm area. I am far more interested in:
a) keeping the torque curve as flat as possible, and;
b) trying to get as much area under the curve as I can mange.



In terms of the design idea, I was looking at an LS-X manifold, or rather a picture of one, and I was struck by the similarities between my concept and the LS design. It was strictly unintentional, but it is noticeable.

The reason I was looking to incorporate curves is because having curves in the runner allows me to get more length in the runner without having to go WAY above the hoodline or out over the valve covers. I will be disappointed if I can't get at least a foot of runner length.


BTW, is there anyone in the NY/NJ area, or in the Ashtabula, Ohio area that would be willing to take a look at my design once I get a mockup?

Originally posted by LameRandomName
In terms of the design idea, I was looking at an LS-X manifold, or rather a picture of one, and I was struck by the similarities between my concept and the LS design. It was strictly unintentional, but it is noticeable.
That is what I pictured when I was reading your post.

Why can't somebody bring out a manifold like the LS6's, but designed for an LT1? :(

Originally posted by AdioSS
That is what I pictured when I was reading your post.

Why can't somebody bring out a manifold like the LS6's, but designed for an LT1? :(

Well, if my ideas work out I'll be happy to churn out a few.

I think I'd like to build it in a semi-modular way so that runners and/or plenum can be switched out.


BTW, if anyone has an LT1 manifold that they aren't using, I need one to destroy for R&D purposes.

If i donate my intake manifold do i get something in return, If things go well? J/W

Originally posted by SSlammedlt-4
If i donate my intake manifold do i get something in return, If things go well? J/W


Well, if you agree to donate your manifold, understanding that it will get destroyed in the process, then I'll agree that if I ever manage to create a working and useful manifold that I will give you the second working model.

I also notice that in the calculator section there is an intake runner calculator and it refers to 2nd, 3rd & 4th harmonics.

when the intake valve closes, the column of air has considerable mass and velocity, which translates into inertia. the column of air doesn't just hit the closed valve and stack up. it's inertia 'bounces' it off the valve and sends the column of air back towards the plenum. this is known as a reflected wave. when the reflected wave arrives at the plenum, it collapses and moves back toward the valve, which is known as the returned wave. when this always moving column of air gets back to it's starting position, the cycle is referred to as a 1st-order harmonic. when it repeats the cycle again, you have a 2nd-order harmonic, and so on. the number of harmonic that will occur before the intake valve opens will depend mostly on the amount time between intake valve closing and opening and runner length(measured from the intake valve to the plenum). naturally the longer the runner, the more time required to achieve each harmonic, and of course the higher the engines rpm: the less time available to achieve a particular harmonic. with each harmonic cycle, the column of air gains mass but loses velocity. as the harmonic number increases, you will lose some cylinder-filling efficiency at the tuned rpm, but gain some on either side of that desired rpm. 1st- and 2nd-order harmonics tend to have very good V.E. at or near the desired rpm, but have a sharp drop-off outside of a narrow band. thus they tend to produce high peak torque but have a "peaky" torque curve. 3rd- and 4th-order harmonics tend to trade peak V.E. for a broader (flatter) torque curve.

for practical purposes, 1st-order harmonic are not really achievable as they require very short runners and very high engine speeds. 2nd-order harmonics are generally achieved with single-plane manifolds and a lot of intake lobe duration(shorter interval between the valve closing and opening). 3rd- and 4th-order harmonics are what dual-plane and tpi manifolds are designed for (with the appropriate cam selection).

Ahhhh.... I see. Thank you.

Originally posted by LameRandomName
I'm picturing a square tube coming almost straight up from the intake port, with a slight narrowing in two opposing tube walls shortly before the port, in a sort of hourglass shape.
My thought is that this narrowing would have a venturi effect, "narrowing" and accelerating the incoming air charge just before it enters the head.

If you're looking to put in an hour glass shape, then you're going to have the air just slow down again as the hour glass widens out. However, maby (just guessing) it would have some benefit to air/fuel mixing, if the injectors were before the venturis.

BTW, some aftermarket intakes are being made with plastic. I saw a 3-piece design in the back of a magazine. Very simple concept. I think it would actually be possible to make one of those, using something like clay for the molds and then vacuum-forming the runners, like some kids made stuff in high-school shop class, but with thicker plastic. Then, just find some thicker plastic for the header flanges, injector bosses, TB flange, etc.

Originally posted by mnorwood
If you're looking to put in an hour glass shape, then you're going to have the air just slow down again as the hour glass widens out. However, maby (just guessing) it would have some benefit to air/fuel mixing, if the injectors were before the venturis.


I'll be out in Ohio in the near future. I used to work in the testing lab at a place out there in Ravenna called Allen Aircraft. They make mostly fuel pumps and valves.

Anyhow, the old man knows a whole lot about fluid dynamics, and I mean in an old-timer way.

He used to win those mileage races...
Anyone remember those? You get a small amount of gasoline and whoever makes it farthest wins...?
Well, he used to win those races with a dual quad corvette.

I'm going to stop in there and talk to him. Maybe talk to the engineers too.

I know that once air passes through a restriction in an otherwise constant sized tube, it actually continues to accelerate for a short distance.

I don't really understand why, and I know longer remember where I learned that. I only remember it being presented as an interesting oddity.

I think it had something to do with the vortices on the downstream side, that are created by the air passing through the restriction.

Originally posted by LameRandomName

Anyhow, the old man knows a whole lot about fluid dynamics, and I mean in an old-timer way.

He used to win those mileage races...
Anyone remember those? You get a small amount of gasoline and whoever makes it farthest wins...?
Well, he used to win those races with a dual quad corvette.



Ask him if he cheated...and how.

Ask about a well hidden small tank about the size of a hand-held propane cylinder with some gas in it and then pressurized with air with a solenoid valve and a small wire-sized line that lead to the intake. Maybe it used the horn ring to actuate the solenoid. I think that might work. See if that gets a smile from him. :)

I find it amusing, to see it stated, that following specific 'rules' is required/recommended to achieve a satisfactory end result with sheet metal manifolds (which I agree with), yet this forum embraces/accepts/tolerates whatever rules breaking 'butchery' is necessary to get an LT4 intake to conform to LT4 head ports. :lol:

OLD -

Actually, I've already had that conversation with him. Assuming that he wasn't BS'ing me, he didn't use that trick.

He wouldn't tell me how he did it, but he did give me a hint; apparently a carb'd engine gets the best mileage at low revs with the throttle wide open.
I asked him how you could keep the throttle open the whole race without accellerating to top speed and wasting gas.
He smiled and said; "Well that's the trick, isn't it?"

Originally posted by LameRandomName
OLD -

Actually, I've already had that conversation with him. Assuming that he wasn't BS'ing me, he didn't use that trick.

He wouldn't tell me how he did it, but he did give me a hint; apparently a carb'd engine gets the best mileage at low revs with the throttle wide open.

Well maybe not wide open, but close. One of the way to achieve that is with a very (numerically) low final drive ratio and small primary carb bores. Think Quadrajet. Before OD automatics, some full size cars had axle ratios as low as 2.29 although 2.41 was more popular. A 3.23 with a .7 OD is 2.26. A 3.23 with .5 OD is 1.62 final drive.

How about if that Corvette had two small (500 cfm or less) 4 bbls with modified progressive linkage so only the primaries on the rear worked until almost WOT. Combine that with a very low axle ratio, a wide ratio trans (maybe even an OD!) to get the thing up to speed, very high compression ratio and optimum timing for the low rpm, low manifold vacuum, large throttle opening conditions and you might get there.


I asked him how you could keep the throttle open the whole race without accellerating to top speed and wasting gas.
He smiled and said; "Well that's the trick, isn't it?"

A very slick progressive linkage perhaps?

I think he used 390cfm holleys, but I cannot swear to that.

LOL. That reminds me. Bought one a couple of decades ago for mileage purposes. Didn't get around to using it....:shaking head:......:shrug:

Just wanna say there is some very good info on here.. Keep it up.

Peace,

Originally posted by LameRandomName

I know that once air passes through a restriction in an otherwise constant sized tube, it actually continues to accelerate for a short distance.

I don't really understand why, and I know longer remember where I learned that. I only remember it being presented as an interesting oddity.

I think it had something to do with the vortices on the downstream side, that are created by the air passing through the restriction.
I know what you are talking about, but the "mild hourglass" shape you originally suggested would allow the airflow to not do what you are talking about. An orifice, or a half of a venturi with an abrupt enlargement after the throat, would accomplish that better.

The turbulance would reduce airflow; in other words, the vortices you mentioned are the highly turbulent region that will greatly increase friction.

Consider this: you could accomplish the same thing by having the ports on the intake manifold smaller than the ports on the heads. That is totally contrary to port-matching, which is done to improve performance.

I think what you may be after it that transient affect whereby the intake valve closes, and the fast-moving air comes to a sudden halt; the rapid stop causes the pressure to spike (kenetic energy converts to pressure; or the air in back runs into the air in front, same thing), and when the intake valve open again, all that high-pressure air is waiting to go. If that's what you're after, probably the basics (small port size, small long runners, give up power for torque, etc) will be the solution.

Sorry, I've been having computer trouble, so I made this a 2-part post.

But, seriously, if you're going to have your own intake built, don't go on some wild goose chase like venturis and air flow that doesn't touch the sides of the runners. Go for something tried and true, and totally kicka$$, but expensive, like variable-length intake runners.

Check out http://www.fuelairspark.com/Technical/Products/Manifolds/

Thta's a plastic manifold. See the velocity map (red, blue, green, yellow); the area where the red area (upper right) is next to the blue area below it is the area where the runners (red) are next to the plenum (blue); this would be an awesome area to add some butterfly valves (one per runner, 4 per side) to give you an intake with some serious low-end grunt and high-rpm breathability-plenty of R&D needed to get the geometry right, etc. Nissan and honda do it on production cars, and I've seen them as aftermarket parts on american V8s, but $$$.

If it's still on stands, check out the inside bak cover of "Best Of Tech", winter. It has a picture of the internals-very simple concept.

I don't know that I want to do something that complex, plus of course I don't have the engineering background or the testing equipment to do the proper R&D on that.

I'm looking more at doing something that will give me longer runners and a bigger plenum.

I understand what you're saying about my "hourglass" idea and why it wont work the way I was thinking it would.

I don't agree with the notion of having a mis-matched port either, although I think I don't like it for a different reason than you don't like it; because it's too close to the back of the valve at that point.

You mentione the notion of both a restrictor and a half-venturi.

What do you mean by "half" venturi?

I would picture a "full" venturi as being a shaped tube within, centered and parallel to the runner, with the only question being; do I want the venturi to have more surface area "inside" or "outside" the venturi? In other words, do I want the air to move faster toward the walls of the runner or toward the centerline?

Or for that matter, do I want have varying velocities at all?
Perhaps my entire concept of trying to speed the airflow just before the head is altogether worng in the first place and I would be better served with merely having a mild taper from the mouth of the runner (at the plenum to the exit (at the head).

Originally posted by LameRandomName

You mentione the notion of both a restrictor and a half-venturi.

What do you mean by "half" venturi?


A venturi is an hour glass shaped pipe. There is nothing inside the pipe-basically, a reducer and a diffuser (expander) back-to-back. This causes the air to speed up at the neck (the narrow part). In a properly designed venturi, the expander expands at about the same rate that the air naturally wants to expand.

A half-venturi, in my mind, would be a reducer between two equal-sized pipes, so the air goes fast at the narrow part, and then abruptly enters the larger pipe (the same size as the inlet to the half-venturi.) This way, you would have high flow in the center much faster than you would normally have for a pipe of that size. The thing is, turbulance (and thus drag) would be higher than just going into a smaller pipe.

Originally posted by LameRandomName

Or for that matter, do I want have varying velocities at all?
Perhaps my entire concept of trying to speed the airflow just before the head is altogether worng in the first place and I would be better served with merely having a mild taper from the mouth of the runner (at the plenum to the exit (at the head).
Tapered runners seem to be used alot, based on what little I've seen. I'd go with that. How long do you want to go with your runners? Any performance goals yet (or, did you already post them?)

I sort of posted them, but I never got specific.

Something relatively mild. Something with peak power at about 6000-6500, with a readline at 7000.
I'm thinking long rod 383 with appropriate heads and cam and extensive use of coatings, with this manifold I'm planning to build.

An orifice would create a fairly large pressure drop, so that would be a no-go, same as a sharp edged transition like a venturi with the diffuser chopped off. The loss coefficient (Ka) is the measure of how efficient transitions are. For instance, the difference between a 10 deg taper and a 180 angle (sharp corner) is .05 compared to about .25 (factor of five) for an area change of 2:1. The orifice is basically the 180 deg, but the effective area of the hole is even smalled due to the vena contracta. With the venturi, you need the diffuser to effectively recover the pressure. I'm not sure how a venturi would be beneficial. They are never 100% efficient, so you'd have frictional losses, without a real benefit. As was mentioned, the center of the venturi has faster air with a lower pressure. As soon as the area diffuses back to the port size in the head, the speed drops back down, and pressure is back up, but you've just wasted a bit of energy making the air change direction. If you've got the air separating from the wall with large sharp edge transitions, then you're killing the flow efficiency. The basic fluid dynamics concepts still apply - smooth transitions, radii as large as possible, etc. There are many opinions on plenum size, but runners seem to be a compromise between runner length and volume, and the Helmholtz harmonic tuning is proven (length from back of valve to plenum). This is a good discussion...

OK< I'm going to scrap the "hourglass"/venturi idea and stick to making the smoothest turning pipe I can.


Question...
How does the air behave when travelling through a tube of gently decreasing cross-sectional area?

Does it speed up? Slow down?

What about the pressure?

Bernoulli's equation states that as velocity increases, pressure drops (air moving over the top of an airplane wing is going faster than the air below it, creating low pressure on top, sucking the wing up). Air moves from high pressure (atmosphere) to low pressure (combustion chamber).

Moving air going to a smaller area must speed up (think of putting your thumb over the garden hose - the water speeds up to get past the small opening). Since it speeds up, pressure drops.

Keeping pressure high will force more air into the chamber (and more fuel = larger bang, for more torque & HP). Since air has mass, you can't just have huge 3" dia runners, since that whole mass of air has to get moving before it gets into the cylinders. Likewise, having tiny 1/2" dia runners would be very restrictive. Once you choose your heads, you have the end dimensions for the runner (area), and now you just have to choose length and diameter at the inlet, along with plenum size. The rest of the design is just efficient packaging for the most part. I'd suggest getting Engine Analyzer Pro software to play with various runner & plenum sizes, which is a lot easier for trial and error compared to cutting & welding aluminum :)

You've got a point.

Say, on a related but different question...

If someone were to make a cold air intake (this happens to be for an LT1) where the diameter of the "main tube" is larger than either the MAF or the TB openings, how would that affect airflow into the engine?
(In this case I am talking about a 4" diameter duct.)

Would it behave as a sort of secondary plenum area or expansion chamber, and help the improve the airflow into the manifold and eventually the engine, or would the expansion of air into the "chamber" end up reducing airflow by forcing it to expand and then contract again?

Yeah, that would end up reducing airflow by causing extra expansions and contractions. It would be big, too. The only benefit I can think of is that adjacent cylinders that fire one after the other can rob each other of air (the first robs the second), and by using your idea maby you could prevent that somehow. That would probably take more R&D than the variable length runners.

If you want power in the 6000-6500 rpm range, you basically want short runners. Good engineers develope good products; great engineers copy good engineers. If you want to design something you should look at something similar. Chech out other manifolds built for 6000 to 6500-rpm peak power in a 383, and start from there.

Z28tt: :cool:

Actually, that last question was in regards to my motor as it sits NOW.

I was playing around with different idea for a better intake system and I bought myself a TB boot from the '95 LT1 Vette.
Now I'm trying to come up with a hardware store solution for the rest of the hookup.