Conventional wisdom is that long intake runners help with low end torque, while short intake runners are better for high RPM engines. People have told me that is true for both carbureted engines and the new EFI engines.

I don't see it. Can anyone tell me why a long tube is better for moving air into a cylinder than a short tube? Also, what is long? 3", 6", or more? Is it better to be a straight tube, or are bends allowed?

Thanks -

 

It's got to do with the pulsing of the air - when the cylinder takes in air, there is a vacuum made in the runner. At the right RPM, with a long enough runner, those pulses will actually push more air into the cyllinder. The longer the runner, the lower the RPM it happens at.

Bends are not good for air-flow, and for carb'd motors bends will cause the fuel to fall out of the mix.

Look at any modern motor and the intake will have matched length runners, and if it's a truck motor (like the V10), it'll have longer runners than a passenger car motor.

For carb'd motors, it's almost impossible to match the length of the runners, and they may not be as long as you would like for ultimate low-end torque. EFI makes it easier to match the length of the runners and elongate them somewhat.

Ever see four downdraft Webers on a SB Ford? What a sight!

Same works with headers - the longer the tube, the lower the RPM where they reach peak effficiency.

 

I believe it is a combination of runner length and size. A longer/thinner runner will increase intake charge velocity at lower rpm, creating better throttle response, fuel atomization and cylinder filling. A shorter/bigger runner will have poorer low end performance because of it's lower intake charge speed, but will allow more fuel/air, and increase cylinder filling at higher rpm.

You can also have long/big/straight runners, as in a tunnel ram, or cross ram. Their purpose is to create enough inertia in the intake tract to "pack" or "ram" air into the cylinder at high rpm. Kind of an inertial supercharger.

"Long" and "short" are relative to the diameter of the runner. Typically it is sufficient to consider a dual plane manifold, like the performer, to be long. Dual plane means that there are actually to plenums, at different levels, fed by the carb. This is done to increase runner length, and keep them roughly the same. Single plane manifolds, like the victor, feed all cylinders from a single large plenum, through roughly equal length large diameter runners. Because each runner connects to the plenum at it's closest point their length is limited, and would be considered short.

-Scouder

 

That's one thing I forgot to mention - the diameter of the runner. Smaller, lower RPM, bigger, higher RPM - like the difference between the Edelbrock RPM and Performer intakes. Same length, probably, but different diameter (cross section).

You're right about the high-rise intakes, longer runners, you get that bam-bam-bam effect, and it's an "intertial supercharger"...

You can actually get more than 100% volumetric efficiency if you can keep it in the right RPM range.

 

What about the fact that with gases and liquids a bend will increase the velocity of movement? I understand the effect it could have on a mixture where particles are suspended, but curves in headers for example will increase the velocity just as the air over the wing of a plane. Yes? No?

 

Could be. I would bet that any velocity increase is very localized, right at the bend. Also, on the short side of the bend, fuel is going to tend to drop out of suspension, just like sand in a creek makes a sandbar on the inside of a bend. In a perfect world, all intake runners would be perfectly strait, and change size with engine rpm.

In an exhaust header, the hot gases don't seem to care that much about bends, but are still critically effected by length and diameter.

-Scouder

 

Thanks. guys, as usual you bring up good points and send me in the right direction. I have studied the stock dual plane 4 barrel manifold I have loose here in the shop and then drew up some pictures to aid my understanding.

I don't know why I am always surprised at how well engineered Ford parts are, but I am. Somebody put a lot of thought into this manifold. All runners are the same size and length, as near as I can see. The runners actually run between two cylinders (one on the right bank and one on the left bank,) with a center connection to the plenum.

The front runner of the upper plenum connects 1 and 6, while the rear runner connects cylinders 4 and 7. I did a timing diagram and found that the cylinders connected to the same runner are always separated by 180 degrees, and none of the 4 cylinders connected to a plenum are ever in their intake cycle at the same time.

The lower plenum works exactly the same, except the front runner connects 2 and 5, and the rear runner connects 3 and 8. The two plenums overlap a bit - they are separated by only 90 degrees.

This means that the instantaneous air demand in a plenum is never more than one cylinder's worth, and even above the plenums, past the throttle body, at most your demand will be somewhat less than 2 cylinders' worth - one cylinder has its intake valve closing while the other has its opening. These numbers work out as they show a 390 at 6k RPM needing a bit less than 650 cubic feet per min.

I am asking these questions as I am trying to determine if a FE manifold will be effective when converted to port injection for a sequential EFI system.

 

If I were doing port injection I would still match the manifold to the RPM range of the engine, but wouldn't worry about anything else. If anything, I might err a little on the oversize, but not to extreme, since the port injection will eliminate almost all of the negatives of an oversize plenum.

Now that I have written that, I second-guess myself. Now that your manifold runners are no longer carrying air and fuel, but just air, I wonder if that will make the manifold seem larger to the engine? Would that mean you could err on the small size, and still get decent upper RPM performance?

Hmmmmm.

-Scouder

 

Yes, lots of unknowns. The silly thing that worries me is the timing of the squirt from the injectors and the unequal mixing of fuel to air. As you know, the pressure on the gas is held steady, and the computer adjusts the time that the injector is open to produce the supposedly correct fuel to air mixture. The injector pumps the same amount of fuel per unit of time whenever it is turned on.

But consider the incoming air - the valves starts to open, ramps to a max lift point, and then quickly closes. The air flow starts slow, reaches a max and then drops to a stop. Perhaps a bit more air flows in as the valve closes due to volocity forcing it in.

But whatever happens you are trying to match two unequal things - the steady flow of fuel to a weird shaped pulse of air. It is even worse when you consider that the computer will likely shut off the flow of fuel at some point before the valve closes.

It seems to me that this is a formula for unequal mixing of the fuel to air mixture in the combustion chamber, which could add a lot of stress on your piston and could also form hot and cold spots on its top.

With a conventional carb fuel is added at a steady rate to the incoming air stream by the ventures. If the air slows down, less gas is added. If it moves faster, more air is added. Plus we take care to insure some turbulent air flow to help mix the air and fuel. None of this is true with port injection, or, at least, not much of it is true.

I have the source code for the computer and I plan (as of now) to rewrite it such that the injectors are turned on for a period centered around the max lift point, depending on the geometry of the cam. But maybe it will be better to turn the injectors on when the valve opens and hope that any fuel still not mixed with the incoming air when the injector closes will do so during the rest of the cycle. I am open to suggestions here.

I keep telling myself that sequential port injection must work or Ford would not have changed its entire model line to it, but I still have my doubts.

 

OK. Your'e gonna make me disagree with you on almost everything you said.

I understand your concern, and it is valid. Maximizing EFI efficiency occupies some of the best engineering minds in the country, believe me.

Let me start with the statement you made about carburetors. Fel is NOT added at a steady rate except in theory. And since most of us don't have the expertise, time or resources, they are never tuned to add fuel at anything approaching the CORRECT rate at each given load/throttle position. Consider the transition from each fuel circuit to the next in a carbureted application and it is easy to see how brutish, it is. On top of that, add a large duration cam at low rpm and watch how escaped cylinder pressure (reversion) actually REVERSES the flow through the venturis! We do agree that we take care to have turbulence in the intake runners to mix the air/fuel, but the important question is, why? The fuel/air was mixed in the venturi wasn't it? Ahah! You and I see the same thing two different ways. You see the long trip from venturi to chamber as an opportunity to further mix the air/fuel charge. I see it as a necessity to RE-mix the air fuel charge because air and fuel do not stay mixed. As soon as it leaves the venturi it begins to separate, increase droplet size, and drop out of suspension. Now what happens when this pre-mixed charge when it slams up against the back of the closed intake valve? Nothing good, in my mind. Fotunately we agree on something else: "None of this is true with port injection, or, at least, not much of it is true."

As to sequential injection. It is the best we have. It eliminates virtually all of the problems of carburetion. It is smart, it reacts to changes in very small fractions of a second, and it is virtually immune to the problems caused by low rpm air flow. The fact that the biggest concern you have is WHEN to inject the fuel in the intake window is a testiment to the efficiency of the system.

I have heard that the injection pulse should happen late in the intake window. I think that would correspond with your plan to time the pulse with the peak flow. But consider that most cams hold the valve open from seat to seat between .0036 and .0042 seconds at 6000rpm, we are talking some pretty fine timing. Regardless of the timing of the pulse, it will be well atomized, better than a carburetor for sure, just by the nature of the system.

I too, will have a SEFI system on my 511 at some point. I am looking at the MegaSquirt II because of cost. No more cold start issues, no more floats, jets, power valves, accelerator pumps and cams, idle bleeds, high angle flooding, etc... Now remind me, what was good about carburetors again?

-Scouder

 

Cost. That's the only reason... However, some would argue against using a computer of any type to control an FE

I didn't take a lot of time to read all of you guy's responses, but did get the major portion. An air/fuel mix going around a bend will, as Scouder(?) said drop some gas out of suspension on the inside of the bend just like a creek, but even more important, as the mix is going around the bend, the gas droplets are heavier than the air and will not all make it around the bend but slam into the side, pooling there (atomization is NOT vaporization)

As for headers, the less bends the better - even air has some inertia, and when you push it through a tube and then change direction abruptly, the molecules actually want to slam into the outside of the bend - when they do, they slow down limiting the flow.

And if you're doing any high-angle travel with EFI - make sure your gas tank is full! (and you're oil pan has the appropriate baffles and pickup tube) (and... the list goes on and on...)

 

I like engineering - facts are facts and opinions just point out where you need to gather new facts. We don't disagree - you are just ahead of me in my thinking. I was looking at steady state cruising only. I agree with SEFI.

I too am looking at MegaSquirt II, partly do to cost but mostly because they give out their source code so you can modify it if you want. One guy I talked to added almost 2 MPG to his 302 by modifying his MegaSquirt I to run extra lean when cruising.

I agree 3 to 4 milliseconds is fast for a mechanical system, but that is a long time for computer controls. The computer you are reading this on is likely many thousands of times faster than that. You should be able to recalculate everything and change the injector squirt and the timing spark in the time it takes to go from one cylinder to the next in the firing order.

I went on the MegaSquirt site and read their "Success Stories" for Fords - Every one that they listed started with an EFI manifold - not a single one of them built their own. This is why I am looking so hard at FE manifolds.

 

I don't see why a Performer RPM wouldn't work perfectly with multi-port fuel injection if you have the manifold drilled for the injectors. I think the only requirement would be to make sure the injector is spraying in the general direction of the intake valve - I think someone in this forum mentioned an intake that had the bosses for nitrous injectors - might work with a small-diameter injector?

Does the MegaSquirt include a throttle body? What type? Holley 4-bbl flange?

All this talk has me thinking about my dream project of putting a Ford modular or SB into a small English sports car (TR7)... only downfall is the 150+ lbs of concrete in the trunk to make it turn corners somewhat OK

I'd love to engineer my own injection system - two clustered computers, two sets of injector circuitry - two crank position sensors, everything double.

 

MegaSquirt is the computer to control the EFI system. You have to find or fab the mechanical parts you need. The current system comes in kit form and you have to build the computer. It is also batch fired. But it costs only about $200 and does work.

They have a much faster computer (actually an ECU or engine control unit) in beta testing. This one does sequential port injection, has ignition timing and knock sensing.

A good example and a before/after charts can be seen at http://home.comcast.net/%7Eqcrispin/...Megasquirt.htm This example is a 302 running the original batch firing system with no ignition control.

 

I'm just going by memory here, but it seems to me that all of the EFI manifolds I have seen are just single plane manifolds. Nothing special. That gives me hope. I am operating under the assumption that the reason they offer EFI manifolds is just so the average guy can convert to EFI in his garage over a weekend, not because there is anything special about them. Other than placing the bungs, it doesn't look like converting a wet manifold to SEFI would be any big deal. Kinda makes you wonder if those old Streetmaster single planes are gonna get valuable again, doesn't it?

As to speed. Yep. I would bet that the system is able to recompute thousands of times per second. In essence, it is sitting around yawning, waiting for the next time it can fire the injector.

I am going to research the injector timing in relation to the intake valve and post what I find.

-Scouder