designing and building manifolds.

[autronicDSM]

does it matter if all runners are the same length?

i was thinking about making mine similar to magnus manifold but with throttle body mounted sort of on a corner. similar to skyline manifold or some audis.

now, i think how it narrows down toward #1 helps to get more air into that cylinder since 1 or 2 are known to run leaner. just my $0.02

i'll post a pic soon as i figure out how to get around a firewall at work

 

[JET]

I made a quick sketch of my manifold idea, but I couldn't post it on here. If you would like it, please e-mail me and I will send you a copy.

 

[JET]

I made a quick sketch of my manifold idea, but I couldn't post it on here. If you would like it, please e-mail me and I will send you a copy.

 

[autronicDSM]

You must be logged in to view this image or video.

here's sketch of my idea, any feedback appreciated.

 

[JET]

You manifold narrows toward #1 like the Magnus. #1 is also known to run hotter, so I think this is a bad idea. This is a better idea than the stock one though and I am sure it will make an improvement. I think it is even a better idea than the Magnus.

Which program did you use to get the sketch to post? Mine came out a bit map image and I couldn't get it converted or onto the board.

 

[marshall]

Originally posted by autronicDSM
does it matter if all runners are the same length?

if you vary the lengths of the runners, the amount of air reaching each cylinder will differ.

marshall

 

[autronicDSM]

Originally posted by JET
You manifold narrows toward #1 like the Magnus. #1 is also known to run hotter, so I think this is a bad idea. This is a better idea than the stock one though and I am sure it will make an improvement. I think it is even a better idea than the Magnus.

Which program did you use to get the sketch to post? Mine came out a bit map image and I couldn't get it converted or onto the board.

i just used PAINT program in accessories menu in windows to make a quick sketch. i think reason for it to narrow down like that is to have better spoolup. if it was squared like bj's manifold, i would think it would lag more since it takes more air to pressurize plenum. i'm not sure what difference that makes though. most noticeable would be down in low rpms...i think

this design was just an idea i had, i could build few different mockup manifolds and flow bench them...soon as i find a machine shop that culd do it. otherwise i would have to dyno my car with each manifold, but that's little too much pita.

 

[autronicDSM]

Originally posted by marshall


if you vary the lengths of the runners, the amount of air reaching each cylinder will differ.

marshall

thanks. i thought that would matter off boost only. once everything is pressurized i don't think it would matter as much. but i could be wrong. that question actually applied more to JET's design of 1 to 2 to 4 design using collectors. on a manifold i would beuild runners would be the same length

 

[JET]

Ok, got it figured out. Here is my idea:

 

[JET]

The problem we are trying to figure out is getting the same amount of air through all 4 runners. If you get 30% thru #1, 25% thru #2 & 3, and 20% through #4, that is bad! We want a design that will give all of the cylinders the same amount of air so that none of them run leaner than the others.

 

[autronicDSM]

that looks really good but is there anough space to make sure runners are samelengths and that there's nough space?

i would like find a flow bench in my area and play around with my design, perhaps varying runner length would help even out the flow? i'm not sure, what are your thoughts?

i just got another idea that i will sketch and post when i get home.

 

[GalantVR41062]

I thought the forrester cast manifold had ever flow to all runners? Or do 4indavidual runners going to the TB,get a Merge Collector form BernsSS and run 2" runners, equal length, then there would be equal flow to each runner for sure.
http://www.burnsstainless.com/MergeCollectors/mergecollectors.html

~John
4 TB? IC piping would get interesting.

 

[autronicDSM]

or 4 individual throttle bodies with air horns (velocity stack?) here's some pics for ideas
http://www.twminduction.com/ThrottleBody/ThrottleBody2000.html

mount air horn or velocity stack between TB and plenum and i think you're golden?! i could be wrong but seems like this would be the best solution for even flow. for plenum just use 3" pipe and you'll have killer spoolup too

 

[JET]

Runner length will not help even out the flow. After talking with an engineer a while back we decided that you do need a plenum of some sort to even out the inevitable variances. I would really like to use pipe, but trying to weld runners onto a pipe is difficult. I also kicked around the 4 runner idea, but decided that would take too much space. I will eventually try my idea, probably this summer.

 

[autronicDSM]

so i guess best solution is to have 4 TB or play around with a position of a TB on plenum like marshall did. i have one idea off his design that i will post later. this is geting interesting

as for welding runners to a pipe, you would need a flange > 1/2" thick and machine it to match the radius of the pipe and weld it on

 

[mrmadmax]

Have you guys read up on this subject? There are a few books that mention the calculations plenum volume, runner length, plenum inlet considerations, ect. The problem is, I can't seem to find anything good that refers specifically to turbo intakes.

I've done some research for my senior project in school on N/A manifolds, and the results were amazing. Has anyone got anything built yet?

 

[autronicDSM]

i haven't really read up anything on it other what i got in college physics about air flow etc. and talking with some guys who build hot rods and bikes and just reading different forums and such. since i could have access to fab lab at school i was thinking about doing it just to see what end result i'll have, i don't mind building manifold with TB relocated but i'd like to try one with individual TBs and velocity stacks except the cost might be high.

can you recommend some reference books, even if it's for n/a?

 

[mrmadmax]

Well, depending on your level of understanding of engineering stuff (or your patience level), for a simpler explaination of intakes, pressure waves, etc; take a look at Advanced Engine Technology by Heinz Heisler, or for more detail, look at Introduction to Internal Combustion Engines by Richard Stone.

Funny, I just noticed the "Advanced Engine..." title gives a basic explaination, and the "Introduction..." title is harder to understand; go figure...

With these two books (and some fab skills) I was able to design and actually build some intake designs that worked very well at whatever RPM I was building them for.

You might want to see if you can check them out at a library before you buy them. They were kinda spendy...

 

[autronicDSM]

JET, how about something like this? just like marshall's design except TB faces intake ports instead of being on top. let me know how what you guys think

You must be logged in to view this image or video.

 

[JET]

That is probably the best design you are going to get without having 2 runners feed into the plenum. If you want to keep it more simple and mount the TB on the plenum that is probably the best design. You are going to be getting a some more flow into #2 and #3 because that is where the air coming out of the TB will be pointed.
This is almost exactly the same design I had before I decided to go with feeding 2 runners in. It is going to be nearly impossible to get even flow to all 4 runners with the air coming in at one spot. Having 2 runners come in is the simplest way I can think to do it. One runner between 3&4 so they split the air coming in and the same between 1&2.
Hmmm....How about making a little change like this? This is going to be very difficult to get right though because there are going to be vortices coming off the ends of the splitter in the middle and it will also impede flow a tad, but should get the air moving in the right direction.

 

[autronicDSM]

only one way to find out...somebody has to build it. i will start looking for some scrap metal i can use to build this and then try to get this thing flowbenched. this should turn into interesting project

i still want to try 4 TB with velocity stacks though...

 

[mrmadmax]

Originally posted by JET
...You are going to be getting a some more flow into #2 and #3 because that is where the air coming out of the TB will be pointed.
This is almost exactly the same design I had before I decided to go with feeding 2 runners in. It is going to be nearly impossible to get even flow to all 4 runners with the air coming in at one spot...

That's not exactly correct. The problem here, gents, is that you're all thinking about constant laminar flow of air; which is somewhat flawed. An intake manifold (hell, the entire engine for that matter) behaves more like a musical instrument than a bunch of pipes with air flowing through them. This happens because a single cylinder's intake stroke occurs for just a short time, then the valve is closed and air stops flowing. It then starts again, and stops again... and so on.

I will say that you are correct about using "velocity stacks" to take advantage of the bell mouth in the entrance to the runner. Doing so will reduce the narrowing effect (vena contracta) of the runner entrance due to the square corners.

The best way to even out the flow to each cylinder is by:

1) Provide enough plenum volume so that each cyl can draw in air with minimum local pressure effects.

2) Make the entrance to the plenum from the throttle body smooth (another bellmouth) and large enough (much larger than the runners) so that the velocity of the air entering the plenum is low and also does not cause local pressure effects in the plenum.

3) Keep the walls of the plenum at least 1 x runner diameter away from the runners.

Take a look in those books posted above for an explaination of the pressure wave and resonance theories on intakes.

There are people that spend 20+ years of their life trying to understand exactly what is going on with fluid dynamics, so I can't spell it all out in one post.... but maybe this will help you guys some.

 

[JET]

I think you are forgetting that there is a large amount of flow. Yes there are slight pauses, but at 7000 rpm you are taking air in 67 times per second! You also have an intake cam duration of 260+ degrees and are intaking air every 180 degrees so there is always a cylinder taking in air and sometimes 2. The higher the RPM, the less effect the pressure waves will have because they will have less time to impact the airflow. It will create pressure waves for sure, but without computer simulation we are not going to be able to map them.

Bell mouths are always good, they help eliminate pressure differentials around sharp serfaces (vortex). I always port out everything possible and create the bellmouth.

All molecules have weight and therefore they have inertia. They do not want to change direction unless they have to. If you have nothing to change the flow in that manifold design, #2 and #3 are going to get more air.

I worked for 2 years with zinc injection molding. We had to create vents out the side of the dies to get the zinc back there. This is exactly the same except for the pressure waves and the main thing that will affect that is runner length. You are also always going to have dead spots as long as you have sharp corners and this impedes airflow around the dead spots (the sides of the manifold should be a ways away from the runners.

Does anyone know if a flow is even made that will measure 4 different ouputs at once? It would be EXTREMELY interesting to flow bench an intake and be able to see how much air was actually moving though each runner and not just a number that was the sum of all 4 runners.

 

[mrmadmax]

Originally posted by JET
I think you are forgetting that there is a large amount of flow. Yes there are slight pauses, but at 7000 rpm you are taking air in 67 times per second! You also have an intake cam duration of 260+ degrees and are intaking air every 180 degrees so there is always a cylinder taking in air and sometimes 2. The higher the RPM, the less effect the pressure waves will have because they will have less time to impact the airflow. It will create pressure waves for sure, but without computer simulation we are not going to be able to map them...

...All molecules have weight and therefore they have inertia. They do not want to change direction unless they have to. If you have nothing to change the flow in that manifold design, #2 and #3 are going to get more air.

I worked for 2 years with zinc injection molding. We had to create vents out the side of the dies to get the zinc back there. This is exactly the same except for the pressure waves and the main thing that will affect that is runner length. You are also always going to have dead spots as long as you have sharp corners and this impedes airflow around the dead spots (the sides of the manifold should be a ways away from the runners.

Does anyone know if a flow is even made that will measure 4 different ouputs at once? It would be EXTREMELY interesting to flow bench an intake and be able to see how much air was actually moving though each runner and not just a number that was the sum of all 4 runners.

Jet,
First off I'll start by saying that nothing I say below is intended to be offensive, I'm just trying to participate in some healthy discussion of the subject to improve all of our understanding of the topic, including mine. I think this is a really interesting topic, and I like to hear people's opinions.

Regarding the flow volume, I realize exactly how much flow we are dealing with, as this was the subject I studied at great length in my mechanical engineering senior project. I would not, however, call the "pauses" in air flow 'slight'. You have to consider each individual runner as it's own entity drawing air in a cycle from the plenum. When this happens, there are pressure waves that travel back and fourth in the runner, which affects the air pressure in the plenum near the runner ("local effects"). Yes, it is true that the pulses overlap each other, having the effect of damping the pulses seen by the entrance to the plenum, but that's not the point. The pressure wave effects are taken advantage of at the runners. It is the timing of this wave that is adjusted with the runner length.

One of the points of my other post was to say that a large entrance to the plenum will slow the air entering and decrease the "inertia" that makes it tend to keep going toward cyls 2&3.

As for the zinc injection molding, we are comparing apples to oranges here... Air is compressible; Zinc is not (for all practical purposes). Injection molds don't have an outlet, hence the need for the vent; Manifolds have an outlet--into the cylinders. The Zinc flow is constant; Air flow is periodic (pulsed). Zinc is quite viscous; Air is much less viscous. Temperatures, densities.... all very different.

As for the computer simulation, I have Star CD Computational Fluid Dynamics in the very lab I am writing this from, but due to the complexity of the software, I have chosen to rely on some simple hand calculations for intakes, and I've had pretty good sucess when verified on the dyno, then further using pressure transducers at points in the manifold.

For the flow measurements on a flow bench (although I feel such measurements would not be very realistic because the flow is not pulsed), you could run the air in through the throttle body and out all of the runners, then use a velocity probe to measure the flow in each runner. Velocity (ft/min) x Runner area (square ft) = volume flow rate (CFM).

 

[autronicDSM]

correct me if i'm wrong but it shouldn't matter that flow is constant or pulsed since we would like to see if there's difference in flow to each runner? just my $0.02


also somebody said earlier that air entering TB flows less to #4 because it's so close. what if we make the plenum longer and offset it so TB is placed farther away from #4?