Designing an L6 Intake Manifold

[rejracer]

I'm trying to come up with a new plenum design on an L6. I have this concept that it's the length of the L motors that makes the plenums hard to design for. As a result I am trying to get the plenum output ports closer together by offsetting the plenum output ports by a set amount for all the ports. I also want all the ports identical. Here is version 1, which has issues with the runners clearing each other, so this particular design is not buildable.

 

I am in the process of moving the ports on the plenum around to make this possible.

 

The port design is intended to be a large velocity stack. The diameter is larger at the plenum side. I think they are 2.5" at the plenum and 1.5" at the flange.

 

When designing a velocity stack, what are the variables and formulas needed to calculate the optimal design for a given application (displacement, VE, RPM...)

 

Any help is appreciated.

 

..and the eyecandy

 

Clearance is an issue between 1-2 and 5-6. 3-4 cleared, but that was an easy fix by increasing the vertical offset.

 

This gives you a better feel for the diameter difference between the plenum output and the flange input diameters.

 

The design seemed simple, but the offset and diameter differences are shooting holes in this concept.

 

 

Thanks.

[johnc]

You also, somehow, need to get at the studs to put the nuts bolts, and fat washers on and off. And you need clearance for the header flange and primary tubes, steering shaft, and motor mount bracket.

[rejracer]

For the bolts I think they would clear, but if not a stud / nut combo would easily clear.

 

As far as header flange I think it's good as I made the flange based off dims from Ron Tyler. Clearing primary tubes, Good question! Aa far as the steering shaft and motor mount, the Plenum is 10" tall, so it hangs down 5" from perpendicular to flange port centerlines. Plenum base is 9" away from manifold mounting surface on head.

 

but here is some info on it:

 

Plenum outlet ID is 75mm.

Flange port ID is 38mm (1.5")

All ports have identical offsets. and design.

 

Here is the last version I had made up.

 

I did not like the layout of the previous, so I rearranced them. then I discover I don't like this very much... so back to the drawing board..

 

 

I ended up reversing ports 3 and 4 on the Plenum. I also reduced the port entrance diameter to 50mm.

I rearranged the ports to make the overall design smaller. Next I will try and increase the offset.

 

I changed the port offset again, Increased the port entrance diameter to 55mm, increased the plenum offset to 10". This latest version the plenum base is 8" tall.

[rejracer]

I decided the stubby concept is not the most ideal design. It's doable, but it's not turning out like I had hoped. The biggest problem is the volume in the plenum, I thought it would slow the buildup of pressure, and thus response.

 

So here is the new concept. It uses a splayed runner layout to get the runner diameters where I want them. I have yet to design the plenum, but it will be a log design with ports going out to each runner. I am splaying the runners so they will clear one another. Both the forward slant and the cone/velocity stack design aspects cause interference between the runners.

 

 

The purpose of this exercise is to come up with designs that will mitigate the efficiency lost as a result of the different velocities between the plenum and runners. If it's possible to lower the rate of change in velocity between the plenum and runner, then I believe the overall design will be more efficient. The my current understanding is the different velocities are caused by 2 design parameters. 1. The diameter difference between the runner entry and the plenum. 2. The near 90 degree angle between the runner and plenum.

 

I've not had the time to work on it as much as I would like, but I think it's a solid concept. Build-able? Probably not. Once I get the "ideal" design down, I can start breaking this thing down to individual components to determine feasibility of construction.

[rejracer]

Here is the latest, I did not like the angles so I split the runners into two pieces, it's much more compact so far.

 

I'm still playing with the runners, but I think the geometry is there, now I just need to tweak diameters.

 

 

from the front:

 

Current runner diameters are 60mm inlet and 38.1mm outlet. I don't yet know how to calculate the ideal, but then again, with the theory I am taking with this, I don't know if established guidelines apply. If anyone has any advice, I'm all ears.

[Leon]

The straighter the runner the better. Bends introduce pressure losses. All of your concepts have bends in them, and in my opinion, the performance cost of the bent runners outweighs any advantage for the sake of a "better" plenum. It just doesn't make sense to me to design the runners around a plenum and not the other way around.

 

My 2 cents...

[rejracer]

Leon,

One point to consider: All L6 intake manifolds have the air entering from front to rear of the vehicle. Since this is the case, making that 90* turn over the largest arc a will consume the least amount of energy. Velocity change should be made over the longest distance (within reason). Velocity change could be the air speeding up / slowing down, or changing direction.

 

The question remains, is this more efficient than a straight runner.

 

Why does it make sense designing runners after a plenum? Is there some math behind it, or does it just seem logical?

 

Thanks,

Robert

[Leon]

Runners have, or should have, a design rpm range in which there is a VE boost caused by pressure wave resonance. Runners should also be designed with as few pressure losses as possible, meaning as straight of a shot into the combustion chamber as possible. I think it makes more sense to design runners with those criteria in mind, and then adapt a plenum.

 

If I were putting so much effort into designing an intake manifold I would be working with an ITB setup. IMO, it would be easier to get right and you'd get better performance in the end.

 

I don't mean to sound harsh, so take it for what it's worth. More power to you if you keep this up and get a nice design out of it.

[rejracer]

Thanks Leon, this particular project is not going the ITB route, as it's whole purpose is to explore different designs. I agree with everything you are saying, just looking for the supporting math to design a runner, rather to design a proper velocity stack for a given flow / pressure at a certain rpm range. I've not found what I need just yet. So far, the design has been based on packaging needs, not really on proper design of the diameters of the runners.

[SKiddell]

Echoing what Leon says, the overall runner design will greatly affect the characteristics of the engine and as such its difficult to work at a this level without knowing the intended purpose (Road, Race, Rally) or induction method (Carb, TB, Forced) and the exact specification of the intended engine.

 

Just as background info

From a recent rolling road session we experimented with different ram pipe lengths on an individual TB NA L6 for a whole day, the results were astonishing, long pipes gave great low to medium range cylinder fill and thus great torque (252 ft/lbs at one point which is a BMEP of over 200 and a VE of 111%) but the car ran out of steam around 6700, shortening the pipes brought the low to mid torque down a little (we settled at 242 ft/lbs ) but lifted the peak power up to 7300. The worse results were for open ram pipes and no pulse plate this absolutely killed the power curve (theory was…out of phase reflected pulses from the inner body work), we gave up on pinning it down as the results were so poor.

Of course these results were a net result of changing the overall tract dimensions resulting in velocity changes, possibly enhanced venturi effects and pulse wave tuning.....this experiment no doubt would have delivered different numbers but probably the same overall "effect" on Carbs and gives a little insight into just how critical the dimensions can be, ….I guess what I am saying is that you need some idea as to the required dimensions and then they will influence the design

 

On a side note would really like to try short body TB's in an attempt to get the overall tract length below 300mm.

 

Now you watch TD or JC come shoot me down

Edited June 9, 2011 by SKiddell

[Leon]

Thanks Leon, this particular project is not going the ITB route, as it's whole purpose is to explore different designs. I agree with everything you are saying, just looking for the supporting math to design a runner, rather to design a proper velocity stack for a given flow / pressure at a certain rpm range. I've not found what I need just yet. So far, the design has been based on packaging needs, not really on proper design of the diameters of the runners.

 

A "velocity stack" is nothing more than an extension of an intake runner. The only difference being the ends, having a taper and bell-mouth. This improves pressure wave reflection. I have the calcs somewhere but don't have time to dig them up. They involve knowing your valve timing, air temperature (for speed of sound calc), and runner geometry among other things.

 

I guess the question is, are you willing to give up X amount of performance for ease of packaging? Runner design will have a huge effect on your engine, as SKiddell points out.

[Drax240z]

Good discussion here that might help you out:

 

http://www.atlanticz.ca/forums/index.php?topic=578.0

[Leon]

Interesting discussion, although I did not have time to read it all. There were a few unknowns and some misinformation though.

 

One thing I want to comment on is runner taper. The reason for the 7 degree number is because it's the maximum theoretical taper before flow separation (bad). A smaller taper will give your intake a wider tuned bandwidth but overall magnitude of VE gain will be less. Reverse that for a bigger taper.

 

Another post that irked me:

 

Without horns, there is turbulance that reduces the flow

 

Tapered runners increase pressure/velocity/density so that air flows faster into the chamber and it is harder to push back into the plenum.

 

The flow in an intake is always turbulent. The air horns simply increase/move the VE boost caused by wave tuning. The second sentence is essentially true, just stated in a somewhat awkward manner.

[Drax240z]

With the curvature/routing of the runners being modeled in this thread, I'd be wary of using any number (ie: 7°) as a theoretical maximum taper, and use some CFD or flowbench options to verify the actual design. If I was going for an all out race build.

 

Thus goes the theory of intake design... I'd be shocked Nissan considered half of this stuff when designing the L-series cast log manifold. At some point, you make compromises in the theory, and build something within your constraints.

 

Fits in the allowable space? Check.

Gets air from a filtered source and directs it into the head? Check.

Possible to build? Check.

 

Those tend to be the more important constraints than velocity stacks, runner taper, pressure waves, etc. etc.

[Derek]

Fits in the allowable space? Check.

Gets air from a filtered source and directs it into the head? Check.

Possible to build? Check.

 

 

Don't forget "look cool, Look really cool, actually runs"

 

That was my design criteria on my manifold!

The one thing I would change at this point on my project was to run some CFD software on it. I wasn't looking for performance but it would have been nice to have the opportunity to tweak the design a bit.

 

Derek

Edited June 10, 2011 by Derek

[rejracer]

thanks Drax, I've been reading it. i understand the helmholtz principle, now trying to understand the math behind it... You posted some good info there, I'm still trying to digest it.

[TimZ]

One thing I want to comment on is runner taper. The reason for the 7 degree number is because it's the maximum theoretical taper before flow separation (bad).

 

Maybe I missed something, but my recollection is that this only applies if the taper is opening up the cross sectional area, as opposed to this case where the taper is causing the cross-sectional area to get smaller.

Edited June 11, 2011 by TimZ

[Leon]

Maybe I missed something, but my recollection is that this only applies if the taper is opening up the cross sectional area, as opposed to this case where the taper is causing the cross-sectional area to get smaller.

 

You may have a point Tim, I had restrictor design stuck in my head.

 

The important effect of taper in the case of starting big and decreasing as you approach the head is the bandwidth/magnitude. A longer and slower taper increases bandwidth while decreasing magnitude of VE gain, and a shorter but steeper taper will do the opposite. An rpm range must be specified in order to properly design an intake manifold.