Twin plenum design questions? - Now finished 1Y based manifold

[regcheeseman]

Just trying to clarify a few things about twin plenum design, with reference to my high quality CAD drawing...



The plate in blue that divides the two chambers is slotted, should the slot be tapered as shown or should it be parallel? Should the area of the slot be equal to the area of the single pipe coming from the intercooler i.e about 5 sq inches?

[theman53]

I would contact the AKI76 guy. He made mine and it is definately parrallel, but no idea on the area of it.

[regcheeseman]

I've sent him questions in the past and received nothing...  

Reading through 'TurboJ's' Jetta thread, you are suggested as the man to talk to.

I'm building a motor up based on a 1.6 bottom end, VNT turbo and a 10mm headed pump.

Currently I'm thinking of using the original 1.6GTD head (with a little flow work) unless a 1.9 head turns up or do you believe the 1.6head is adequate for 150bhp?

I have either a NA 1.9 or NA1.6 inlet manifold and was thinking of using the 1.6 one and building a twin plenum chamber inlet manifold.

Do I keep the long runners or shorten them to limit low down torque?

Are each plenum volume = 1600cc or greater

And the tapered slit between chambers? How big? how tapered and does the taper get narrower nearer the intake pipe connection?

I'd really appreciate any help you could offer me with this project or pointing out someone or something (book/internet) that will help.

Thanks

Regcheeseman

Sent sept 7 2009 - shows how long this has been on the 'back burner'

[theman53]

There is the one I am not using until I get a VNT. IIRC AKI said something about it is a ratio of the inlet pipe, but i don't know how much. I woud guess that it is right around the piping area or maybe just a bit more.

[regcheeseman]

Thanks for that, I've seen that pic before I think, and I still had to look twice before I could tell that the slot was actually parallel - the shadow makes it appear tapered.

[blackdogvan]

I think Andrew hit it on the head. A large volume single plenum like the passenger performance intakes is the best DIY without some proper maths.  If anyone has access to a SolidWorks workstation with the flow works add on you can do some pretty great modeling like this very easily.



On a side note, check out the intake restrictors on those turbo's!!

[Vincent Waldon]

Pretty picture...sadly I've forgotten everything I used to know about flow or FEA.

Well, almost everything.  Next time you're at a fancy-smancy resturant... one where their idea of ambiance is to remove the ceiling tiles and spray paint everything flat black... have a look at the HVAC while you're waiting for your lobster bisque.  If the air distribution pipes are long enough you'll probably find that they are stepped down in diameter gradually as the air is distributed... on purpose to help keep the velocity consistent as the pressure decreases.

Makes you wonder a little about how much you actually gain by slapping on a gasser manifold or anything else with "straight pipes"?

[theman53]

I know very little about this myself, but I do know that Aki asked many questions to get an idea of how to build it for me.

He asked what I was going to go for upper boost wise. Intercooler piping size. If I was going to put a different cam in it. Daily or drag race, or road race type of driving. Then all the obvious questions like what engine, pump, car, etc. He asked all of that for some "math" equation that he had to plug them into to get to come up with pipe size, size of the slit, and something about the 2nd plenum.

You could probably achieve the conical effect on the inside and still make the outside look like a regular pipe.

[R.O.R-2.0]

Pretty picture...sadly I've forgotten everything I used to know about flow or FEA.

Well, almost everything.  Next time you're at a fancy-smancy resturant... one where their idea of ambiance is to remove the ceiling tiles and spray paint everything flat black... have a look at the HVAC while you're waiting for your lobster bisque.  If the air distribution pipes are long enough you'll probably find that they are stepped down in diameter gradually as the air is distributed... on purpose to help keep the velocity consistent as the pressure decreases.

Makes you wonder a little about how much you actually gain by slapping on a gasser manifold or anything else with "straight pipes"?

Gasser intake runners are far from straight. they are VERY tapered.. they are FAIRLY HUGE at the plenum..

or were you talking about some other sort of "straight pipes"?

[Vincent Waldon]

Yeah, poorly worded, what I was trying to describe is the 4 identically-sized runners (tapered or not) that come straight off the main plenum.

[gldgti]

The purpose to the dual plenums is to reduce the turbulence going into the second plenum and evenly distribute the air to the individual runners.  My imagination tells me that without the diminishing cone compensating for the loss of pressure from the initial air flowing into the second plenum, the result would be added turbulence in the second plenum and unequal pressure to the runners.  I also imagine that the plenum being straight and the slit being larger toward the inlet would be highly detrimental as it would allow excess air in at the highest pressure area and very little air in where the pressure would be least and the slit smallest.  It would seem to be better, if the initial plenum is a "cylinder" rather than cone, that the slit should be smallest toward the inlet and largest toward the later runners in order to compensate for the pressure loss across the initial plenum.  Designing the initial plenum to be an appropriately sized cone would be easier and more effective.   I've also been of the opinion that without knowing what the goals were and the design parameters and calculations to achieve those goals one could easily cause very little benefit and possibly cause more harm as good.  I suspect that many DIY dual plenum intakes are of the "wing it" design and their effect is questionable.  Not to criticize AKI in any way (his efforts are very remarkable) but before I emulated his intake design in any way I'd want to know what calculations he used for the design and how he arrived at using those calculations and so far, despite seeing several people on this board ask, zero info on calculating the size/shape of the plenums/runners has been provided.  

A couple of things -

> the area of the slot should be LESS than the inlet pipe area. This is because we want the first plenum to be "pressurised" compared to the second one.
> the diminishing cross sectional area of the first plenum that we see on the audi race engines is ideal, but tricky to make
> the way you have described the "slot width relationship" is not correct. The thing you need to remember is that the air mass has plenty of velocity, and hence, inertia. Lets imagine there was no slot, just a single plenum (like a gasser manifold). This design will almost certainly preferentially feed the cylinder furthest away from the inlet pipe so some extent. In reality, to try to balance the flow by changing the slot geometry, you need to minimise the area of the slot at the end furthest from the inlet, and maximise the area at the end nearest the inlet.

In the other thread with my manifold, I have make the slot area 85% of the inlet pipe area, so as to ensure that the first plenum is pressurised compared with the second. This is effeectively the same as reducing the velocity of the air, and hence the iniertia. This means any feed preference due to the inertia of the inlet gasses is minimised.

[gldgti]

I have not calculated my runner lengths. They are just as long as they are.... I don't think its vital for my application to "tune" them to a certain length because all I'm really trying to do is improve over the standard aaz manifold, which has no runners at all to speak of.

As far as the total slot area parameter - I managed to get a guy on another forum who was building a 2.5 5 cyl TDI drag engine (for a transporter no less :-) to tell me how he calculated the size of the slot area.

As far as the information I gave about inertia and feed preference - this is based on a bunch of thinking that I did a while back when I was thinking about making a manifold, and I spoke to a colleague of mine who did a lot of work in fluid dynamics with our CSIRO (Commonwealth Scientific & Industrial Reseach Orgasnisation). Basically it became obvious that the fluid inertia is the biggest problem, and that keeping that in mind, what you need to try and do is make flow through the slot the same velocity from one end to the other, if it is parallel.

I imagine that the better the primary plenum design is, the closer in area the slot can be to the inlet piping - so you achieve better flow. I have made my slot that much smaller than the inlet pipe because my primary plenum design is poor.

[theman53]

You could split the IC piping 4 ways equally and then put a dual plane on each cylinder???

[Mark(The Miser)UK]

You could split the IC piping 4 ways equally and then put a dual plane on each cylinder???

EDIT... I  haven't actually drawn it yet!But my description is  clear as ever
Haha,
I've just downloaded a 'free' CAD to draw my 'same' idea.  
'Tiger CAD'
http://www.freewarehome.com/index.html?http%3A//www.freewarehome.com/Graphics/Drawing_and_CAD_t.html

On the straight part create an adaptor that goes one into four then with gentle curves, tubular runners into simple flanges.

IMO, the problem with the balanced slit, which would give equal pressures along the chamber, is the actual pressure drop across.
The crude [no disrespect intended, I wish I could weld like that ] blunt cylinder must suffer from lots of turbulance and 'dead' or even very much alive and interfereing air.
The nice AUDI cone feature reduces that, but still has to be a resistance, to perform.

Compare with the equivalent sharp 90 degree bends, like in cheap modern central heating which are terribly inefficient and create hammering when taps are opened. [Contrast with the earlier "good ole days's" central heating's swept bends]

Likewise, the laminar flow you and I envisage  must be better, and easier to construct.
Perhaps a little home casting to make the one into four.

[gldgti]

Andrew,
Your right, I know my "references" are poor ones. I don't want to "toot my own horn" so to speak.... I don't know what training you have, or what you do for a living, but I know my own training and my own expertise, and I know other people with expertise that may or may not be pertinent to this subject. I can't give you any hard references about what I have said about the fluid dynamics, all I can say is that my education and training allowed me, upon investigation myself over the past 12-18months on this subject, to speak with people who's job it is to know fluid dynamics, and understand the how's and why's about turbulent flow through these types on pipe/chamber arrangements.

I am a university trained mechanical engineer, and I work as a mechanical engineer, but I do not work in fluid dynamics, so I can certainly say I'm NOT an expert on it. But, like I say, I know people who do work in this field and asked them about it. My own training merely gives me the capacity to more easily understand the explanations I seek. At the end of the day, this is my weekend tinkering hobby, not my job, so I'm never too careful about writing anything down because... I'm lazy. It all just sits in my head.

Now, I should have been a bit more tactful when I stated that something you wrote was "not correct" - especially given all of the above. However, if you would like me to try and explain why I said what you wrote was incorrect, please reply to this post and I'll go about it to the best of my ability. One piece of "evidence" (or lets say interesting data) I might look to, to try and explain, is a reply to my thread about the stock manifold, referencing bent conrods.

Cheers :-)