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

[gldgti]

OK.

Introduction to fluid mechanics, 6th ed, Fox, McDonald, Pritchard. Chapter 12 - Compressible Flow - Subsonic Flow -

"In a converging section [a reduction in area, if you like] pressure must decrease, and velocity must increase" - This principal should be known to us all - like a venturi - a reduction in area of the pipe causes the fluid to increase in velocity and decrease in pressure (Bernoulli Principle). The reverse is also true - "a diverging channel must lead to a pressure increase and a velocity decrease" - a subsonic diffuser.

What does this mean for us? I'm still thinking about it now actually....but put it this way - try to keep that relationship in mind when you think about the fluid flow through transitions in cross sectional area, and what it might mean.

As for calculations, the time may come, but its not an attractive option for me as it means solving many many high order couple simultaneous equations... so i would need to write a matlab program at least....

All I can really say libby is that you've made me dig out my books, and now I'll have to do the bloody thing properly instead of just living in my happy little world of ignorence.

[Mark(The Miser)UK]

Books on fan/ventilation engineering, rare, sometimes pricey, but a good read, cover much. Bernoulli never used a right angled elbow to demonstrate his theories did he?
I thought in trying to get the most out of the engine, creating inline resistances were things to be avoided.

Why not make the slot enormous and stick a coarse filter across to remove swirl perhaps like scotchbrite...

[CRSMP5]

ive got another to add question/idea..

why not add a intercooler directly to the runners and cone shape the end cap where the turbo feeds in?

reason...

mk1 jetta coupe, has ac and keeping ac with my tdi-m build. no matter what i have to make a hole in my hood to do a intercooler set up as is..

figure in a 2 piece design, add a flange at the runners going into the inter cooler, so can be removed easily so the valve cover can come off. there is enough 16v manifolds in junkyards for the flanges IMO..

you can also then have a factory gasket for the flange too, then add in the 45mm ish hole size of that to the runners you could build them into velocity stacks also.. aka big hole feeding into small hole..


yes i have thought it out.. just wonder why i see no one else try this

[Aki-76]

for normal street driving, the charge air tubing should be made small enough, to get faster throttle response. 50mm or even 45mm tubing should be good for under 250hp. when planning the dual plenum intake manifold dimensions, the charge air tubingdimensions shoud be also considered. It's very important to avoid unnecessary increases/decreases in charge air tubing diameter.even the ic should have the right size inlet/outlet but sometimes it is not possible.about the dimensions of the first plenum: the diameter should be the same as thecharge air tubing, and the length the same as the second plenum.the area of the groove using 50mm tube should be 19.5-20cm^2, rounded up. In a 30cm tube this means about 6.6mm wide groove which is rounded up to 7mm.earlier it has been said that the groove should be smaller, but I disagree. The same size or slightly bigger.this applies to street-driven engines, racing engines have slightly differentdimensioning.


Hopefully this helps a little.

 

[gldgti]

for normal street driving, the charge air tubing should be made small enough, to get faster throttle response. 50mm or even 45mm tubing should be good for under 250hp. when planning the dual plenum intake manifold dimensions, the charge air tubingdimensions shoud be also considered. It's very important to avoid unnecessary increases/decreases in charge air tubing diameter.even the ic should have the right size inlet/outlet but sometimes it is not possible.about the dimensions of the first plenum: the diameter should be the same as thecharge air tubing, and the length the same as the second plenum.the area of the groove using 50mm tube should be 19.5-20cm^2, rounded up. In a 30cm tube this means about 6.6mm wide groove which is rounded up to 7mm.earlier it has been said that the groove should be smaller, but I disagree. The same size or slightly bigger.this applies to street-driven engines, racing engines have slightly differentdimensioning.


Hopefully this helps a little.

 

Aki,
if you use 50mmID tube, the area is 1963mm^2. If you have a 7mm slot width, that is 260 long (which is the length of the plenum on my design, and about the same on others of these 4 cyl vw ones) then the area is 1820mm^2. This is smaller!

Andrew,
Don't hold your breath on the book :-)... I may be a while

I don't have any hands on experience with solidworks, but I've never heard of it in the context of a flow dynamics program. There is however a VERY good program made by Ricardo (who would have guessed? hehe) but I don't know of its availability to the GP. I will look into the solidworks side of things, it might be interesting. I would be curous to see how what maths it uses.

[RadoTD]

I have an older version of Solidworks as well, a semi functional student version. It's got CFD in it, but unfortunately I can't get it to work, my program keeps crashing. I was even trying to run a simulation on a dual plenum intake with it
It's fairly simple setting it up; there's a good walkthrough in the help files, so if you have a copy of solidworks, making the model of the intake will be the toughest thing.

I've been reading this thread from the beginning... I have a general understanding of fluid dynamics but have definitely learned a few things here. Somewhere on the interwebs, there is a 30+ page thread on someone designing/building a dual plenum intake for his car and having CFD simulations run on it all. The one reocurring theme I saw with every incarnation was that it was somewhat unpredictable; as is a lot of fluid dynamics as far as any of us can calculate. I'm not trying in any way to say it's a moot point that it shouldn't be discussed, but short of dyno testing (even with EGT probes on each cylinder?) and/or CFD running many simulations we can't prove anything and shouldn't get too worked up over it
I'm searching for that thread; IIRC, the guy running the simulations was a mech engineer working with fluid dynamics so there was a lot of good info in that thread. I should've bookmarked it!

One idea I had for an intake that was briefly mentioned earlier is running a proper merge collector near the injection pump, then running parallel charge pipes over the valve cover, each going into the middle of a plenum for two cylinders and runners dropping down from that. I'm still tempted to try to build it, that's the way I would most likely go if I was building an intake mani. It may not be the most perfect design, but I think it would be the easiest way of getting similar flow to each of the 4 cylinders.

[gldgti]

RadoTD - I agree, CFD is a bit of a dark art. There are applications where CFD works well, and others where it becomes more theoretical - mainly because of complexity.

The real world situation is so complex, with a variable frequency/position outlet port that has a transition from open->shut->open again, and on the inlet we have a more continuous supply....

Not only that, but depending on the design of it all, the the operational envelope, sometimes there may be some laminar flow, but mostly there would be turbulent flow. All these factors change the mathematics, and any model we make is ofcourse based on assumptions, that may or may not be correct.

When speaking on this subject with my associate, he said to me "just the shear fact that the plenum feeds to only one port at a time makes it difficult to get it really wrong - this is why VW can get away with such a seemingly thoughtless design".

There comes a point where you just throw your hands in the air and laugh at the the complexity of it all....

[regcheeseman]

the diameter should be the same as the charge air tubing, and the length the same as the second plenum.the area of the groove using 50mm tube should be 19.5-20cm^2, rounded up.

Aki, thanks for that. It's exactly how I will do it. you've also confirmed that whilst not 'scene friendly' my 2" pipework will be fine for practical applications.

[RadoTD]

In that same thread I still can't find (but I swear it exists!) there was a chart that listed pressure drop given different sizes and airflow. I remember grazing this subject once, the name guy who came up with the law/equation started with a P I think

But anyway, I googled it and found an online calculator here http://www.engineeringtoolbox.com/pressure-drop-compressed-air-pipes-d_852.html
If you scroll to the second chart, that has 2" and 2.5" ID pipe listed with the flow values we're looking at. It's at 100psi initial pressure and 100m long pipe though, but it'll give you an idea. You can plug values into the second calculator up top to get reasonable values.
2" is probably the lowest I'd go; that's about where the pressure drops start climbing faster than you want it to.

For reference, 25psig, 300cfm, 10ft and 1.5" is 1.21psi, 2" is .29psi drop, 2.5" is .09psi. And 300cfm is about the most air you'll pump through these engines unless you're going nuts

I'm personally going 2.5", but you're not losing much with 2". 1.5" would be a different story...

[theman53]

The tube diameter I don't look at as loss. I see it as how much extra air will the turbo have to compress to get it to spool with a bigger tube.
125.6 cubic inches is what I roughly figured for 2"
196.25 is same only 2.5" tubing
So you would have a little more cubic inches to fill up before the boost reaches the valve. That should increase spool time. From a daily driver stand point the smaller tubing should give better response, but all out racing or max flow you would probably want the bigger tubing. I think this is what AKI is refering to as well.

Same thing with the exhaust manifold. I am guessing if you had tubing the exact size of the exhaust ports and as short as possible to the turbo it would spool best. Not overall flow best but spool best. Then after the turbo the biggest tubing to make sure it doesnt' hold it back.
Just theories...like Andrew said earlier I probably am not helping clean the water, but muddy it more since I have no "proof positive" reference. Sorry guys just thinking out loud.

[gldgti]

but you're not losing much with 2".

depending on what else goes on, you might not be 'losing' anything at all. The pressure drop is greater than that of 2.5", but the gas velocity is much higher. Depending on a lot of different things, it might be better to have the higher velocity air.... Its all just so complex that the only way to say if one solution is better than another for a given application is to do a back to back test.

[Toby]

Am I to understand that the consensus of opinion is that the larger the plenum the better? If that is so, why not start with a 1.6 N/A intake? If one boxed in the filter side one would have a very large plenum for the slot to feed. IIRC they have pretty straight parallel ports with a bell at the plenum end.

As for the tapered plenum dimensions, I think the photos have enough info to get us in the ball park. The length we know from the pics is just over the distance between runner 1 and 4. The inlet diameter appears to be the same as the inlet pipe diameter. It looks like about 3". The narrow end looks to be close to zero on the 4 banger and under an inch on the Audi motors. Some blown up prints and a protractor might help nail down the angle of the taper as well

If you used the N/A intake for a test mule you could fabricate the closure and second plenum to clip on like the black plastic N/A air cleaner housing. Probably not optimal but very quick to take apart for playing with slot dimensions while doing pulls on the "long hill dyno".

[Toby]

Lots more photos here:

http://www.bufkinengineering.com/intake%20manifolds.htm

[Thezorn]

I think im gunna try and make one of these for my AAZ also, only thing im not clear on is what size tubbing is going to be used for the runners?

[gldgti]

FWIW - the main reason most aaz users will want to go down this path is because the early intakes and gasser manifolds all mean you need to have the turbo mounted down low, which is a PITA on a mk3 with a T3 or other larger turbo.