PDA

View Full Version : Dual Plenum CFD



joe's gt
07-01-2010, 04:27 AM
Just in case any of you are curious, this is why a dual plenum kicks ass.

The CFD shows how equally this configuration distributes air. This can be optimized a lot more of course, I just wanted to show a quick and dirty analysis that shows why this design is so successful. It consists of a tapered plenum that the throttle body feeds. Then the air is transferred from this tapered plenum through a slit with an area approximately the size of the intercooler intake pipe into the main plenum that feeds the runners. This smaller top plenum and slit helps distribute the air equally to all runners.

http://i23.photobucket.com/albums/b396/92ct-27gt/dualplenumflow.jpg

Dual Plenum Cutout Pic

http://i23.photobucket.com/albums/b396/92ct-27gt/dualplenumcutout.jpg

karl
07-01-2010, 03:56 PM
gorgeous.

Johnny
07-01-2010, 05:10 PM
I did not like how these type manifolds looked like but its beautiful how they work

RedRkt01
07-01-2010, 07:17 PM
That's pretty cool, actually.

Mafix
07-01-2010, 08:41 PM
engineering always beats appearence.

takka
07-02-2010, 01:17 AM
hmmm this performs pretty well doesn't it, might have to attempt this after my other manifold is finished an do a comparison. no shortage of 3sge manifolds to hack up in Australia :) actually the design that just popped into my head would suit me down to the ground!

ChrisD
07-02-2010, 03:03 AM
Very cool.

Hiko
07-02-2010, 07:05 AM
That's freaking awesome!

Is that SolidWorks? If it is, what tool kit is that?

joe's gt
07-02-2010, 07:14 PM
eh, its just COSMOS FloWorks. Not the best but good enough for diy purposes.

I could probably get Ansys Fluent off of a torrent, but I don't really feel like learning a new software.

I honestly just wanted to see what all the hype was about because the dual plenum looked like a brilliant concept. Based on the CFD, it is.

ChrisD
07-02-2010, 09:11 PM
Question: would this cause a pressure differential between the 1st and 2nd plenums? (ie. higher PSI in the 1st)

joe's gt
07-02-2010, 10:59 PM
Question: would this cause a pressure differential between the 1st and 2nd plenums? (ie. higher PSI in the 1st)

Pretty much. That's why it works so well. As you know air moves from high pressure to low pressure and its forced to do so through that slit which ends up equally distributing the air. You'd think that slit is a restriction, but its actually not, because its suppose to be around the same area as the intake pipe going to the throttle body. So its not restricting the air, rather just redistributing it.

takka
07-03-2010, 03:43 AM
what im thinking is that the dual plenum would be no good for part throttle an driverbility, but im pretty drunk so review when im sober

rizin
07-04-2010, 02:00 AM
Hey Joe the pic is feed from the side by the first plenum. I was wondering on your CFD are you able to do a 3d on that design? If not no worries just would like to see what it looked like.

joe's gt
07-04-2010, 08:17 AM
It is 3D already, I just do a front view cuz its a lot easier to show the distribution in a 2D front view.

I can upload a 3D pic. DC power jack is out right now tho, so I got to solder a new one in. Probably gonna be a slight delay.

rizin
07-06-2010, 06:17 AM
I just want to see a east to west view that is all.

Edit: Sorry the CFD at the top looks like a top feed by the 1st plenum. So ya you got my looney mind thinking about designs. Even a slit that can adjust to throttle body opening beacuse of other post. Thinking is one thing and making is the other and working.

Hookecho
07-06-2010, 03:23 PM
would this concept work the same with the front feed intake?

joe's gt
07-06-2010, 07:08 PM
would this concept work the same with the front feed intake?

I think it actually would, but instead of one long taper, you would have basically a symmetric taper on each side of the center feed. Just like with the other design, this would be much more effective with velocity stacks.

Top feed tho, is just all around better because the air shoots straight into the runners versus having to make a 90* bend into the runners.

Mafix
07-07-2010, 09:26 PM
i'm actually suprised nobody has taken a stock 3sge manifold and added a chamber to the front of it.

Mafix
07-15-2010, 07:56 PM
mabye you and i can work on something like this soon joe.

joe's gt
07-15-2010, 08:03 PM
I was just gonna say, that seems like a really good idea. I said takka's manifold was gonna be my last, but this idea seems too tempting to pass up. Its a lot of work with all the modeling and trial and error.

Possibly add something like this (http://www.jasperintegration.com/distribution-plenum.html) to the front

You could also make your own by cutting flat plates at an angle and welding them to form an elongated triangle or pyramid.

Magnum
07-16-2010, 08:59 AM
For the front plenum your best bet is to use round tube and cut it with a tapering jig. Cut it so your cutting a V out of the length of the tube. Like this l/\l . The edges will line up perfectly. Weld the outside joint and cut the slit size on the inside edge and weld the rectangle connection tube.

---------- Post added at 07:59 AM ---------- Previous post was at 07:23 AM ----------


would this concept work the same with the front feed intake?yes
http://www.kmdtuning.com/images/uploads/DBRa4manifold.JPG


http://www.kmdtuning.com/images/uploads/DBRa4manifold.JPG

Mafix
07-16-2010, 06:33 PM
something like that joe but that's ass backwards from what i was talking about. time for me to photoshop something.

Macdaddyjaz
08-04-2010, 02:27 PM
I was going to Pm you dude, but the forum won't let me send PMs without posting a bit first.

I was told to come to this site and have a look at your CFD because I am trying to teach myself. I want to test a charge cooler I have designed and was going to get your advice on modelling this sort of thing.

Are you a CFD guy by trade of just a hobbyist?

Regards,

Chris

joe's gt
08-05-2010, 03:47 AM
I was going to Pm you dude, but the forum won't let me send PMs without posting a bit first.

I was told to come to this site and have a look at your CFD because I am trying to teach myself. I want to test a charge cooler I have designed and was going to get your advice on modelling this sort of thing.

Are you a CFD guy by trade of just a hobbyist?

Regards,

Chris

Sorry Chris, just a hobbyist. I haven't been able to do anything lately cuz I have been working 50 hrs. a week and preparing to move. Its not too hard to learn. If you are using Cosmos, they have some good tutorials to get you started. I was told by a fellow mechanical engineer to use a mass flow rate for the inlet and a pressure for the outlet. Ansys Fluent is a much more advanced flow program that can do transient analysis along with animations a lot better.

Let me know if you got any questions. CFD analysis is tedious because of the numerous trials you have to do to get the result you want.

Roundy
08-09-2010, 01:06 PM
What i see from this plenum is the slowest velocity in the runners is in the runner closest to the throttle with the highest furtherest away from the throttle.

And it looks significantly lower as well, the left seems to have velocities mostly in the 120-140m/s bracket whilst the right seems to have velocities mostly in the 60-80m/s bracket.

That velocity difference would hugely affect cylinder filling i would imagine.

What sorta flow rate is that?

With the rather low runner velocities it is still showing some velocities at 200m/s through the 'slit' which is quite fast.

ChrisD
08-09-2010, 05:05 PM
I think the manifold design could be optimized a lot to even those velocity rates quite a bit. Pretty sure Joe just wanted to see what "could be" and did a quick analysis on one. 1st plenum taper, TB angle, and even the 2nd plenum can all be adjusted.

JDM SNUKUMZ
08-09-2010, 07:44 PM
interesting thread

Roundy
08-10-2010, 10:22 AM
I think the manifold design could be optimized a lot to even those velocity rates quite a bit. Pretty sure Joe just wanted to see what "could be" and did a quick analysis on one. 1st plenum taper, TB angle, and even the 2nd plenum can all be adjusted.

Without a doubt, I am just putting a few questions to him to further my understanding.

ravensguildassasin
08-10-2010, 09:54 PM
interesting very very interesting. marked for later viewing. :popcorn:

Mafix
08-10-2010, 10:19 PM
no worries. in due time i'll work together with joe on some measurements and make a dual plemun ge manifold.

rizin
08-11-2010, 03:34 AM
Can't wait to see what you guys come up with.

joe's gt
08-11-2010, 04:22 AM
Honestly guys, right now I am swamped. I am in the first month of my new job right now and am working 50-60 hours a week and made a kind of quick decision to buy a house instead of rent.

We could brainstorm ideas and come up with possible configurations, but the actual CFD part takes A LOT of trial and error and is very time consuming. Each run through takes about 15 minutes on my crappy CPU. Times that by all the different lengths, angles, and configurations that are possible and you have quite a bit of development time.

I am totally willing to do this, but the actual CFD part will have to be a month or two down the road once I am more comfortable in my job. My lease ends soon, so I got to move all my stuff into storage while I live at home and save up for a house down payment, then move into a house when I have that. I am also having to help my Dad with his up and coming business. So my life is a little hectic and kind of in transition right now. I guess you could go so far as to say I am transitioning from the responsibilities of a kid to a man.

Mafix
08-19-2010, 03:24 PM
i'm in no big rush ATM. busy making huge changes to the car as is.
long runner FWD manifold where the turbo sits on top of the trans is the current problem, lol.

ChrisD
09-05-2010, 12:38 AM
Looks like we will see a dual plenum manifold based on the 3rd gen runners very soon!

http://www.pballandmore.com/images/securedownload_001.jpg

http://mr2oc.com/showthread.php?t=401392

Car_Barn_Bandit
09-05-2010, 03:00 AM
The other physical aspect of the dual plenum design is the air-multiplying effect.

Think of the walls of the first space like a river current. The current travels faster on the outside of the curve (along the 1st plenum wall) then the center. As the large volume center decreases in size and the slower air is in effect multiplied, or sped up through the long slit by faster moving air along the chamber wall, it helps create an even flow at a high velocity.

I wonder if some changes in the slit, either in distance between plenums, or the shape of the slit, so air flows evenly into the second space, could further enhance the design.

joe's gt
09-05-2010, 07:09 AM
As you can see form the first flow analysis, TB angle is a crucial part of directing flow. Having a TB go straight into a conical 1st plenum is easy and all, but it doesn't exactly optimize the flow through the slit. The conical 1st plenum can only do so much redirecting as air is flowing straight to the back of it. It is important to angle the TB to ease that 90 degree change in direction. Distance shouldn't matter too much, as long as its not excessive, but shape could have a significant effect. What is VERY important is to ensure that the cross sectional area of the slit is equal or greater to the cross sectional area of the TB, otherwise you will be restricting flow.

That 1st conical plenum helps to greatly increase air density and obviously helps equalize distribution. Velocity is not equal to the velocity at the TB of course, but it is increased as compared to a traditional TB to large plenum volume transition.

Mr.Shaun91
09-06-2010, 12:28 AM
HEY JOE,

question. did you build this or was it a pic you found to show everyone?

if you did build this could you send me the measurements as a basis for me....

i would like to build a couple diff sizes for my 7m and see what will work best with a turbo setup.
thnx

joe's gt
09-07-2010, 04:30 AM
Sorry man, it was just a pic I found and the modeling was just something I mocked up in Solidworks quickly.

You could try something like this (http://www.jasperintegration.com/distribution-plenum.html) or make your own out of some flat plates in a pyramid configuration or cutting a round pipe at an angle and welding the 2 halves.

Mafix
09-09-2010, 03:53 PM
no worries about throttle area as i never plan to upgrade mine to a larger one. this thread on mr2oc is nice to see and hopefully will help us in our designs.

ChrisD
10-01-2010, 05:18 PM
The dual plenum by RacerX that I mentioned above is now complete. He gained 3WHP and 9ft-lbs over his already custom manifold shown below. On the GT3076 he is now making 482whp, 383 ft/lbs or torque at 18psi. (2.2L)

http://i213.photobucket.com/albums/cc148/racerx1715/IMG_2653.jpg

http://i213.photobucket.com/albums/cc148/racerx1715/IMG_2654.jpg

http://i213.photobucket.com/albums/cc148/racerx1715/IMG_2659.jpg

---------- Post added at 10:18 AM ---------- Previous post was at 10:16 AM ----------

He is also building another one for a customer with stock 3rd gen runners.

http://www.pballandmore.com/images/Manifold25.jpg

http://www.pballandmore.com/images/Manifold26.jpg

Mafix
10-01-2010, 10:28 PM
been following them with great interest. i can only hope he does a stock runner 2nd gen manifold.