Well I have been snowed in since yesterday due to this east coast blizzard and had some time on my hands. I decided to do some basic simulations of typical square intercooler end tanks out of boredom. I figured I would share these with you guys as I know alot of you build intercoolers. These are just some quick simulations so these are by no means verified real results.

With mass flow rate of 3 lb/s




With mass flow rate of 20 lb/s



Here is a backdoor style one , sorry its not the best view




Its pretty much what you would expect. What are your guys thoughts on this? I am also willing to try some other designs if you guys want as Im going to be snowed in for the next two days or so.

 

would the sim be the same if you push it or pull it? Interesting. I also have read that cooler air moves slower. is this material heated or does temp have any effect on this?

 

how bout the intercoolers with the triangular type end tanks?

 

some fins in the tanks are not uncommen in a good fabricated intercooler tank..

 

I'd love to see some backdoor action

 

Temperature does affect this as an intercooler is basically a heat exchanger. Decreasing the temp increases the density of the air. The efficiency of an intercooler is essentially the heat put in / heat rejected out by the intercooler. I have never thought about using fins inside the endtank that is an interesting idea. Here is a link that has some pretty good intercooler theory.

http://www.gnttype.org/techarea/turbo/intercooler.html

 

X2. I would like to see simulated backdoor airflow also.

 

I can do a backdoor design in a little bit. Do you guys have any dimensions or spec's you want me to go by or will a rough design work for you?

 

So an aluminum rectangle box, with an inlet courtesy of a hole saw, welded to a heat exchanger, doesn't provide evenly distributed airflow!?!?!

 

What boundary condition are you using for the right hand side?

 

yes this is known, no need to be an a$$ about it. This is why I dont post much on h-t.

 

I actually wasn't trying to be a dick, but now that you mention it... Why did you post this anyways? Do you have a solution, or?

 

Thanks for the info and the time to look it up.

 

sharing information? why did you post?

 

I am curious to purpose of the thread.

 

If you would have read, I posted this because I have been snowed in and I was bored. I figured I would make a thread where we can discuss different inter-cooler end tank designs and have as an informative thread on this topic. I am sorry I didn't make a thread posting pictures of a turbo manifold or down pipe we've all seen a millions times. so please quit ruining this thread and stop posting your useless comments.

 

 

.dave I sent you a pm about the boundries

 

There wasn't anything I said to imply that you should have posted a manifold or a downpipe, you are getting butthurt for no reason. Lets see some tests on some different tank designs

 

I like the back door one, I dont get to play with stuff like this so simple things amuse me. I dont think Passanger was trying to be a ***, I think he was hoping for you to have some ideas on getting some even flow across the exchanger, or trying to inspire you to try some solutions. Hes pretty smart on things like this

If you have time can you try a backdoor with a radius on the front tank?

 

Idk passenger has been trolling lately....

When i think about intercooler end tank design i just look on google for race cars and take into account what they have done, and some of the nicest i have seen are rally cars.



this one is off a Bentley LM car


They like to flare the inlets out into a oval shape. I could only guess to distribute the flow over the core better. also they alternated sides on the inlet and outlet.

 

he included a it's all about the guys, he wasn't serious!

 

most inter cooler designs are such crap. People sometimes make such bad end tanks that their big *** intercooler only cools a couple of rows. Sure it looks pretty and big and impresses people, but the truth is that they are wasting space and decreasing air velocity. The problems are that people dont have the room to fit the right shape in, or the shape becomes too much of a fabrication problem for them.

http://www.mipowertechnology.com/hrsg.htm

This website shows a little CFD analysis on adding fins to direct flow. Such techniques are used in jet fighters called "air fences", which help slow moving air from 3D flow (left to right movement) on the wing. Keeping air velocity high with no piping losses is a bit tricky, but having the right shape and fins along with appropriate CFD for the matched mass flow the inter cooler can then have its best cooling potential.

SolidWorks is a good and easy way to see whats happening! Now put some new models up and lets discuss them!

 

I think one of things to remember when we say that most intercooler end tank designs are horrible, very few customers are willing to pay for 12 hours of labor making hammer formed end tanks. For most customers, sharp corners and flat sheet will be enough to keep everyone happy.

I've done some flow testing of some intercooler tanks I did for a project, and as expected, they flowed like crap. I have a pretty good sense of visualizing air flow in my mind, and I know just by looking at my end tank that it is bad. But what can you do?

I've tested using CosmosFloWorks, but I don't really know what I'm doing with that software, so I don't know how setup the air flow. Sure I can get the arrows up and it looks cool and all, but what are my actual results ... (I wanted to post some pics, but for some reason my remote desktop is not responding).

 

What you want to look for first is pressure drop. The pressure related to fluid velocity and Reynolds number. The higher the Reynolds number the higher the turbulence. Buy hand you can crunch only a spot at a time or just focus on a certain spot and known/assumed parameters (diameter,length, fluid viscosity etc). This can help you improve slowly one step at a time the problem areas. Nothing is an easy one step fix and takes time to refine. If I have time I'll open flowworks up and get something going.