Intercooler CFD Testing

[TimZ]

Hey Jeff...

Something I've been meaning to ask - I'm curious about the effect of the air making a 90 degree turn before entering the IC. I would guess that there would be a density distribution such that there is more air towards the outside of the turn. So, for instance, your setup should split the airflow pretty much in half, but if I did the same mod to my IC above I think I would end up biasing the airflow to the bottom half.

 

Would it be a big deal to re-run this model with a 90 degree turn before the splitter just to see if this works out like I think? Maybe one with the turn oriented like you setup and one like mine?

[bradyzq]

Hi All,

 

I know this thread has been dormant for a long time, but what I have to add/ask seems topical, so here goes:

 

Your flow analysis has been from the point of view of the charge air, since all else is external. The charge air sees a cross-sectional area in your case of 18W x 3.5D or 63sq. in.. This number is proportional to the flow rating of the intercooler, all else remaining equal. So, in theory, a 21W x 3D or a 25.2W x 2.5D intercooler core should flow the same as your 18W x 3.5D core.

 

So having a shallower but wider core should not impede flow. But, would it not cool the charge air more effectively? The ambient air that is going through what it sees (in your case) as an 18W x 6H core has the biggest temperature delta (difference between charge and ambient air temps) as it just begins to pass through the 3.5D core. So, it stands to reason that it has the smallest delta temperature just as it exits the core. This means the ambient air cools less and less efficiently as it passes through the core.

 

So, if the above is accurate, wouldn't a shallower core be more efficient than a deeper one?

 

I realize you were just trying to salvage a crappy design, but the thread seems to have gone beyond that, even though it has barely touched upon the cooling abilities of different types/shapes of intercoolers (air-air in this case).

 

If you're interested in doing the simulations, I'd be curious to see whether a 21W x 5.14H x 3D core or a 25.2W x 4.29H x 2.5D core would flow any differently than yours, all else being equal. These all have the same frontal area (108sq. in.) and same end tank x-sectional area (63sq. in.) Or try them with 6H, too?

 

I have zero experience with the type of software you're using, but if you can introduce the cooling effect of the ambient air, it would be very interesting to see which version "wins."

[TimZ]

As I recall, the main thing we were looking at in this thread dealt with method for decreasing the pressure drop across the intercooler core for a given amount of airflow. The effect this has on efficiency is twofold:

 

First, you maximize the amount of core being used, thus maximizing the amount of cooling surface area available.

 

Second, for a given amount of boost in the intake manifold, the turbo has to create lees outlet pressure, so the air going into the IC is cooler to begin with.

 

That's an interesting aspect for discussion, but I don't think the software that Jeff is using is designed to actually look at the thermal transfer properties of different core shapes. Perhaps there is something that could be done there, but I think that that type of analysis would be even more involved, and you'd need some pretty advanced models of the heat transfer interface (cooling fins, both heat transfer and airflow models, and much more) that probably don't exist with enough fidelity to be useful for everyone (i.e., conclusions for one core might not be applicable for another, and it would be really difficult to know what is applicable and what is not)

[jeffp]

I was looking at the picture of the intercooler and noted that the flow through the core is 90 degrees of the end tanks. Did you choose this type of cooler so that you could run the inlet and outlet on the same side? I took an all together different flow path for my setup.

 

I also had spearco do my cooler, and I have to say it looks like their quality has gone down since they have been bought.

So I got the SP-209 core to start and had the tanks welded to the ends. I did send them a drawing of how I wanted the tanks made. The core tubes run horizontal instead of vertical with the end tanks on each end.

So air goes into one end of the tank, through all of the core passages, and out the other end.

I did have to run a longer discharge tube to get back to the intake inlet, but it is in front of the radiator and is not subject to heat soak from the engine.

To me that seems like the better option for the intercooler.

I think there is a better cooling effect with the longer core tubes, and flow is less turbulent in and out of the core.

anyway.