13

Project [Rotary] FD RX-7: Part 6 - V-Mount Cooling System (Overview and Intercooler)

by M-P Spierer

The cooling system in the FD RX-7 is widely regarded as its most glaring weakness. This is partially due to the design of the rotary engine itself and partially due to issues specific to the design of this generation RX-7. We can't fix the flaws inherent to the rotary engine, but we can do something about those external to the engine. Most notable of these is the fact that many of the engine's cooling components are constructed out of plastic, the intercooler is wildly undersized, and the orientation of the radiator does not allow for proper airflow to the intercooler. In the next two installments of this project we tackle these issues by implementing a 100% custom v-mount cooling solution.

The first question you might be asking yourself is, "What is a V-mount cooling system?" Before we answer that question though, let's take a look at the stock cooling system configuration and its flaws. Unlike many turbo cars where the radiator and intercooler sit vertically one after another behind the front bumper, the stock FD RX-7's setup is different (I know, what's new?). In the RX-7 the radiator is angled forward while the intercooler is located above it and is fed by a small duct. The intercooler is also a tiny thing that consists of a 4.5x10.5x2.5" tube/fin core fed by a pair of plastic end tanks.

 

A graphical representation of the stock cooling system orientation. Notice that the duct for the intercooler gets very little priority compared to the radiator, which is for all intents and purposes unducted.

Because the stock intercooler is so small, it really needs to be upgraded if any RX-7 (even a stock one) is going to see any track use. I have personally felt the power sapping effects of super high intake temps due to the stock intercooler within only a couple of laps at a short track. The main issue with upgrading the intercooler in the stock location is that there is barely enough room for one that is significantly larger and not enough room for the ducting that is required to feed such an intercooler. The best option is to flip the radiator so that the intercooler and radiator make a "V" shape...hence the name V-mount.

 

A V-mount radiator and intercooler orientation is fundamentally similar to the stock orientation, but allows for much better ducting of air to a larger intercooler without sacrificing flow to the radiator. Proper duct work is extremely important for a V-mount cooling system to function properly. 

Now, if you remember back to Part 1 of this project, there was a HKS v-mount intercooler system that was already installed on the car. Clearly this is not the first time we have had this thought. Originally the intent was to keep this system in place and only improve it by ducting it better. However, as time went on and the build progressed deeper into the world of uncompromised one-off custom fabricated components, we began thinking about how the cooling system could be improved even further and decided that a new ground up solution was in order. 

 

The old HKS V-mount intercooler setup that was previously installed on this project was nice, but there are several things about this design that we thought could be improved. Increase intercooler efficiency and reduce pressure drop by moving from a tube/fin to bar/plate core design, increase radiator cooling performance by changing the radiator angle and increasing cooling area, and increase both radiator and intercooler performance by implementing a true set of sealed ducts.
A significant amount of time was spent developing the RX-7's cooling system and nothing says this more than the fact that we created an entire 3D CAD model of it along side the cardboard mock-ups. Physical prototypes are still completely necessary when determining space claim for new components such as these, but there is something reassuring about having a CAD model to play with as well. It helps to further validate the design before wasting real dollars on scrapping parts because they didn't work exactly as intended.
Page 1 of 6 Next Page
Bookmark and Share
Comments
SM_Clay72
SM_Clay72link
Tuesday, June 14, 2016 7:57 AM
Looks great. Nice work.

When is it getting an LS swap?
andre333
andre333link
Tuesday, June 14, 2016 8:38 AM
Thanks for the info on the V Mounts, it's always nice to see a visual.

I don't know if it's appropriate to ask in this space but I've wondered about this question for awhile now.

Seeing as how the Rotary generates massive amounts of heat out of the exhaust and that heat can cause the internals to expand and in some cases cause poor sealing and general mayhem.
Wouldn't it make sense to move the turbo further away from the engine? Obviously this has some downsides in terms of ultimate spool and exhaust energy...but it would reduce strain on the engine and reduce under hood temps as well.
I don't know about full rear mount or anything, but some version.

Just a thought and for a track build where RPM's are always up the transient response wouldn't be terrible assuming a properly sized turbo was used.
Basshed
Basshedlink
Tuesday, June 14, 2016 9:17 AM
Would you share which bead roller used? I'm rerouting some aftercooler plumbing in a procharger kit & need to rebead the ends exactly as pictured. Appreciate the timely article!
M-P
M-Plink
Tuesday, June 14, 2016 10:29 AM
@SM_Clay72 | Thanks...and to answer your question, never. :)

@andre333 | That's an interesting question and one I have never really considered before. Moving this mass of heat further from the engine's exhaust ports could potentially decrease the amount of heat conducted back to the rotor housings from the exhaust manifold and thus increase engine life. I don't know how it would improve underhood temps though unless you relocated the turbo out of the engine bay. The heat generated isn't going to change, it's just where it is dissipated that would change.

Personally, I like the idea of keeping the system simple and feel that placing the turbo anywhere but on the right side of the engine would be more of a waste of space than anything else. A properly sized turbine housing and exhaust system combined with some good tuning should keep exhaust temperatures in check enough to not worry too much about this. From a purely academic stand point, it would be very interesting to look into this deeper.

Also on a side note, the Full-Race manifold used on this project has some of the longest runners I have ever seen on an off the shelf manifold. So it does have that going for it.

@Basshed | Sure. It is the Jegs Tubing Beader (P/N: 80083). It's officially rated for a max wall thickness is 0.050" but I decided to risk it on the 0.065" wall aluminum and it did fine with two passes. Quality is just okay and there seem to be better ones out there, but at more than double the price I couldn't justify them.
theneil
theneillink
Tuesday, June 14, 2016 12:57 PM
Cool! I thought the motor blew up or something, glad to see it around. I have a question about your tuning, will you be running any sort of anti lag (drive by wire throttle that stays allittle open, like a more passive form of the bang bang wrc type) or are EGTs already going to be close to the limit with the rotary?
rhocken
rhockenlink
Tuesday, June 14, 2016 1:59 PM
I've always been unconvinced about a V mount setup - though never done any calculations on it. It seems to me that a V mount takes a portion of that air flow for each cooler. A 'conventional' setup with the intercooler in front of the radiator will see considerably more air through the intercooler. That air will be heated up, then that same volume of air (much higher than through a V mount radiator) will go through the radiator. It will be hotter, but I've always thought that added airflow (twice as much if it is a simple 50/50 split) would compensate for that. V mount does have the advantage that the split point can be adjusted which is harder for a front/back setup. I've never seen any engineering comparisons however.
M-P
M-Plink
Tuesday, June 14, 2016 6:20 PM
@ theneil | No, I don't plan on implementing any kind of anti-lag when tuning, but it isn't necessarily because I'm scared EGTs are going to be too high for it. Though that is a big consideration when talking about rotaries. Tuning is going to be mostly straight forward with an emphasis on reliability. The primary goal of this build is to be able to go to a time attack event and not be worried that something is going to blow up. That way the focus can be on the driving and chassis setup.

@ rhocken | I completely understand the skepticism regarding v-mounts. As I hinted at in the article and will be talking a lot more about next time, ducting is the real key to a good v-mount setup.

In a conventional front mount setup, most people don't duct at all because it isn't truly needed for the reasons you stated. However, with a conventional setup there is still a lot of air that goes around the intercooler and radiator, not to mention the fact that in most cars there is a large portion of the intercooler or radiator that is blocked behind the bumper.

The idea behind a v-mount is to properly duct (meaning fully seal) both the radiator and intercooler to build up pressure on the face of each core. With a front bumper with a large enough opening to properly size the duct inlets for the given core sizes, you should be able to catch nearly all the air coming through the front bumper and force it through the coolers. At speed this should be more than enough air. V-mounts also allow for insanely short intercooler piping and more centralized weight distribution.

Do I personally have any actual data to back all this up? Unfortunately no, but the fact that almost all of the fastest time attack RX-7's (Panspeed, RE Amemiya, Garage Revolution, etc) run a v-mount setup gives me confidence that my theory is sound. I guess we will see.
Van_1986
Van_1986link
Tuesday, June 14, 2016 6:59 PM
I believe i read in some old Kojima article that only the first three inches of depth of a heat exchanger are efficient. If this is true then a stacked intercooler and radiator would be a nightmare for a car that needs extreme cooling like a rotary. Whether its true or not, It's become a rule of thumb in the rotary community to not use front mounts for anything other than street or drag.
spdracerut
spdracerutlink
Tuesday, June 14, 2016 9:32 PM
@andre333: what determines temp inside the engine? Well, the energy of the fuel goes into work and heat. Work pushing the rotors and the rest in heat which goes into the engine (coolant, oil, heating up everything metal) and out the exhaust. So I guess one way to keep heat out of the engine is to keep more in the exhaust gas coming out of the rotors. I guess what you’re thinking of is hot exhaust gas heating the manifold and that heat conducting back into the engine block. For a turbo car, remember that heat is energy, so you do NOT want to lose heat between the exhaust ports and the turbo or you will sacrifice performance. So, I guess you could try to minimize heat transfer from the exhaust gas into the engine block/rotor casings and also into the exhaust manifold. And then insulation of some type between the exhaust manifold and engine block.
@rhocken: long story short on v-mount, the frontal opening area for the air going into the heat exchangers is the same as stacked configuration, but you can greatly increase the area for the air to exit with a nice big fat vent in the hood ducted to the intercooler. Remember, air going in has to go out somewhere. Look at the extreme case of no exit area for the air (pretend a completely sealed engine bay). Well, no air will flow through the heat exchangers because it can’t get out on the backside right? With V-mount and ducting the IC to the hood, a lot of exit area was just added to the previous area which was through the bottom of the engine bay.
M-P
M-Plink
Wednesday, June 15, 2016 6:22 AM
@spdracerut | Good point with the extra exit area comment. For those that don't know...increasing the exit area for the same intake area (front bumper) means lowering the pressure at the exit. Low pressure on the back of a coolers combined with high pressure on the front of a coolers equals flow.

Funny enough, when I first installed the old HKS v-mount intercooler with the stock hood I could see the hood physically bowing up at highway speeds. It sounds crazy, but I swear that it didn't happen before redirecting flow up through the new intercooler. Only thing I could think was that the air that was going through the intercooler was putting pressure under the hood. I then switched to a vented hood and what do you know, intake temps dropped significantly since there was somewhere for that air to go.
JR
JRlink
Thursday, June 16, 2016 12:02 PM
Question:

You say that running a thicker intercooler aids not necessarily in better cooling, but instead you get less pressure drop. My question then is how is it that air to water intercoolers have such a small pressure drop being that they are so small compared to some of the large air to air intercoolers?
M-P
M-Plink
Thursday, June 16, 2016 12:24 PM
@JR | I have very little experience with air-to-water intercoolers, but I think the key is that they have narrower water channels running through the core than air channels because the heat capacity of water is so much higher than that of air. So the percentage of open hot-flow core area is greater on an air-to-water core than on an air-to-air core (which is about 50/50 on my Garrett core). Think of it like a radiator who's tubes are much narrow than air channels.

One way or another, the same rules will apply for both when it comes to hot-flow open area and pressure drop. Maybe someone with more experience designing air-to-water systems can chime in.
spdracerut
spdracerutlink
Thursday, June 16, 2016 1:01 PM
@JR, an air-to-water core can be sized smaller than air-to-air due to the higher specific heat capacity of water. Now, you need surface area to get heat transfer. You can size the core with essentially a large frontal area and short length or small frontal area with long length to reach the same surface area. Think big diameter/short straw vs small diameter/long straw. From a flow pressure drop point of view, big/diameter/short is the preferred way to go.

Uh, anyway, assuming air-to-water and air-to-air have the same frontal area, the air-to-water can be made shorter which reduces the pressure drop.

joeycrakk
joeycrakklink
Tuesday, February 21, 2017 8:35 AM
every week i check this project for updates and every week i am disappointed. What's up?
Post Comment Login or register to post a comment.

MotoIQ Proudly Presents Our Partners:



© 2018 MotoIQ.com