Thanks for returning to the final portion of this build. Now that we have finished testing, it is time to review our data and see the improvements our cooler provided. To start things off, here is a shot of the cooler fully installed!
Let’s dive into the data. First, we’ll compare the basic specs of our cooler versus the OEM unit.
The Mishimoto intercooler features a 37.9% increase in internal flow area compared to the stock intercooler!
The Mishimoto intercooler features an 80% increase in internal core volume compared to the stock intercooler!
The Mishimoto intercooler features a 79.5% increase in overall intercooler volume compared to the stock intercooler!
Now that we have the physical specs covered, let’s evaluate the data from our testing. For intercooler efficiency and AIT benefits, this data was collected during road pulls. We tested both the stock intercooler and the Mishimoto intercooler on a secluded road. The vehicle was driven from 3,000 rpm to 6,500 rpm in third gear. Each cooler was tested with three separate pulls, and these pulls were then averaged. To replicate consistent testing, we allowed the vehicle to cool for exactly five minutes after each run. The results are displayed in the graphs below!
These two graphs represent the big improvements you want to see with an intercooler upgrade. The Mishimoto intercooler was 10% more efficient at cooling AITs and was able to reduce outlet temperatures by 12°F on average. This is a great improvement on a nearly stock vehicle! Our engineers anticipate even greater improvements on a vehicle with higher boost, generating more heat.
Next up was our dynamometer results. Check out our plot below.
This testing was completed in a single day with little variation in ambient temperatures. The vehicle was strapped to the dyno and the intercoolers were swapped with the vehicle still on the dyno to ensure that none of the variables changed other than the intercooler itself. Each intercooler received a warm-up run and 4–5 additional runs until we had three consistent pulls to average. The fan shroud we fabricated provided the intercooler with 50 mph of airflow. The Mishimoto intercooler provided an average increase of 6 whp and 8 wtq compared to the stock intercooler. These are nice gains for a heat exchanger swap!
While we had the vehicle in the garage, we fitted it with our throttle body hose as well. Because the stock throttle body hose is quite prone to blowing out, we will be including our silicone replacement hose with this intercooler. We are assuming a majority of those upgrading their intercoolers are running higher boost pressures, so this is a great inclusion for some additional reliability.
So that essentially wraps up this project. Time to recap our goals and see how we did with meeting some of the tough requirements we set out to tackle.
Must be completely direct-fit for the 2008–2014 Subaru WRX and also utilize the stock-style turbocharger flange connection.
The Mishimoto intercooler bolts into position just like the stock intercooler and functions perfectly with all stock equipment. No vehicle modification is necessary for installation.
Support vehicles from stock power levels to 400 whp.
Our engineers recommend this cooler for vehicles from stock to 400 whp. We are setting a maximum rating of 500 whp, but we do recommend a front-mount intercooler setup for vehicles that achieve above 400 whp.
Reduce efficiency and AITs compared to the stock intercooler.
Our data showed a 10% improvement in intercooler efficiency and an average of 12°F lower temperatures compared to the stock intercooler.
Improve power output compared to the stock intercooler.
The Mishimoto intercooler provided 6 whp and 8 wtq average gains compared to the stock intercooler.
Increase cooler volume to support higher power and heat.
The Mishimoto intercooler features an 80% increase in core volume compared to the stock intercooler.
So that rounds out our project! We were very successful on all fronts of our testing and development. We are very confident that this intercooler will provide the temperature reduction needed for your vehicle. Check out a shot of this cooler viewed from the scoop!
Thanks for following along with the build! Feel free to follow up with any questions regarding the intercooler or our data!
Did you think we would just develop a top-mount intercooler (TMIC) for the GD WRX and leave out the STI? No way! At around the same time we began development on the WRX, we picked up a new shop vehicle, a 2010 STI. When a new vehicle makes its way into shop service, we need to be sure it accelerates and handles properly. The STI passed our checkup. We then set out to determine what products we wanted to develop first. We already had a few items on the market, including a performance aluminum radiator and a silicone radiator hose kit. Our first target had been a direct-fit oil cooler that proved to be extremely successful, providing a 20°F drop in oil temperatures during our testing. Next we created a plug-n-play fan shroud with dual 12” electric fans. The key component of this shroud was the use of stock-style fan connectors and a very slim profile. Finally, we tackled one of our favorite projects, a performance cold-air intake. This was actually the first performance intake we developed, and it was quite a learning experience. With a garage stocked with fabrication tools and testing equipment, we were able to prove a power increase of 25 whp and 25 wtq by simply bolting on this intake. Yes, you read correctly, no tune is required for these power gains, and AFRs remain well within spec. With all these projects complete, we now turned our attention toward improving air charge temperatures. So we began with a TMIC and would later move to development of a front-mount intercooler (FMIC). Keep an eye on our blog for information regarding both projects.
Check out our 2010 STI shop ride!
This vehicle has been great for hauling supplies, product testing, displaying our current products, and occasional lunch trips (don’t tell the boss). In fact, I have been waiting for my S4 to break in the parking lot (shouldn’t be long, it’s a C4) so I have an excuse to take it home.
So, on to the project at hand. We started by evaluating products currently on the market, because our product will need to stand out from the others in some way. We developed a few goals and then set our engineers loose to see what they could do.
Must be direct fit and require very limited vehicle modification
Support vehicles having up to 400 whp
Provide reduced IATs
Increase core volume
Improve power output
This is a simple goal for all the products we develop. There is nothing more frustrating than having to modify that brand new component you purchased. We want our customers to have a hassle-free installation, so we will be including any additional components needed for proper installation. We will be test fitting this intercooler on numerous occasions to be sure any bugs are worked out.
The stock top-mount is reasonably efficient; however, with higher-than-stock boost levels, IAT’s are prone to rise. Heat-soak is also fairly common with the tube-and-fin core used with the stock cooler. I am in no way implying that the stock intercooler is insufficient. Our goal is to provide improved IATs for stock vehicles up to our 400 whp power range. At any point above this, we highly recommend an efficient FMIC, which would eliminate any residual heat-soak and provide the temperatures needed for rotated or high-boost setups. As we have seen on many of our other intercooler projects, reaching 400 whp will not be an issue because of the space we have to expand this intercooler.
All about Coolness
Everyone wants to be cool; this is the primary target of this project. The reason for upgrading your intercooler is to achieve lower induction temperatures. If we fail on this front, the product will not reach shelves. We are hoping to provide concrete data of temperature improvements on both a stock and lightly modified vehicle.
Large and In-Charge
This goal ties into the others we have for power support and temperature reduction. The consensus is that a larger TMIC would help with both these components. That being said, there is certainly a sweet spot of size, heat transfer, and flow that would be the true challenge for this project. A core that is too dense is going to impact air flow, which may even reduce power. This component will require precision on the part of our expert engineering team.
An intercooler upgrade is a supporting modification, not technically targeted toward power increases. Obtaining additional power by bolting on a new cooler can be quite a challenge. We’re hoping that we can pull a few ponies out of our design via flow increases or increased timing from lower temperatures. Although this will be a pending or secondary goal, we are hoping to achieve it even if the numbers are not substantial.
Jumping right into the project! We needed to evaluate the stock intercooler, determine our design constraints, and begin creating a prototype. Let’s take a quick look at the stock intercooler.
The intercooler removed!
And the engine bay we are left with upon cooler removal!
The stock intercooler is constructed completely from aluminum, unlike the plastic end-tank cooler used on the WRX. The end tanks on this unit are cast and crimped to the tube-and-fin core with a rubber gasket to seal this connection point. Overall the cooler is rather well designed, fits well, and the flow is reasonably efficient. At one point during development we evaluated the Y-pipe system for the inlet and found that it is well designed. Airflow through this section is more than sufficient for our planned power level support, so we would not need to modify the Y-pipe in any way.
Additionally, as you see above, the intercooler features a dense tube-and-fin core. It will be interesting to see how the heat transfer capabilities compare to our planned bar-and-plate core design. The core dimensions are as follows:
Volume: 350 cu in
Total intercooler weight for the stock unit comes in at 10 lb, good information to keep in mind for those trying to keep weight down.
With all the stock dimensions noted and a vehicle at our disposal, our engineering team began designing a prototype using Solidworks software. Once complete and reviewed several times, we manufactured a prototype for testing purposes. Check out this raw prototype!
Check back next time for full details on our prototype unit!
With our finalized prototypes in hand, it was time for some testing! We needed to perform a back-to-back comparison of the stock oil cooler and the Mishimoto oil cooler. First, check out a few shots of our CNC-machined sandwich plate!
Let’s see how this guy looks installed!
Throw on a few banjo fittings!
And then we attach and route our lines.
Here’s a shot of the rendering of the assembled kit. These Solidworks software renderings are always neat!
Testing time! We spliced our temperature sensors into the lines and hit the road for some testing. Our testing conditions are listed below.
79°F–80°F with 62% humidity
65 mph cruise for approximately 8 miles
Identical tests on the same day conducted for both the stock setup and the Mishimoto oil cooler
Special care was taken to ensure a smooth path of airflow in front of the vehicle during testing.
Check out the sensors!
And the results are in! Check out our temperature and pressure charts below.
The addition of the Mishimoto oil cooler provided a 25°F drop in oil temperatures during normal driving conditions while lowering oil pressure by only 5 psi. Our engineering team predicts that even greater differences between the stock setup and the Mishimoto unit will be evident during track/aggressive driving and with higher-powered vehicles.
With successful results and our goals accomplished, we were able to close this project and put another successful oil cooler into production.
Thanks for following along, and feel free to follow-up with any comments or questions!
Keep an eye on our Engineering Blog for updates on more of the projects we are working on for the 2015 WRX!
The most obvious thing to do with your brand new WRX is to take it to the track immediately after leaving the dealership. Fortunately, the WRX is a rather competitive package right out of the box. Our initial driving impressions were very positive, and we have no doubt that this vehicle is going to make a big impact on the enthusiast world. We monitored fluid temperatures at stock power levels during normal driving conditions and found that everything seems to be in spec. We are assuming a majority of these vehicles will not remain at stock power levels for very long. Soon enough, many enthusiasts will be bolting on intakes and exhaust systems and supporting everything with a solid tune. As we have seen with our initial intake testing, the FA20DIT responds very well to modifications. We expect the 2015 WRX should be able to attain 300 whp with a few simple modifications.
With increased power comes increased heat. As we have said time and time again, high oil temperatures are detrimental to the longevity of your engine. Warm climates, increased power, and track/aggressive driving are all contributing factors that necessitate an oil cooling solution.
Our talented engineering team is not new to oil cooler development. Not only do we offer a direct-fit oil cooler solution for the entire year range of Subaru WRX/STI models, but we also cater toward a variety of other enthusiasts’ vehicles such as the S2000, BRZ/FR-S, 350Z, Genesis Coupe, and many more. The primary focus of these products is an easy-to-install kit that requires minimal to no vehicle modification. We want to take the guess work out of installing an aftermarket oil cooler. Along with offering an all-encompassing kit, we conduct real-world testing and then publish this information so our customers can appreciate the benefits our products will have on their vehicles.
Enough about our product line, let’s jump into this WRX! With less than 1,000 miles on our vehicle, we began thinking about new-product development. Our first products included some very nice silicone hose kits. As with previous generations we developed, our radiator hoses are constructed with 4-ply silicone, a nice upgrade in reliability. We also released a silicone ancillary hose kit featuring 13 coolant and PCV lines that can benefit from more durable material. Finally, we released our FA20DIT airbox hoses, a two-piece kit featuring the upper and lower hoses from the turbo inlet to the airbox. We are presently conducting some power tests to see if the smooth interior of our hoses provides any flow benefits compared to the accordion-style design of the stock units. More on this later!
Now, on to the goals for our next target, a direct-fit oil cooler. We do not release products without at least one goal in mind, such as reduced IAT’s from an intercooler or lower coolant temperatures with one of our aluminum radiators. This oil cooler would be no different, and our goals would be simple enough to allow our engineers to explore their creativity.
Project Goals: 2015 WRX Direct-Fit Oil Cooler
Must be completely bolt-on and require no vehicle modification
Must provide substantial temperature reduction without significant decrease in oil pressure
From a design standpoint, an oil cooler for the 2015 WRX provided some interesting challenges. I’m sure you are aware that the oil filter is no longer on the bottom of the engine. The FA20 utilizes a top-mounted oil filter with a cup assembly to catch any overspill. This design is very similar to the BRZ oil filter setup; however, the WRX utilizes a liquid-to-liquid heat exchanger below the filter to regulate temperatures. For our BRZ kit, we developed a specific CNC-machined spacer to fit with one of our oil sandwich plates. This strategy would not work for the WRX, and hood clearance became a concern when a spacer and a sandwich plate were added.
Instead of tweaking a preexisting product, we decided the 2015 WRX deserved a completely new sandwich plate of its own. This component would require the longest time to manufacture, so we targeted this first to get a jump on the completion of this kit. Bringing a product to market quickly is important to our team, as we knew enthusiasts would be tracking this one right out of the gate. Despite this, we did not sacrifice quality or due diligence in any way throughout the design process.
First the lead engineer for this project, Kevin, evaluated the dimensions of the factory oil cooler and filter assembly. He then created a sandwich plate design using Solidworks software, followed by 3D printing. This printer has been an extremely valuable asset for numerous recent projects. Being able to have a prototype in-hand for test fitment in less than 12 hours saves us a great deal of time. The fewer delays on our part means we can bring products to you quicker and get you out on the track having fun. Check out the prototype of our sandwich plate!
We then attached our standard oil sandwich plate O-ring.
And finally we attached our -10AN fittings and dowty seals to complete the mock-up unit.
Next, we installed the prototype of the oil sandwich plate on the WRX to decide on the lines we wanted to use and where to route them. First we tried using a pair of 45° fittings to route the lines downward and around the radiator. Although these lines fit, we found that clearances were quite tight with the overflow bottle when using a radiator that is thicker than stock.
These images provide a glimpse into the space we had to work with. As you can see, the 45° fittings forced us to twist the sandwich plate sideways for clearance – not what we wanted to do. You can also get a good look at the stock oil cooler and the coolant lines running in and out of it. Now check out a few shots with banjo fittings used instead of the 45° fittings.
The banjo fittings provide the clearance needed to mount our sandwich plate straight, and they even function with our thicker X-line radiator installed. We always strive to provide an entire line of products that fit together perfectly, so this was important. Banjo fittings will be the way to go for this particular design.
Now let’s talk about oil warm-up. You might notice that this sandwich plate does not utilize any internal thermostat. You can also see that we did not remove the stock oil cooler unit. Our engineers spent a great deal of time discussing and exploring different potential designs. We concluded that keeping the stock oil cooler intact and adding our liquid-to-air heat exchanger as a supplementary cooler would give us the best results.
The stock oil cooler also functions as an oil warmer. Coolant flows through the unit to bring engine oil to operating temperatures as quickly as possible. (Cold oil is just as harmful as oil that is too hot.) By retaining the stock liquid-to-liquid heat exchanger, we can assure a speedy oil warm-up. Additionally, our external liquid-to-air heat exchanger works to lower temperatures once it has warmed up. Although this will need to be tested, we found that this system works perfectly on the previous generation of the STI.
So, the basic design is settled. Our data will be the true test though, so stay tuned for that.
The next detail was mounting of the oil cooler. As with any liquid-to-air heat exchanger, optimal heat transfer requires proper airflow, and we want that heat transfer to result in lower fluid temperatures. The obvious choice was to mount the oil cooler in the front grille area. Airflow testing on numerous vehicles helped us to determine the optimal location. After much deliberation and testing, our team selected the passenger side of the lower grille as an ideal location. Mounting brackets were fabricated, and we then did a mock-up of our 19-row oil cooler. Check out a few shots below!
And now with the bumper installed!
And finally, the lines were routed. These will go behind the crash beam, around the radiator, and up to the sandwich plate. All our kits utilize preassembled stainless steel braided oil lines with -10AN fittings. These high-quality components ensure consistent, durable, and leak-free operation.
At this point we were waiting for the CNC-machined prototypes of our oil sandwich plates to arrive for testing. Once they arrived, we could install the entire kit, compile our testing data, and evaluate the results of the project.
Testing time! Well … almost. We had a few quick testing details to sort out and some preparations before we actually began. First, we needed to drill and tap the stock intercooler as well as the Mishimoto prototypes so we could install our pressure/temperature sensors. Additionally, we decided to create a shroud for our TMIC fan to ensure that all cores would receive appropriate airflow on the dynamometer.
Tapping the intercoolers for sensors is a relatively straightforward process. We drilled into both end tanks and then tapped them to 1/8” NPT for our sensors. The stock intercooler tanks are thick enough so we could adequately tap them for testing. Check out a few images of our coolers getting this treatment.
Once the coolers were set, we prepared our fan rig for operation on our test vehicle. Intercooler airflow differs substantially when you compare fan use to actual wind speed on the road. Our temperature testing would be conducted on-road to obtain the most accurate data. Power data collection would be conducted on our dyno, so we needed a shrouding setup that would optimize our fan.
First, let’s take a look at the footprint of our fan and understand why this needs to be done.
As you can see, the fan features an oval shaped outlet, while our intercoolers are square. When placed on the cooler during testing, a ton of airfow misses the actual core and would skew our results. The previous generation TMIC was more rectangular, which catered a bit more to our fan outlet. To produce maximum airflow in a directed manner, this shroud would need to be completely sealed, similar to the hood scoop ducting on your WRX.
To start the project, Dan welded a frame for the outlet.
Once complete, shrouding was attached to convert from the oval outlet to the square shape of the intercooler core.
After everything was assembled, the gaps were sealed and weather-stripping was applied to the contact points to avoid damage and to provide a press-on seal for the shroud. We then strapped the fan and shroud assembly to the car.
We also needed to consider the difference in core size between the Mishimoto and stock intercoolers. The Mishimoto intercooler is 1.5 inches higher, so we have a removable adapter to reduce the size appropriately. This ensures that both cores are receiving equal amounts of airflow from the fan.
Now that we were all set for testing, it was time to bring in a vehicle. Our project manager, Jason, is very involved with the Subaru community and talked his friend into bringing his 2012 WRX in for a few days. This vehicle was lightly modified, essentially a stage-one WRX. This would be adequate for testing and provide a nice idea of what a majority of our customers should expect when bolting on this cooler.
Check out the car!
Our first goal was to capture baseline data for the stock intercooler. After bolting up our tapped stock cooler, we put the car up on the dyno to prepare it for some pulls.
And the pulls began, filling our garage and offices with the wonderful tone of an EJ!
As mentioned before, we designed two cores with varying densities for testing. Once we had our baseline data, we installed the first of our Mishimoto prototypes and hooked it up to our sensors. We made pulls until we had a nice average for our data collection purposes. For consistent and accurate comparisons, we ensured that the test vehicle was properly warm by timing the breaks between pulls. We then repeated the testing on the second prototype.
We spent the day making pulls and collecting data. Once completed, we compiled our findings and determined which core we would use to move forward. Check out a neat video of the dyno testing!
Check back next time for the results of our testing!
As I am sure you are aware, we have been working diligently to develop innovative components for the new 2015 Subaru WRX. To continue our video series regarding the new model, we have captured some of the new products in various stages.
Check out these videos featuring our in-house Subaru expert, Jason, as he provides an inside look at products we already have on the market, as well as two unique products we are currently working on.
Last time, we left you with a teaser of the rendering for our prototype of the top-mount intercooler (TMIC). Since then Dan has been hard at work refining his design. Our team worked up some CFD analyses of both the stock intercooler and the Mishimoto prototype. And, we also worked up a few functioning prototypes so we could begin real-world product testing. This project is moving quickly and our team is rather excited to see the outcome. Take a look at the finalized rendering of our cooler below!
The core we designed is rather dense and features an offset fin style for greater heat transfer contact points. Despite the density, we saw very similar if not better flow results compared to the stock cooler. Take a good look at the comparisons below.
So what do all these colors mean? The dot is a graphical representation of airflow. The dot color denotes the velocity of air at that point in the cooler:red (highest velocity), yellow, green, and blue (slowest). This analysis was conducted with 22 psi at the inlet. Our goal here is to improve velocity through the core while still allowing for appropriate heat transfer. The results are very positive!
Now that we had successful flow data with our desired core, it was time to work up a functional prototype for real-world testing. All our products undergo testing to ensure proper functioning. Our facility is well equipped with a Dynojet dynamometer and fabrication equipment so we can modify our design as needed during testing.
We made two prototype coolers, one black and one silver. We designed two very similar cores with slight differences in density. Check out the much anticipated prototype below!
Initial impressions: This cooler is very well designed and constructed. The cast-aluminum end tanks are extremely smooth on both the interior and exterior surfaces. The core looks great and our branding is visible, yet subtle. The primary concern with fitment and design was the two flanges that this cooler features: the BPV flange and the hot-side turbocharger compressor outlet flange. Take a closer look at these flanges below.
Both flanges are CNC-machined and provide completely flat surfaces. The hot-side turbocharger connection flange has provisions for the stock-style O-ring seal. To save you the trouble of locating one of these gaskets or reusing the old one, our new TMIC will include a new gasket.
Next, it was time for a few quick comparison shots of the stock cooler and our prototypes.
The main points of visual comparison will be in the physical size. See the images below where the stock cooler is dwarfed by the Mishimoto prototype. First up is the height comparison.
The next important measurement is the thickness. Remember, the stock core is 2.5” thick.
The Mishimoto prototype intercooler is a full inch thicker than the stock cooler. The total core dimensions are 11” x 11” x 3.5” for a total volume of 423.5 cubic inches. The Mishimoto intercooler provides 162.25 additional cubic inches of volume for a total increase of over 60%!
Now that we have our prototype coolers, it is time for some testing and performance data collection. Check back next time for our testing preparation and initial round of results.
Nothing sets off a Subaru engine bay like a nice beefy top-mount intercooler. Not only does a big TMIC look great, but it is also instrumental for keeping AITs low and your EJ happy. For those seeking low AITs on a stock to mild build, an efficient top-mount setup is the way to go to avoid inheriting the pressure drop and lag from a front-mount intercooler. Not only this, but the TMIC prevents any road (or off-road) debris from damaging your heat exchanger. This is key for those who are involved in any rally, off-road, or rallycross events.
If you keep up with any of our social media outlets, you are likely aware that Mishimoto has been working to develop a TMIC for the 2008–2014 WRX. We are very active within the Subaru enthusiast community, and a majority of our best-selling products are designed specifically for the WRX/STI. We cater to all makes and models, but it would be tough to deny that we are slightly partial to Subaru, especially considering that our three shop vehicles are a 2003 WRX wagon, a 2010 STI, and a brand-new 2015 WRX.
With the recent release of our WRX GD top-mount, which has been very successful, we decided it was time to tackle the WRX GR and take a look at the new flange-style intercooler setup. Although typically well known for our aluminum radiators, Mishimoto has been designing and manufacturing performance intercoolers for quite some time. Our universal line of coolers find their way into some really neat builds, and more recently we have put some of our focus into direct-fit intercoolers. With more and more turbocharged vehicles reaching production, this really opens the door for our team to innovate on the stock setup and provide the cooling needed for those who modify their vehicles. Aftermarket tuning has found a way to get the most out of these stock turbo vehicles, with massive power increases from the press of a button. There is one issue, however: Factory intercoolers are designed to efficiently cool stock power and temperature levels. This creates a need for an upgraded heat exchanger to manage intake temperatures and heat soak. This is where our team can help you find a solution for dealing with high temperatures.
Now to get vehicle specific! You are probably aware that the WRX GR utilizes quite an interesting stock intercooler, so let’s get a good look at what Subaru put together for your WRX.
This is a pretty typical plastic end-tank intercooler with a fairly decent tube-and-fin core. The main difference as mentioned before, is the use of a flange-style connection point for the turbocharger compressor outlet. The BPV mount flange has a similar diamond shape to the previous generation. The use of plastic is a bit concerning, as we have seen several higher-boost vehicles separate the tanks from the core of these coolers. This is not something you want ruining a day at the track.
Based on the core design and placement near the turbo (not to mention the direct contact with the compressor housing) heat soak is certainly going to be an issue with this cooler. Check out the BOV flange and end tank close-up below!
You will notice that our end tanks have openings. This is for the installation of our pressure and temperature sensors for testing. We will get into testing a bit later down the line, so stay tuned for that. Now, another key thing to consider is size. With greater volume comes increased efficiency (when correctly designed) and higher power support. The factory core measures at 11” x 9.5’ x 2.5’ for a total of 261.25 ci. The entire intercooler assembly weighs 6 lb.
To get the ball rolling on this project we would need to first evaluate the stock cooler. Our plan of attack was to create a 3D model of the stock intercooler and test its flow characteristics. We could then use the dimensions and information from that model to design our intercooler. Pretty straightforward, yet time consuming. Check out a few cool shots of the process.
The Romer arm we use is portable, which allows us to collect data points when the component is both on and off the vehicle. With data points collected on the vehicle, we can provide a better look at exactly how much larger we can make this intercooler. Once we determine this, we can decide how large the intercooler needs to be to meet our goals with the projects. This doesn’t mean we will be leaving potential power/temperature benefits on the table. Our goal is to provide the highest gains possible for the largest possible range of vehicles/modifications. Despite this, we are aware that a TMIC begins to lose its efficiency at a certain point, necessitating the need for a front-mount setup. We have no reason to make wild claims that our intercooler will support 500–700 whp. We want you to get the most out of your vehicle, even if it means explaining that this cooler is not sufficient for your goals. Not to worry though, as we are also working on a very slick front-mount setup for this particular vehicle.
Now, take a look at a few cool shots inside the end tanks of the stock cooler.
Once we had all the necessary information from the stock cooler, we could begin to discuss the design aspects of our first prototype. As with any project, we would need to lay out a few basic goals to be sure we are meeting the needs of our customers.
Must be completely direct-fit for the 2008–2014 Subaru WRX, and utilize the factory-style turbocharger flange connection.
Support vehicles from stock power levels to 400 whp.
Provide an efficiency and AIT reduction compared to the stock intercooler.
Provide improved power output compared to the stock intercooler.
Increase intercooler volume to support higher power and heat.
Fitment is key when developing aftermarket components. Many of us have had an experience with a so-called direct-fit product that required trimming, cutting, and drilling, resulting in a ton of frustration. Our primary goal is always perfect fitment, especially with a product that requires extreme precision. Now that we’ve captured all the data points from the factory cooler and its mounting points, it should be a simple process to match the critical dimensions. As always, we will be test fitting this product on multiple occasions to ensure there are no issues.
The tricky part of this particular cooler is the unique two-bolt flange that secures the turbocharger compressor outlet to the intercooler. This connection utilizes a high-temperature rubber seal that we would need to emulate as well.
The BPV mount flange will also require precision. This flange is rather unique and will need to be replicated so it will accept the factory bypass valve.
Our research has found that a majority of enthusiasts with vehicles having more than 375 whp are selecting a front-mount intercooler to cool the air charge. Because of this, we will be targeting the range of stock to 400 whp for this particular cooler. By solidifying a range of power levels, we can really hone in on our core composition to provide optimal cooling and flow. We will have to properly balance core density to take advantage of our plans for larger volume. Our engineering team has tackled a similar project with our 2001–2007 Subaru WRX/STI intercooler. After testing a few different cores, we saw the effects of our changes on power, pressure, and temperature, enabling us to select the core that would provide the best results for our goals.
Cooling is another top goal here, obviously the primary objective of a heat exchanger. We want to see AITs plummet compared to the stock intercooler. Increased boost pressures and higher power numbers will generate unwanted heat, which requires a remedy to avoid heat soaking the stock cooler. We are hoping that our cooler is the solution to these high-temperature issues.
Jeremy Clarkson says it best: We want more speed and power, and I am sure you do to! Swapping intercoolers does not always imply that more power will result. An intercooler upgrade is more of a supporting modification. That being said, with all our other intercooler projects we have managed to squeeze out power compared to the stock intercooler. We are assuming that 5–10 whp is certainly an achievable bolt-on number for this cooler.
Make It Big
Going back to the first sentence of this post, everyone wants a nice thick intercooler that invites onlookers to peer into your engine bay in awe. In all seriousness, the increased volume will help us provide a greater cooling surface area, more heat transfer points, and increased volume, which will help achieve our other goals noted above.
Back to the project progress! Once we had collected our points, the lead engineer for this intercooler, Dan, set about to model the stock intercooler. This would allow us to compare flow through the core with whatever design we actually developed. Check out his work below!
That’s it for this portion of the build. Next time we will be showing our new prototype design and compare that to the factory cooler using CFD software. This will allow us to test the flow of both the stock cooler and our prototype to be sure our increased core density does not affect flow in a negative manner. Check out this teaser of our prototype design!
Finally, we made it to our day of testing! We had previously confirmed fitment and made a few minor adjustments to some external end-tank features to improve coolant flow and ease of installation. A member of E90post was kind enough to lend us his vehicle for this round of testing. This 2007 sedan was equipped with a manual transmission, JB4 software and an intake. This vehicle would be ideal for collecting testing data, we have found that the “Average” E9X owner has similar modifications.
First, we needed to remove the factory unit and replace it with the Mishimoto radiator. We would be testing our unit first followed by the OEM radiator. Our main goal here was to gather comparisons of several features of the radiator as well as temperature data. Check out a few shots of the Mishimoto unit next to the factory radiator we removed.
It is nice to see a BMW with an electric fan, no more antiquated mechanical fan!
Another shot of the stock radiator. It seems this one has built up some fin debris. The vehicle owner is certainly driving his car! This debris will have an effect on cooling performance and heat transfer, it may be wise to check yours out every now and again to be sure buildup is not excessive.
Once we had the old unit out, the Mishimoto radiator was installed. All of the adjustments we made to the previous prototype were a success and this unit fit like a glove. Check out the Mishimoto radiator in its new home.
And finally we could tap the factory hoses for our sensor bungs and install our testing equipment.
Now first, let’s cover the physical specs of the Mishimoto radiator compared to the factory unit.
A few more additional points of important information in regards to cooling efficiency and heat transfer.
Mishimoto radiator provides a 35% fluid capacity increase
Mishimoto radiator provides a 25% increase in total core thickness
Mishimoto radiator provides 10% greater cooling surface area
At this point it was time to review our temperature data comparisons from the stock radiator to the Mishimoto unit. This is where things become tricky for a variety of reasons. We are completely transparent regarding our testing processes as well as our results, nothing has been skewed in any way to improve the appearance/value of our products. After reviewing the data, we came to the conclusion that the inlet and outlet temperatures of both radiators were nearly identical. This may be shocking based on the specs listed above, but a few things regarding testing conditions need to be considered.
Ambient Temperature: Our data collection was compiled with ambient temperatures in the low 70°F range. This is not ideal for recording coolant temperatures in an extreme manner. Preferably we would perform this test at a minimum of 85°F to see the result of this on heat transfer.
Testing Process: Keep in mind the vehicle we are using is being loaned by an enthusiast. All of our driving (off the dyno) is performed in a normal manner and on public streets. This places far less stress on the radiator and cooling system, which certainly effected our results. For this particular test we drove on the highway for five minutes with each radiator at a set speed of 65 mph with a clear path of air in front of the vehicle. This testing is not exactly capturing the capabilities of a performance radiator. Ideally we would have a test track, our own vehicle, and we could really put the car through its paces to see what the effects would be on temperatures.
These factors all played an important role in the data we collected. We are not making excuses for the performance, just providing an educated explanation of what is happening. With our radiators capacity increase, surface area increase, and heat transfer improvement we are 100% confident that this radiator will outperform the stock radiator in both a warm climate and during aggressive driving.
Our plan is to circle-back to this testing during the summer (Temperatures should hit the 90’s) and see the effect on our data. Additionally, if anyone would like to volunteer their time to test this product on the track, we would be very pleased to discuss setting that up.
We performed similar testing on our new E46 M3 radiator. Our team designed a fantastic product and flew down to Orlando over the winter for track testing in a warm environment. Check out the details here: http://engineering.mishimoto.com/?cat=81.
As of now, we are confident with the specs and performance of this radiator design and will be moving forward with mass production. We will have a better idea of release date and pricing within the coming weeks.
Along with our radiator testing during this particular week, we also began development of an oil cooler setup for the E9X. Check out this teaser shot!
Thanks for following along! Feel free to follow-up with any questions.
Thanks for returning for the second portion of this build! As mentioned in the previous post, our team was prepared to conduct extensive testing on this induction hose. Our main goal with testing was to determine the particular combination of components that would provide the most power output. Power output collection would be via our Dynojet. Each combination of components received an average output based on three consecutive runs all completed on the same day with essentially similar conditions. The variations of testing are below.
Resonator and noise amplifier both installed
Resonator plugged with noise amplifier installed
Resonator attached with noise amplifier disabled
Resonator and noise amplifier both plugged
We would be testing each of these conditions with both the Mishimoto prototype induction hose and the factory induction hose to provide a variety of results for all combinations.
Our test vehicle is the 2013 BRZ you see below. This particular example was stock (engine-wise) and featured a manual transmission. A big thank you to our test vehicle donor for the use of this BRZ!
Once we had the BRZ strapped to the dyno, it was time to get the vehicle warmed up and to check that all our readings were functioning. Check out the engine bay shot below with the Mishimoto hose installed!
It was quite a day! We made pull after pull and swapped components in between each set of runs to provide the information you see in the chart below. As mentioned earlier, this testing was going to be a learning experience for our team. We also wanted to educate the community about the differing power outputs based on the removal of the noise generator and resonator. Enough talk, check out the chart!
So let’s break down this data in an easy-to-digest fashion! First portion: OEM intake with altering intake components.
The completely stock setup put down consistent runs of 167 whp and 133 wtq. This is around what we expected from the BRZ. Also note that we are capturing noise level in terms of decibels (dB). This was captured in the cabin of the vehicle for comparison purposes. Despite the purpose of the resonator, it appears that the most power output returned from the factory induction hose occurred when the resonator was removed and the noise generator was retained. The removal of the resonator increases intake noise quite significantly as well. Although horsepower was unchanged, torque showed a slight increase, and dB measurements increased significantly. It should also be noted that we saw a 2 whp power loss when both the resonator and noise generator were removed. Now, take a look at the chart for the Mishimoto induction hose testing.
This is the fun part! With the Mishimoto induction hose installed and all other factory equipment remaining, we made 2 whp and 2 wtq compared to the factory hose. This is an impressive increase! Our continued testing showed an additional horsepower increase with the noise generator removed, bringing our total gains to 3 whp / 2 wtq. For a component that installs in minutes, this was an easy way to achieve bolt-on power. We could have stopped at this point, but we decided to perform some additional testing. All previous testing was completed using the factory paper air filter. We installed an aftermarket high-flow air filter to see what further increases we could obtain from testing. Check it out below.
As you can see the filter alone adds 2 whp to the factory setup. We tested this with the Mishimoto induction hose and were pleased to report a 6 whp and 3 wtq gain compared to the completely stock setup. All our testing results point to the resonator as a supplement for performance with the Mishimoto induction hose. We found that removing the resonator had a negative effect on power output with our induction hose.
With our findings from testing it was time to decide on a final iteration of this hose to move it toward mass production. We decided that our hose would provide provisions for the resonator, but we would be eliminating the sound generator. We knew that in general, the enthusiast community prefers to eliminate the noise generator. This strategy provides the best power output, which was our primary goal. Additionally, we will be providing a plug for the resonator for those who wish to eliminate the resonator and gain some additional intake noise.
Our engineers also evaluated the decibel readings based on the components installed to the induction hose. We found that there is a definite pattern in our test results: The resonator works to eliminate noise, while the noise amplifier adds noise. We saw this essentially as a 1:1 ratio; the noise generator added 2–3 dB of sound while the resonator removed 2–3 dB. Essentially these two components cancel each other out when installed. Check out the graph below depicting this information!
With all our testing complete, it was time to assemble a final prototype and ensure perfect fitment. Check out a few installed shots below!
The fitment was perfect and it was time to move these hoses into production! Now, let’s recap the project goals and see how we did with this one.
Silicone material should be used for increased durability.
The final design of the Mishimoto induction hose features five-ply silicone with internal metal support rings. The five-ply silicone provides unmatched durability compared to the factory counterpart and will provide improved resistance to heat, pressure, oil, and fuel. The internal metal support rings will eliminate any chance of this hose collapsing under pressure. This hose is built to last!
Potentially include ports for resonator and/or sound generator.
Our testing showed that the noise amplifier consistently reduced power output while the resonator provided increased power. With this information we designed our induction hose to eliminate the noise amplifier, yet still have provisions for the resonator. For those who wish to eliminate the resonator as well for greater intake noise, we include a CNC-machined plug for the hose.
Test the effects of the sound generator and stock intake resonator.
Our testing results showed that the sound generator hampered performance while the resonator supported performance. We tailored the design of our hose to the results from this testing.
Hose should install like the factory unit.
The Mishimoto BRZ/FR-S induction hose installs just like the factory unit and is an easy way to add horsepower in minutes!
Offer in a variety of colors.
After a lengthy discussion our team decided to move forward with black, red, and blue options for the induction hose.
With our goals met, this project could finally be closed. Our team has deemed this product a complete success, with the hose being a great upgrade to the factory unit. Either alone or when combined with a high-flow air filter, this induction hose provides a nice increase in power that is certainly noticeable. Check out a few of the major benefits of this hose below!
Proven gains of up to 5 whp
Internal metal support rings protect against high vacuum and collapse
Direct fit for the FR-S, BRZ, and GT86
Installs in a matter of minutes
Multiple configurations using the included CNC-machined resonator plug
Five-ply silicone for added strength
Available in black, red, and blue
Mishimoto Lifetime Warranty
And check out a few more images of the completed product!
Along with all this development and testing, our team put together a few neat videos for this particular hose. Check them out below!
Features and Benefits
Thanks for following along with our build-thread! Feel free to follow-up with any questions or comments!