Interested in purchasing our WRX/STi top-mount intercooler? Check out our product page for more information!
Finally in the comfort of our own garage, we could begin further testing and data collection with our intercooler design. Below is a first look at the Mishimoto intercooler fully installed!
I may be biased but in my honest opinion this cooler really has a robust, beefy appearance and feel. The unique end tanks definitely set this intercooler apart from any other top-mount cooler currently available.
Once we returned to our facility we compiled the data from our initial testing. These initial runs were very positive. The Mishimoto cooler was making great power gains compared to the factory WRX cooler. Check out the dyno plots below!
In the last two runs we cranked the boost up slightly to see the effect on power output and cooling efficiency. This explains why these plots are slightly higher than the others.
It was now time for some additional testing. We reached out to the Subaru community and received a ton of great feedback as well as volunteers for testing.
Our research showed that a substantial portion of modified WRX models were using the STI cooler, and it was actually deemed rather efficient by online communities. The STI core seemed to flow reasonably well, and it supported mildly modified vehicles with little heat-soak problems. We decided to pick-up an STI cooler ourselves and see how it stacks up to our new design.
The first step was to drill and tap the STI cooler for our pressure and temperature sensors.
In addition to testing both the WRX and STI coolers, we would need to acquire a variety of vehicles to evaluate how our cooler reacts to different engine displacements, boost pressures, turbochargers, airflow, and power levels. This was quite a task! It involved a host of flat-fours belching out some nice exhaust tones on the dyno for weeks. What a treat!
One of our first vehicles for the test fit was this very clean 2006 STI shown below. This vehicle was almost completely stock; the only modification was an aftermarket intake.
With the stock intercooler setup we saw around 250 hp to the wheels. We also captured pressure and temperature data to compare later with our intercooler design. After collecting data from several controlled runs, we removed the STI cooler and bolted in the Mishimoto cooler and silicone intercooler hose. Check it out!
Everything fit very well with zero complications. Our sensors were installed on the hot side of the inlet hose and on the end tank of the cold side. One concern during initial design was the hose placement near the turbocharger and the effect the radiant heat would have on the hose.
The silicone hose cleared the heat shield perfectly. Although close, this shield will provide ample protection against the temperatures generated by the turbine housing. We highly recommend using the OEM heat shield when possible. If the shield is not used, a turbocharger blanket would also provide sufficient defense against the heat. Later, we will have temperature data for the effects of the heat on the silicone hose. Stay tuned!
We then checked for optimal fitment between the factory hood scoop and the Mishimoto intercooler. This scoop must seal properly with the cooler to prevent any cold air from escaping around the edges of the cooler. Once closed, the cooler sealed perfectly with the OEM shroud design, so no further modification was necessary with this component.
Next, we ran a few pulls on this stock STI using the Mishimoto intercooler. Instantly we saw the benefits of the efficient core and improved airflow. The Mishimoto cooler showed average gains of 5–6 hp/tq, with maximum gains of 9 hp and 9 tq. The plot below is an average of three runs for each cooler. All runs were completed back to back with a consistent wait time between each run.
The Mishimoto cooler provided gains across almost the entire rpm range, with its peak difference at just under 6,000 rpm. While this is fantastic and we were very pleased with the results, our top priority with this project was to reduce AITs. The Mishimoto cooler reduced AITs by 10˚F compared to the STI cooler. Ten degrees might not seem substantial, but modified vehicles have significantly greater inlet temperatures, so even 10 degrees would enhance the effectiveness of this cooler. Check out the chart below!
This testing was a great start, but we still wanted to make a few changes. After taking a look at the overall design and evaluating real data, our engineering team decided to make a few minor improvements to further increase airflow. We had two primary goals. First, we wanted to enlarge the inlet/outlet to achieve greater flow. We noticed that the castings, while robust, were rather thick, which provided a smaller inner diameter than we preferred. By opening the inlet/outlet we could squeeze out more airflow.
Second, we felt that the heat-transfer capabilities could definitely be improved with an adjustment to the composition of the core. To find the ideal core, we will test numerous designs featuring different fin pitches, fin heights, bar sizes, etc., to see the effects of each change. By doing this we can learn for future projects as well. Check back next time for design modifications and our next round of testing!