Reason:

Since the intercoolers are responsible for bringing the temperatures of the post-turbo intake air down as much as possible, the more effective they are at accomplishing this task, the more power the car should produce. With the intercoolers exchanging the heat from the intake air to the outside air, a higher transfer rate should enable cooler temperatures. I was interested in finding out what, if any; affect cleaning the intercoolers would have on their performance. My motivation for this test was the suggestion that cleaning the intercoolers intuitively must improve their efficiency. Following the replacement of my stock intercoolers with aftermarket IC’s, the opportunity was presented to test this theory.

Test Setup:

My stock intercoolers had been on my S4 for four years and six months without ever being cleaned on the inside, and the outside front face only sprayed with water on a couple of occasions. After removing the IC’s I decided to test them without cleaning, and after cleaning the inside and outside.

In order to test them I did the following. I attached the intercooler shroud, and taped over the remaining exposed section that normally would be blocked by the car’s structure. I attached an IC hose to one side and placed a hair dryer exhaust in this hose, then filled the open space around the hair dryer with newspaper to seal the hose off. On the outlet side of the intercooler I stuck the outside temperature probe from an indoor/outdoor thermometer far enough inside, but without touching the intercooler, to allow me to record the temperature of the escaping air.

I positioned a small portable (~15″ diameter) fan approximately 18″ in front of the intercooler shroud to provide cooling.

The test was begun with the intercooler and shroud placed in an oven for 4 minutes at 170 degrees F. This was to ensure that the intercooler was warm when the test began. A clock was started when the intercooler was removed from the oven. By one minute and twenty seconds I had moved the IC to my test location, attached the hair dryer assembly, the thermometer, and positioned the fan. At 1:20 I turned the hair dryer on and recorded the temperature. I repeated this at 2:20, and then at 3:20. At 3:20 I also turned the fan on high. I then continued to record the temperature reading every 20 seconds until the temperature change during a 20 second interval was approximately less than one half a degree. This I assumed was an indication that the temperature had stabilized near steady state.

I performed this test twice with the intercoolers dirty. Then I used some degreaser to clean the inside of the IC. After two applications of degreaser I washed the IC and used a pressure washer on the outside surfaces. I then ran the hair dryer through the IC from both ends for several minutes to dry out the inside.

Then I repeated the test for times. During two of the tests, when the IC was clean, I applied a water spray to the IC front face as the fan was turned on, to try and simulate an IC water sprayer.

Limitations:

With this test set up there were a number of variables that were difficult to hold steady. I had to assume that the temperature from the hair dryer would remain constant throughout the test. The placement of the IC in front of the fan after removal from the oven was not repeatable; there would be some small variation in location and distance. Additionally the temperature of the IC following removal from the over could not be verified. After the cleaning of the IC I could not verify that all of the water had been dried out of the inside. If someone were so inclined they certainly could have continued to apply degreaser to the inside of the IC until they were very confident all of the oil and dirt had been cleaned out. I was satisfied with two cleansings. Time constraints limited me from doing a larger number of tests that may have provided more data that would help to reduce variability.

Results:

The charts below show the results of the six tests with the outside air temperature plotted at 64F.

The second chart shows the average of the dirty runs by the heavy line, bordered by the standard deviation in the thinner line. The last two clean runs are plotted as well. I used the last two runs because of concern that there was still some moisture inside the IC on the first two clean runs.

Conclusions:

There are a few things I was looking for from these charts.

1. Were there any anomalies that might indicate an error with the test procedure? The fact that the first two clean runs produced substantially lower temperatures led me to believe that some moisture was still present in the IC, aiding in the reduction in the air temperature.
2. Is there any indication that the dirty IC heats up more rapidly than the clean? Answer: Initially, within the first minute, there does not appear to be a difference in the rate at which the intercooler heats. After the first minute it appears that the dirty intercoolers heat up slightly faster.
3. Do the clean IC’s cool off faster than the dirty IC’s? Answer: They do not appear to. The rate at which the dirty and clean IC’s cool appears to be nearly the same. What seems to be happening is that the dirty IC’s begin to heat up more rapidly than the clean IC’s after a while, resulting in a higher starting temperature once the fan is turned on. The IC’s then cool at the same rate, resulting in the dirty IC’s being at a higher temperature at any given time during the cooling phase. A question to ask would be how long would it take for the dirty IC’s to cool to the same temperature as the clean IC’s. Because I only took the test out to the point where the temperature differential dropped below 0.5 degrees per 20-second interval, I cannot draw a conclusion based on the data I collected. Looking over the shape of the curves toward the end of the test, it appears that the temperatures will eventually reach steady state, and the dirty IC’s will be at a higher final temperature.
4. What is the probability that there is a difference in performance given the results collected? I took the data collected and applied a t test of hypothesis for the mean. The hypothesis was that the mean temperature difference between the dirty and clean IC’s was greater than or equal to 3 degrees F; the average difference over the timeframe the test was conducted. At a significance level of 0.05 this hypothesis was not rejected.

Final remarks:

There appears to be a reasonable likelihood that the cleaned intercooler operates more efficiently than the IC that was allowed to accumulate dirt for 4.5 years. Whether or not it is worth the effort to remove the intercoolers from the car to clean them, a large job for sure, for a three-degree benefit, is the question to be asked if one is considering cleaning their intercoolers.