Restriction in the Stock BRZ/FR-S Intake – Results


We wanted this testing to be performed on the road to obtain real world data, as opposed to on a dyno. This method would allow the air dam to obtain actual flow to be received from moving at realistic speeds on the street. The conditions were less than ideal for tire grip (28 degrees F), so tests with tire spin were immediately thrown out and retested. However, since we’re measuring differential pressure the high density of the air due to the low temperatures has no effect on the overall pressure reading.

The test was performed the same each time, on the same stretch of road. The road was uphill, which is beneficial to increase the time of each pull in order to have a better chance of obtaining accurate sample data to combat the low sample rate of the manometer’s datalogging capabilities. We started off in first gear, rolling into the pedal to wide open throttle to avoid wheel spin, shifting at 7300rpm into second gear, straight into wide open throttle, shifting again at 7300rpm, and immediately into wide open throttle through all of third gear. Each run took approximately 14 seconds to complete. We performed this test 3 times for each configuration, measuring pressure drop from:

  1. Snorkel inlet to airbox inlet
  2. Front of airbox to rear of airbox (filter)
  3. Rear of airbox to entry of intake elbow (MAF housing)
  4. Entry of intake elbow to throttle body
  5. Snorkel inlet to throttle body

These runs were performed back to back on the same day, stopping each time briefly (less than 5 minutes) to save the datalog file to the computer, and/or to change pressure test locations on the intake tract.


grimmspeed stock airbox testing

This graph shows what happens across first through third gear, which is clearly shown by the fact that all five components have three clear humps, each with longer durations. These occur during wide open throttle, and the dips show the pressure approaching 0 between shifts. This graph also shows why having such a low sample rate makes for poor data, but we’ve made up for it by increasing the amount of trials. The fact that the graph maxes out for the overall system at about 9.5in of H2O in all three gears shows that that value is most likely correct for the overall system. Same goes for each of the individual components of the system; in each gear they seem to have the same maximum value. The graph also shows that restriction increases as RPMs increase, because as RPMs increase so does the required flow rate. The short duration of first gear shows the weakness of the sample rate, as the peak numbers of the individual components do not exactly match the peak numbers of each component in second and third gear. For this reason, the graph is most accurate for the third gear section (approximately 9 through 14 seconds), and shows a nice curve instead of a quick peak. However, for illustration purposes, showing all three gears shows that the pressure drop is RPM dependent and not speed dependent as would be initially expected. One would expect more air in the front air dam from the increase in speed to change the results in each gear, but clearly it does not.

This graph also does a good job of “double checking our data.” Remember that the orange line (Snorkel to Throttle Body) is the overall restriction of the system, and that it is the sum of the individual components. The graph of this curve is real world data, and is not simply the overall curves added together in Excel. However, if one were to measure the peaks of each gear for each individual component, and add them up, they would find that they total up to about 9.5in of H2O, which is what is shown to be the peaks of the overall system.

The effect the snorkel had on the system was very interesting. The snorkel showed a consistent pressure gain of about 3in H2O. The fact that the inlet is smaller than the outlet lends that the decrease in velocity of the air as it passes through should increase the pressure. However, the fact that this number is nearly high enough to cancel out any one other component’s restriction shows that in stock for the intake is very well designed. Each other component seems to have a restriction of about 4in H2O (air filter, MAF housing, intake elbow).


This answers a lot about the perceived weakness and the performance of the stock intake. It also goes to show that since the pressure drop doesn’t seem to be dependent on vehicle speed that all of this testing could have been performed stationary while strapped to a dyno. Removing the snorkel should yield no performance gain, but leaving it in could be compared to removing the air filter, or the MAF housing, or having a lossless intake elbow. However, all of these perceived restrictions really are not that bad. Each component having a restriction of 4in of H2O is really only equivalent to about 0.144psi, with the total intake’s restriction being equivalent to about 0.342psi. If I were to perform some completely fake equivalency math, and say that this car makes about 165whp in stock form, and at one atmosphere (14.7psi), a restriction of this size would be equivalent to about 3.84whp. So we would expect to see a gain of only about 3.84whp if we were to create a completely lossless intake system that acted at the exact same air to fuel ratio. However, luckily that math is completely fake, and just for illustrative purposes as there have proven to be larger gains than that achieved without creating a theoretical “lossless intake.” This is true because there are so many more contributing factors than just reducing pressure drop on an intake that is already well designed.

Chase – Engineering

Restriction in the Stock BRZ/FR-S Intake – Equipment


Now that we’ve identified potential restriction, we’ll want to use what we know about differential pressure to determine just what effects these potentially detrimental features actually have on the intake system.

grimmspeed cold air intake testing brz fr-s
This is the differential pressure data logger that was used in this testing.

Anyone could perform the testing for differential pressure as it is relatively easy to do. Since we’ve already located what we believe to be potential points of restriction we know exactly where we should tap into the intake system to gather pressure data. One could accomplish this extremely cheaply by making their own manometers out of water and tube, and it has been done before. However, we did not want to rig up two of these (as they are usually large and hard to read) and spill them all over the place while doing first through third gear pulls. Instead we acquired a digital differential pressure manometer, specifically an “Extech HD750.” This was chosen for it’s low range (5psi) which lends to it’s accuracy, as well as the fact that it can datalog. Being able to datalog was ideal because we can show a chart of what is happening as we row through the gears, which is infinitely more interesting than if we were just to post peak numbers. This will also show if the pressure drop effects are due to speed of the vehicle, or if they are rpm dependent. Unfortunately, the sample rate is only 1 sample per second, which means that a longer pull is necessary to get a better data set, but this can be alleviated by performing more pulls. The units used for measurement during testing are in “inches of water.” 1 in H2O is equal to 0.036psi, and 27.67 in H2O is equal to 1psi. This is a relatively small unit of measure, so it is useful for showing small differences in pressure.

The manometer has two inputs for pressure, and will display the difference between the two pressure inputs. With how we hooked up the pressure signals a positive number indicates a pressure drop, and a negative number indicates a pressure gain. The manometer was to be placed in the cabin, so substantial lengths of hose were needed. Since we’re testing for pressure, the length of the hose was negligible. However, two 10ft lengths of .125in norprene hose were used, and were rated not to collapse under the expected vacuum.

Grimmspeed ft86 intake testing stock airbox
Location of each fitting added to the stock intake system for pressure sensing.

The stock intake was tapped in various places, and fittings were added that would connect to the hose for the manometer. A fitting was placed at the inlet of the snorkel, at the top of the front of the airbox before the filter, at the top of the airbox after the filter and before the MAF housing, at the inlet of the rubber elbow after the MAF housing and before the bend, and finally right before the throttle body. Each fitting was sealed to prevent leaks, and caps were added to all fittings.

See Part 3: Testing Procedure to Continue

Restriction in the Stock BRZ/FR-S Intake – Introduction

grimmspeed intake testing brz frs
This photo shows placement of one of the fittings for pressure sensing.

When we began thinking about designing an intake for the twins, we first wanted to evaluate the claim that “the stock intake is good enough.” Its general knowledge that in the last ten years or so, that factory OEM intakes have become very good in design, and are often difficult to improve upon. There are several ways to evaluate this claim, and we wanted to start out with looking at the design of the entire intake as both an overall system, as well as the sum of all of it’s parts.


A visual inspection doesn’t tell an absolute truth about the system, but it does give you a place to start evaluating. The first source of restriction you’d look for is sharp or abrupt entry points. Air entering a pipe without a flared entry (think velocity stack, or a funnel shape) produces a restriction, compared to one that does have a flared entrance or transition. Just the same, when air has to traverse a larger and larger angle bend, there is an increase in restriction. The same can be said for when air has to pass over surfaces that are not smooth, etc. All of these situations add restriction, which can be measured as a drop in pressure. The ideal case to move air from point A to point B would be a perfectly smooth, straight length of pipe, and even that will have a pressure drop as the length of the pipe increases.

So from a visual standpoint, lets break apart the sections of the intake: There is a snorkel, front of airbox, air filter, rear of airbox, MAF housing, intake elbow, and throttle body. The entire system can be looked at as being the area before the snorkel (behind the bumper cover) to just passed the intake elbow (right at the throttle body). Measuring the difference in pressure between these two points will give you the overall restriction of the system. But in order to identify where the weaknesses in the system are, one would be more interested to measure the difference in pressure between components in the system. For example, to measure the restriction the air filter has on the system, you would measure the pressure before and after the filter. And if you add up the pressure differences between all parts of the system, it should equal the overall restriction.

Back to the visual inspection of the system, what do we see as a potential problem area, and why do we want to choose these locations to test? The first part of the system that air sees as it enters is the snorkel. The inlet of the snorkel looks good; there is a well formed velocity stack that has minimal extra material from being molded. It’s a slight oval shape, roughly 2.25in x2.5in. About 10 inches down the air’s path, the snorkel starts to make an approximate 90 degree bend to it’s exit. The bend is very smooth, and all the while the shape is transitioning to a flatter oval, while at the same time increasing in overall cross sectional area. At the point where the snorkel transitions into the air box, it is roughly 2in x 5.7in. The snorkel contains two resonators along the first section, in two different sizes, each containing a small drain hole at their lowest point. The snorkel is sealed to the air box with a strip of foam that interfaces the outlet of the snorkel to the inlet of the front airbox.

The front face of the airbox is angled at the bottom, and contains a circular emboss. Both features are in place to maximize area before the filter, while still clearing the radiator and fan. There is also a large resonator to the left of the entrance. The front airbox has a hole at it’s lowest point just right of the entrance, as does the large resonator, both for drainage purposes. The inside of the front of the airbox is very smooth across all surfaces. The only noteworthy point from a flow standpoint is at the entrance. The half of the entrance below the snorkel has a smooth radius flowing towards the filter. However, the half above the entrance is abrupt, and looks different than you would expect from viewing it from outside the box. Outside the box, just above the exit of the snorkel there is a hump which looks to exist as an area to smooth airflow going towards the filter, but just the opposite appears to be true as there is a void here. One can only assume this is for strength, or some phenomenon that is hard to explain.

The air then flows through the filter, which is not your typical paper filter, and has only 14 large ribs. I am unsure of the media of the filter, but it is similar to a fabric like cotton. The ribs on the front side are longer than those on the back to increase filter surface area.

After the filter is the rear of the airbox, which contains mostly smooth transitions, with a taper at the opposite side to the exit that should promote flow towards the MAF housing. The only noticeable source of restriction in this piece are several protruding ribs that run lengthwise in the rear of the airbox, however small. The exit of the airbox is technically the mass air flow, or MAF, housing. The entrance to the MAF housing appears to have been optimized, as it is one of the most important parts of the entire engine. The rear face of the airbox has a section “dug out” to smooth the transition into the MAF, and the opposite side of that feature has a molded plastic velocity stack. Immediately at the entrance is a plastic matrix that is commonly referred to as an “air straightener.” This is specifically put in place to help the MAF provide the most accurate reading as possible by modifying the flow of air before it. The thickness of the pieces of this matrix is 2mm, and the diameter of the entrance here is roughly 68.5mm. The entire MAF housing is only about 70mms long, and places the MAF sensor about 25mm, or about 1in after the air straightener. The inner diameter at the MAF sensor is 70mm, and the diameter at the outlet of the MAF housing is about 72mm. So there is a taper through the entire section, albeit minimal.

At the exit of the MAF housing is the entrance of the intake elbow. The entrance to the elbow is just under 3in in diameter, and has an immediate 90 degree bend. This bend is very tight, and has a centerline radius significantly under 3in. This most likely means that the diameter of the cross section does not stay a constant 3in as the bend progresses. There are ribs on the outside of the part for strength, but they do not exist on the internal surface of the elbow. There is a tube exiting the elbow for the sound tube, just opposite of the intake elbow’s entrance, and a resonator toward the bottom of the engine bay, both located directly on the bend. Immediately after the bend is a roughly 2.25in long flex section. This section contains 5 smooth ridges that exist on the inside of the tube, and extend outwards of the tube less than .125in. After this flex section is a 5in long straight section, smooth on the inside, with ridges on the outside. This terminates at the entrance of the throttle body.

Based on this visual assessment there isn’t much to expect in the way of restriction. From the entrance of the system to the exit, we expect to see a restriction from: 90 degree bend of the snorkel, air filter, decreased size (in comparison to the air box volume) of the MAF housing, the tight 90 degree bend on the entrance of the intake elbow, and the flex section located right after the previous bend.

See Part 2: The Equipment to Continue