Intake Comparison - 3.5" Injen versus 3" CTS Turbo

B8 S4
#1

PURPOSE
The purpose of this evaluation is to compare the airflow, boost pressures, and intake air temperatures between a popular design of a 3” intake and a popular design of a 3.5” intake for the B8.5 Audi S4. The intakes tested were a modified Injen intake and the CTS Turbo intake which is identical in design to the Roc Euro intake (identical with respect to pipe diameter, filter diameter, heat shield area).

THE INTAKES

CTS TURBO:

http://img.photobucket.com/albums/v714/bhvrdr/IMG_20151012_123107-s_zps8qofl7be.jpg

MODIFIED INJEN

http://img.photobucket.com/albums/v714/bhvrdr/IMG_20151012_120809-s_zps5tft55kp.jpg

PROCEDURE
All efforts were made in order to eliminate confounding variables. As such the following procedures were performed.

  1. All testing was performed on the same open stretch of closed course pavement the same direction of travel during the same ambient conditions (82 degrees F, 54% humidity, 29.95in pressure). The entirety of the testing was completed on the same day within 30 minutes from start of testing until end of testing.
  2. Prior to the beginning of testing the vehicle was driven for 3 baseline runs and then parked for 5 minutes in order to allow heat soaking. This was in order to make all beginning runs of both intake groups the same beginning conditions (each run group initiated after 3 successive runs).
  3. The first intake tested was the 3.5” Injen intake that was run with no heat shield and no portion of the OEM airbox in place.
  4. Each intake was run for three successive third gear runs from approximately 2000rpm to 6200rpm. Each run group was completed after the vehicle idled for approximately 5 minutes. Runs 2 and 3 were performed immediately after Run 1 as the vehicle came down to a 2000rpm speed in third gear from the prior runs.
  5. The intake was swapped out from the Injen intake to the CTS intake in the same 5 minute idle time that was given prior to testing the Injen intake in the beginning of the test.

RESULTS

http://img.photobucket.com/albums/v714/bhvrdr/Air%20mass%201s_zps1wjbjspf.jpg

http://img.photobucket.com/albums/v714/bhvrdr/Air%20mass%203s_zpsl0h4cawi.jpg

http://img.photobucket.com/albums/v714/bhvrdr/boost%201_zpsbkojgyre.jpg

http://img.photobucket.com/albums/v714/bhvrdr/boost%203_zpszz6gebpg.jpg

http://img.photobucket.com/albums/v714/bhvrdr/IAT%201_zpscmwavxoh.jpg

http://img.photobucket.com/albums/v714/bhvrdr/IAT%203_zpsf4yhppws.jpg

SUMMARY

NOTE: Both intake air mass and boost settings will appear universally low for both intakes. This is due to the use of a piggy back that attenuates the ECU signal to see lower boost and consequently air mass (which is calculated form boost). This should not impact the test results in any manner as the attenuation is equal for both test sessions and both intakes.

Interestingly the 3.5” non heat shielded Injen intake appeared to slightly surpass the performance of the 3” heat shielded intake in the categories of air mass and boost, although the difference was negligible. Intake air temperature results indicated a more significant difference between the two.

One may have expected that the 3.5” Injen intake would have been at a disadvantage due to the total lack of heat shielding. At the very least during the first run after the car had heat soaked from idling one may have surmised the non heat shielded intake would have produced higher IATs and less air mass. This is not what the data reflected.

The larger diameter of the 3.5” Injen versus the 3” CTS did appear to offer advantages. This was reported by Injen designers/engineers who originally developed a 3” design and later terminated that project in favor of a 3.5” design after they reported seeing an 8hp increasing using the larger diameter intake.

Perhaps what was also helpful with respect to the Injen was the design of the tube that places the intake filter directly in the path of the incoming air. Refer to the above picture to note the intake air path in reference to each, the Injen and CTS filter.

LIMITATIONS:

The limitations of this evaluation was the omission of testing the Injen intake in its original configuration using the bottom of the stock airbox with their own airbox lid. In retrospect, it also would have been helpful to log actual timing for each intake. It had not been expected that there would be a significant difference in intake temps, and particularly the Injen being the intake with the lower intake temps. Had these intake temp differences been better anticipated, timing could have been logged to see if the reduction in intake temps produced increased timing utilized.

#2

makes sense…less restrictive, more open intake has been the formula from day one on this platform. Nice charts.

#3

mind blown on the IAT’s.
No heat shield is lower IAT than with heat shield???

Wow. I wonder what is going on there.

#4

I was surprised about the IATs as well.

A couple thoughts…

First, i’ll run the test again in reverse order just to make absolutely sure the order of testing did not impact results but I am very doubtful it did. I ran the car hard and idled it prior to putting on the first intake to try and control for that.

The other thing that another forum member from “the other site” lol, brought up was that the heat shield may actually be preventing the best airflow. Lets discuss that…

Some of you remember Audis that had fender mounted intercoolers (b5 s4, b5, b6 a4). Audi placed vents in the fender liners in order for the air to be able to flow from the foglight vent, through the core, and out the fender liner vent.

Putting a heat shield where it is in the engine compartment MAY be inhibiting straight through airflow out past the filter. It could possibly be creating a bit of a stagnant hot air zone.

Of course another factor is the larger filter area and the larger pipe area (larger area to disperse heat).

I can test the heat shield theory by running the CTS with and without the heat shield. It would be pretty funny if removing the heat shield actually helps our IATs in all situations but idling. I’m not convinced but I will at least test it.

Mike

#5

Yeah worth testing for sure.
I wish I had a vagcom to do logs of this sort.
I actually ghetto fabbed an intake with a heat shield and half the stock airbox. I noticed the airbox was right next to one of the headers, so in my head, there is no way you wouldn’t want to run a heat shield there. I mean, look at the opposite side where the coolant overflow tank is - audi put a heat shield there too! Or at least, that was my train of thought.

Would be interesting to see what your test results are when you put the lower half of the box back in.

#6

In a past life, when testing (PC) cooling equipment, we always used the Rise above Ambient (measured temperature minus ambient temperature). The thought being that since ambient air temperatures can fluctuate (short of being a lab/data center environment), but the rise above ambient would always be constant.

I remember doing a test with my APR stage II+ beta file that the Delta T would increase from 10C to over 30C after 3 pulls on a chilly fall night (though I used the data points from VAGCOM). (I had the APR Carbonio intake)

#7

want to buy one? I have an extra. PM me if so.

#8

Right on. Both intakes were measured over three consecutive runs and ambients were the same (injen runs and cts runs were done within the same 30 minute time span). I also monitored coolant temps to make sure the were identical for both sets of runs. I plan to run the study backwards starting with cts and ending with injen to make absolutely sure heat soaking was not a variable but this doesn’t appear to be the case thus far as both sets of runs should have been identically heat soaked. With another look though, thanks.
Mike

#9

The one without the heat shield has the stock “ram air” inlet pointed directly at the filter. The one with the heat shield does not.

That’s the only explanation I can think of. Not sure how big of a difference that makes.

#10

I’m going to say that testing boost pressure and calculated MAF is not really that useful with a spoof device that manipulates the measurement signal in question according to some kind of look up table. While the MAP derived “boost actual” is likely manipulated in a 1-1 fashion, it can’t be certain since the ECM could be conducting corrections and/or trims to correct out what it perceives to be sensor bias error. I’m not saying this is the case, but I wouldn’t try to conduct controlled tests with a measurement device that has intrinsic signal manipulation upstream. Just remove the sensor spoof.

Not to mention, you aren’t logging the bypass valve or cam phase angle, so who knows if the difference in boost is truly due to the difference in intakes, or possibly a result of the ECM manipulating boost by bypass and/or changing the cams which should in turn change the charge pressure in each cylinder. You also have higher IAT’s on the second run which could be implicitly changing the ECM strategy. You probably have these parameters logged? Would be useful to plot them along side.

As for the IAT comparison, I’m not sure if this test is controlled either. You tested the injen first, and then went back to test the second one. You said you did this quickly in order to ensure no change in atmospheric conditions. But what you probably did was inject a bunch of heat into your intercooler coolant loop and then ran the second intake tests with IAT’s starting higher.

There is no reason one intake should have a 10 deg C delta in IAT’s, especially when you first start the test at 2k rpms. I am pretty sure that the old “intake comparison tests” always looked at signals derived from the pre-blower MAP sensor. There is some other metric there. And your CW sensor spoof doesn’t corrupt those results :wink:

#11

Thanks for the feedback.

Based on the IATs in all other logs and comparisons I can locate, it appears we are all logging the G430 sensor. I’m actually not sure if vag-com has the G42 sensor available (not all sensors are mapped out by vag-com) but I can look again and try some out and run logs on both sensors for the next test. Obviously everyone who has logged up until this point has been logging the G430 or else we’d be seeing much different IATs in other peoples logs.

With respect to order of tests, I would agree this possibly could have been a factor (although no other rises in intake temps were noted during subsequent third gear runs) and just to make sure to rule it out i’ll do it in reverse order. I’m also going to run it first with the CTS with the heat shield in place and then second with no heat shield in place on the CTS, third with no heat shield in place and the ram air scoop in place (if it will fit), and then 4th the injen. If heat soak is an issue the injen will be heat soaked to all hell so we can see if there is reliability in the results.

With respect to the piggy, it is the same attenuation map used for all runs so any differences in ECU maps would be the same regardless of the attenuated signal or the non attenuated signal (ie boost and timing changes related to IATs and ECTs). I get just as reliable (read: repeatable) results with the piggy on versus without. I could always run it stock though to satisfy that skepticism. I did monitor ECTs as that is usually a confounding variable that will cause map switching. The bypass valve stayed 100% closed. Ive actually never seen my bypass open ever since the setting I use produces zero timing correction during initial flogging and after repeated heavy flogging and heat soak conditions I have seen a maximum of 3degrees knock correction. I obviously have pretty high quality 93 octane here though.

Good points.

Mike

#12

One thing that also puzzles me is why boost changed so much between runs 1 and 3 (though it was consistent with both intakes). Looks like as much as +50hPA or 0.7 psi difference. And that is the piggyback scaled number which unscaled could be an even bigger difference in actual boost. Seems a little fishy for boost to vary that much with bypass closed and all other things being equal? I wonder if the piggyback scaling really is as consistent as we think?