E30M3

Pinky and I decided to get together and informally test his CAI. It's the "Mingster" version which appears to be very Spoon-like. Pinky has been working hard to deal with underhood engine temp issues in various ways. This is a great idea cuz high temps cut power, especially on high specific output engines. Besides the CAI he also has done some air intake insulating. We used an infrared non contact thermometer to confirm and get a sense of the intake temp "problem". We also used a Race Technology AP-22 dual axis accelerometer to record the car's performance on a real road moving through the beautiful early Fall air. Then the data was uploaded to a PC for analysis. Some quickie summary data is noted below, tons of raw data is available and has not yet been fully examined. As passenger and button pusher I can't say that my butt noticed any difference at all. And these are WOT tests which ignore any cruise and engine response changes which are most noticeable to the driver.
------------------------------------------------------------------------------------------------
Hey ***, I enjoyed meeting you and playing around with the CAI.

Attached please find the raw datalog of the tests we performed. The CAI tests we did first enjoyed air temps in the low 70s F while the near stock test sucked in high 60s air. This is because we started at around 3:30 or so as temps were dropping. We did a set of runs with the CAI and a set without it. We didn't have a chance to switch back and forth as much as we would have liked since it takes some time to sap parts. That is okay though, since you generally need to do a few sessions to really get to the bottom of things. Usually you discover stuff that then requires you to devise special additional tests to investigate properly. It was a nice clear day, top down, two aboard. The test road was dead straight and nearly perfectly flat. We ran back and forth on it many times and used normal starts. Ideally, it would have been nice to do a totally bone stock versus max cool setup test on a hot day, among other things. Still, the results are what they are and are very real-world. The AP-22 was not calibrated to the car. But that is actually not a very big deal since the RELATIVE differences are what we are looking for here and not whether you put out 200 vs 201 HP, or 198 vs 199 HP to the ground. To calibrate we would have to do some high and low speed coast down tests which gives you mechanical rolling resistance and Cd times frontal area aero factors. Also, a chassis squat fudge factor and actual test weight can be determined. I have all of that data and more for my M3 of course. Another issue is the OBDII adaptation issue which can affect the results. The car's ECU adjusts the engine mapping according to parts swaps it notices and how the vehicle is driven over a number of sessions. Since the car runs the CAI all of the time, the ECU was optimized for that and Pinky's driving style.

I have summarized the data from the speed-to-speed tests and plan on posting this stuff to the board. If I need to revise anything I will do so. So far, it looks like the CAI was a bit better than near-stock from 30-60 MPH. Average 30-60 without "shakey" runs was 4.38 versus 4.45 and the necessary distance to achieve that speed change was shortened by 5 feet. In other words the CAI would give you a 1/3 car length advantage in that instance if our tests are credible. This covered rpms from just before vtec until just past peak HP. Runs repeated to 0.08 seconds and 5 feet for the CAI and 0.08 sec. and 6 feet for near stock.

For 30-70 it looks like the advantage reversed! Lessee..8.76 for CAI versus 8.49 near stock. Distance delta about 19 feet, about one car length. This covered the roughly 3000 to 7000 RPM range at WOT. Basically low to mid range. Note that at part throttle and during transitions (driveability related) the CAI may still be better. We can measure that characteristic but did not do so yesterday. The CAI repeated within 0.03 seconds / 4 feet which is a tiny percentage. The near stock had more variation, around 0.17 seconds and 12 feet. Additional runs could be done to examine this variation. Remember that this CAI changes not only the air temp factor but several other things as well.

I'm sure there is more useful info to be found in the continuous run data. But so far within the various limitations of our screwing around, it looks like the CAI WOT is better in vtec but a hair worse below vtec. It could also be better under slow traffic conditions. Since theory predicts that hotter air would lead to more rapid power reductions. If that heat rise can be reduced and slowed down then the driver will enjoy more of the max potential more of the time. Without needing a minute or two of high speed running to cut and stabilize temps. Were we to test things again there are a few things we discovered that it would be fun to try to understand a bit better. And I suspect that tests done on hot days might be very interesting too...

Gee it's tough to go over this stuff in a single post to a message board!!

Stan

Tedster

Stan,

very interesting, thanks for doing this. Please post any more data or analysis that you can...

Thanks,

Ted

Tanqueray

Stan,

While I am disappointed that the ID/Spoon CAI had mixed effects, I am not completely surprised; no one has found any other way of improving the use of the stock airbox.

Do you have any plans to do something similar for the Injen or AEM CAIs? Maybe replacing the airbox completely will make a difference.

Also, I would love to see this experiment repeated in warmer weather. It stands to reason that more power would be lost when the ambient temperature is 100 degrees than when it is 70 degrees, giving the CAI more room for improvement. Anyone in TX have the ID/Spoon style CAI, AP-22 unit and some free time?

Tanq

Giampiero

Stan,I will like to have tested our CAI in carbon also;any idea?

E30M3

>>very interesting, thanks for doing this. Please post any more data or analysis that you can<<

Okay will do. Although I can only do additional tests when I have access to an S2000 as I don't own one. If you are interested I can email you the raw data which I uploaded to my PC from the ap-22. You can pull it up on a spreadsheet or word processor.

Stan

E30M3

>>While I am disappointed that the ID/Spoon CAI had mixed effects, I am not completely surprised; no one has found any other way of improving the use of the stock airbox. <<

It appears that the results were mixed. But more testing and more data analysis wouldn't hurt. Whenever you test, you always think of a better way to do it next time. And there never seems to be enough time.

I can tell you that it's very worthwhile to measure things like CAIs on the move since the airflow changes speed and flow pattern, etc. with roadspeed. You can't duplicate this effect very well on a chassis dyno. I don't think it's a huge problem but it wouldn't hurt to flare the CAI entry point to reduce entry turbulence. Entry turbulence makes the effective crosssection go down just as if the entry hole was smaller.

>>Do you have any plans to do something similar for the Injen or AEM CAIs? Maybe replacing the airbox completely will make a difference. <<

One thing that may have affected matters a bit is the rubatex insulation that Pinky used. It's selfadhesive - so both the Stock air horn and the CAI enjoyed an insulated box. A JR filter and a debaffled airbox were used. I can redo the tests with any intake or whatever provided that I have an S2k to use.

>>Also, I would love to see this experiment repeated in warmer weather. It stands to reason that more power would be lost when the ambient temperature is 100 degrees than when it is 70 degrees, giving the CAI more room for improvement. Anyone in TX have the ID/Spoon style CAI, AP-22 unit and some free time? <<

BINGO! Our ambient temps were near room temperature. Part of the benefit to CAIs is not necessarily to gain power, but to avoid or reduce the loss of power that occurs when things get hot and heavy.

I think that in-car instrumented tests are super useful. They are very real world and you can run torque/HP curves, determine how many feet or car lengths a particular mod gains or loses. That is important on the street cuz it tells you where you'd be if you ran against yourself or a friend so to speak. If you do in-one-gear speed-to-speed tests there is no skill or luck invoved. On a flat road you can get things to repeat very closely.

Stan

E30M3

>>Stan,I will like to have tested our CAI in carbon also;any idea?<<

Sure I can do that, assuming Pinky is also interested, as we used his car. From my inspection of the underhood temps on the S2000, it looks like there are a number of ways to improve matters.

Stan

Giampiero

let me know.

Luis

I am installing the Mugen CAI within the next 2 weeks, but unfortunately my testing will be limited to timing in gear rpm gains. That I can do pretty reliably.

I wonder how much benefits you could get from ram-air effects.

These should be more noticeable past peak power, in the 8.3Krpm to 9Krpm range. To test this in 5th, you really need a long stretch of road and it's probably not doable in public roads (my country is an exception, so I may try it).

In 4th there is still enough speed to potentially make a difference. Can you try it next time?

E30M3

>>In 4th there is still enough speed to potentially make a difference. Can you try it next time?<<

I have been up to about 110 mph on my local test road.

Luis, ya gotta get yourself an AP-22. I guarantee you will hang a picture of me up on your wall after you try it out. Since you have an s2000 (I don't) you can get the gadget totally calibrated to the car - rolling resistance, tilt factor, weight, and so forth. I use mine to fine tune my M3 and can determine subtle differences between engine oils, air filters, launch techniques, all sort of bolt-ons, braking cornering you name it. It is so much better than a g-tech and so much cheaper than a vericom PC unit.

www.race-technology.com Check out the DCAT software on the link. Or you can just use hyperterminal in the windows accessory menu for uploads to your PC. Between that and spreadsheet work you can do almost anything if you are a careful, detail oriented person which you seem to be.

Stan