nitrodann

May I please have a more detailed answer Vlad.

Savington

iTrader: (31)

I don't have easy access to them, or else I would. I've changed hard drives and computers a few times since then. I have posted the raw data copied straight from the logs before on the forum, though (here) so hopefully that satisfies your curiousity.

Savington

iTrader: (31)

This is engineering myopia at its finest, dude. Why bother measuring the IATs in the plenum when they have to travel down the runner? Why bother measuring in the runner when the air has to pass through the port and past the smoking hot valve? Where do you draw the line?

The coldside IC endtank is easy to install, on the correct side of the heat exchanger, and highly resistant to heatsoaking. It's fine, certainly accurate enough for 99% of applications.

I'd argue it more, but I have to go install an IAT sensor in a prototype intercooler.

Girz0r

iTrader: (1)

patsmx5

iTrader: (16)

Uh....

The further the sensor is from the valve, the more error there will be. Putting the sensor as close as possible will reduce the error the most. Common sense. You measure EGTs close to the exhaust valves, coolant temps close to the water outlet on the engine, oil temps close to the sump, all for the same reasons. To get an accurate measurement.

Your "argument" is that putting the sensor by the intercooler is accurate enough, same BS you accuse me of for "where to draw the line" regarding accuracy.

The whole point of having an AIT sensor is to do fuel and timing corrections based on the temperature of the air going into the engine, not the air exiting the intercooler. Two different things, two different temps, and your entire argument is to ignore it and call it accurate enough.

For anyone that's ever had a lean or rich hot restart after a brief shutdown, you've experienced what a change in air temps do to fueling when the temperature correction is not accurate.

aidandj

iTrader: (5)

Pat, your heat soak test only determines if the heatsoaking is occuring because of the metal that the sensor is screwed into is heating up.

What if the sensor element is heatsoaking, it may be the same temp as the air around it, but what if it isn't responding fast enough when a bunch of cold air is sucked in when restarting.

nitrodann

Id say, then its a crap sensor.

Savington

iTrader: (31)

1.8L engine = 1.8L of air ingested per engine cycle (RPM/2) @100%VE

850rpm idle = 425 engine cycles per minute, or 7 cycles per second.

Assume 40% VE at idle, the engine ingests 5 liters of air per second at idle.

Let's guesstimate 2L total volume for the intake manifold. 2.5" OD pipe with a .080" wall thickness (standard IC pipe IME) has a volume of ~71cc per inch, with about 2ft of piping between the intercooler and the throttle body, so let's call that another 1.7L. IOW, the entire volume of air between my IAT location and yours is ~3.7L of air, or approximately the volume of air ingested in 3/4 of a second.

So yes, my argument is that it doesn't matter.

Joe Perez

iTrader: (8)

You're looking at this wrong.

Consider the following two errors:

1: The change in temperature of the air (in a running engine) from the time it exits the intercooler to the time it enters the intake plenum having traveled about two feet at high velocity and encountered relatively little surface error, vs.

2: The error induced into a temperature sensor by being screwed into a hot piece of metal located behind the radiator vs. being screwed into a roughly ambient piece of metal located in front of the radiator.


Sadly, an AIT sensor does not measure the temperature of the air it's exposed to. It adds in a bias based on the temperature of the case of the sensor, and that error is far greater under most conditions than the change in the temperature of the air as it travels two feet at 50+ feet per second.

nitrodann

In the intercooler it takes a LOT less time than that to change the temp by 100f.

Who knows how much the valves and inlet tract can change the temp, especially considering the much greater Delta T.

nitrodann

Great, so you can definitely confirm that heat soak in the sensor body affects output of the sensor significantly?

Dann

Joe Perez

iTrader: (8)

Yes, I can confirm this definitely, for any commonly used OEM-type sensor. (eg: your J-type thermocouple mounted into a hollowed-out ceramic plug with aerogel used as the sealing agent doesn't count.)

Joe Perez

iTrader: (8)

So, the tubing you use to connect your intercooler to your intake manifold is filled with tightly-packed fins and has an equally large, finned external surface area?

No offense, but that's a pretty stupid design. You'll have much better results using smooth-bore tubing, and better yet if the tubing is plastic or silicone, which conducts heat less efficiently than aluminum.

nitrodann

Come on man, argue my point, not some absurdity. The valves are hundreds and hundreds of degrees and literally no one here has data on the heating effect they have on the inlet air temperature.

Thankyou for your info re: GM sensors.

Dann

Savington

iTrader: (31)

Sure, but the intercooler is literally designed to transfer as much heat as possible. A round aluminum tube isn't. The difference in actual air-metal contact between the two is probably an order of magnitude.

nitrodann

And the delta T in an intercooler is <200* not >500*. Additionally the inlet air sits swirling around behind the hot valve for 3/4 of the time the engine is running. Im just saying that we dont know, and your maths example doesnt disprove this.

18psi

iTrader: (76)

I can post up logs of the gm iat screwed into the cast iron manifold of a FFS supercharger pretty much sitting at 200f no matter what kinda driving the car did.

Ironically this same FFS I sold to Pat

But who cares about any of my experiences. Let's talk about more theory

Savington

iTrader: (31)

What's your point? I don't even know what your argument is.

patsmx5

iTrader: (16)

This. No air temp sensor can measure air temperature accurately if it is heat soaked by something other than the air it's supposed to be measuring.


At wide open throttle, like you say the high velocity/high mass flow will result in the temps being practically the same.

Regarding the error you say the AIT sensor base temp affects the measurement, my testing showed otherwise on my setup.

At cruise/idle/not moving/etc, they will be different as the volumetric flow through the pipe is much lower.

I've measured this before, when the car had a SC on it. Intercooler stayed about 5-10*F over ambient at idle/cruise/highway cruise. Air temps near the throttle body changed depending on vehicle speed. At idle the engine bay heat soaked and AITs entering the engine were higher than what exited the intercooler due to heat from the engine bay entering the charge pipe.

A very good example of this is when I warm up the car, shutdown the car, let it sit for 15 minutes, and then restart the car. The engine bay heat soaks and the air temp entering the engine is typically 50*F+ over ambient easy. Why? Because EVERYTHING heat soaked when I shut it down. That is, everything, including the air in the pipe. It doesn't magically drop to ambient in 1 second as suggested, this thing called heat transfer happens.

AIT mounted near the intercooler reported a much lower temperature than what was actually going into the engine, resulting in a rich condition at startup. AIT near the throttle reported a higher value, and AFRs much closer to target.

Not hard to test if anyone is curious, it's just installing a second AIT sensor.

18psi

iTrader: (76)

umm, I'm almost afraid to ask what your air density comps look like if the car is rich on hot restart with a heatsoaked sensor.

cause like 95% of turbo miata's on this site have to battle hot restart lean-ness.

also how would you mount the ait near the tb without heatsoaking it? we've had a whole thread about this before