High intake temperatures - ideas.

[Kevin Doe]

During my drag racing this weekend I identifed an issue that I think was the culprit behind my less than expected times. Aparently when I sit and idle my car the intake air temperature sensor becomes heat soaked, and reads erroniously high (like 145 degree high). I have a table in my tune that pulls timing as a function of intake temp. With intake temps at 145 degres at the beginning of the run I was seeing up to 7 degrees of timing being pulled. This is an issues I've never noticed or seen before while tuning because usually my tuning is while driving, and the sensor doesn't get heat soaked when the car is moving and air is flowing.

I bought a different type of sensor to try out, and I may even try a different location for the sensor. The factory sensor is plastic bodied and mounts in a grommet in rubber. My sensor is metal bodied, and mounts in aluminum. I think this is the root cause of the problem. The aluminum is transfering heat to the sensor body, causing incorrect readings. I'm going to try out the factory sensor, and maybe even move it to a diff location.

Here is my intake setup. What is not pictured is that it is now header wrapped. I overlapped the wrap 3/4 width on each wrap, so its probably equivalent to double wrapping it. I used two 50' rolls on the intake.

It should have several layers of heat protection. Ceramic coating, aluminum pipe, ceramic coating, and double thick header wrap. This is the reason I belive the temp sensor reading to be erronious. I am considering not even putting the sensor in the intake, but putting it down in the cavity where the filters pick up air from. Thoughts? Currently the sensor is in the Y where the two sides come together, only a few inches from the throttle body.

http://i93.photobucket.com/albums/l55/Kevin_Doe/RX-7/910e3272.jpg

http://i93.photobucket.com/albums/l55/Kevin_Doe/RX-7/newwheels7.jpg

[SSpdDmon]

http://forum.efilive.com/showthread.php?t=5501&highlight=IAT

I had to relocate the IAT to the source on my last N/A car. In other words, I put the sensor just outside of the filter...where it was drawing air in from....which was right above the washer fluid tank on an f-body. No more heat soak.

On thing with your setup...even though it's wrapped, metal still retains heat. And, metal right behind the radiator is no exception (unless there's a fan shroud that I can't see redirecting air away from the intake tubes).

I'd see what you could do by relocating the sensor somewhere closer to the air source - not necessarily in the intake tract. Caspers sells IAT wire harness extensions and you can get an extra IAT sensor from Autozone for ~$10. It may cost you $45 or so and a few zip ties to test, but the results may be worth it. :)

[Kevin Doe]

IAT wire extensions? Why would this be necessary? The IAT wire in the harness is just regular ol' wire anyways.

[charcold-bowtie]

You can but the wires, an extension would just save the integrity of the stock harness perhaps. My sensor is in an aluminum intake, right before the TB and my IAT do pretty good. I do know a guy they drilled one into the old EGR spot on the intake, a truck intake, with good results last I heard ??

[Kevin Doe]

Update: I moved the sensor to the volume that the air filters pull from. I drove the car for about 45 minutes really hard, stop and go, to try and get everything really hot and heat soaked. The ambient temp was 65 degrees, the sensor read 65 degrees with the car cold and off. With the car fully warmed up, it read 70 degrees while cruising. With the car all good and heat soaked at idle for 20 minutes and the highest it read was 80 degrees.

All in all I consider this a big improvement. 15 degrees over ambient vs 65 degrees over ambient. The 15 degrees over ambient I think is actual temp of the air entering the filter. The IAT drops within 15 seconds of getting the car moving, so I think its real.

Success is mine! LOL!

[SSpdDmon]

Update: I moved the sensor to the volume that the air filters pull from. I drove the car for about 45 minutes really hard, stop and go, to try and get everything really hot and heat soaked. The ambient temp was 65 degrees, the sensor read 65 degrees with the car cold and off. With the car fully warmed up, it read 70 degrees while cruising. With the car all good and heat soaked at idle for 20 minutes and the highest it read was 80 degrees.

All in all I consider this a big improvement. 15 degrees over ambient vs 65 degrees over ambient. The 15 degrees over ambient I think is actual temp of the air entering the filter. The IAT drops within 15 seconds of getting the car moving, so I think its real.

Success is mine! LOL!
Nice... :)

So, how's the fueling consistency?

[Kevin Doe]

Nice... :)

So, how's the fueling consistency?

I have not checked as of yet. I"m getting ready to put on some new heads (stage 2.5 5.3l heads, 59cc chambers) so I'll probably hold off on any more tuning till that. This effort was really to make the IAT stop pulling timing so I could run better at the track and redeem myself in my local circle of friends. LOL.

But, I'll report back on consistancy findings when I start tuning with the new heads.

[ScarabEpic22]

Now you know how the TrailBlazer SS guys feel with the LS2 stuffed under the hood. Heatsoak is a HUGE issue with them, they see IATs like you were until they move the IAT out of the MAF and closer to the filter.

[redhardsupra]

using IAT's to measure ambient temps does little to proper estimation of air temperature INSIDE the intake...

[SSpdDmon]

using IAT's to measure ambient temps does little to proper estimation of air temperature INSIDE the intake...
That's what the charge temp blending table is for... ;)

[Kevin Doe]

using IAT's to measure ambient temps does little to proper estimation of air temperature INSIDE the intake...

I'm trying to get the IAT sensor to measure actual average temperature in the intake pipe. The sensor is on the edge of the pipe, where there is very little flow, especially at idle. The temperature profile of inside the pipe is very nonlinear. The temp at the wall (where the sensor was) is much higher than average temp of the air. I think the way I have the sensor mounted now will give a much better reading of true air temperature inside the intake pipe.

Plus as SSpdDmon said, once you get the air temp entering the manifold right, you blend that with the ECT, to get a better approximation of the actual air temp in the manifold.

[redhardsupra]

That's what the charge temp blending table is for... ;)
oh yea, and how are you gonna calibrate that? set it to 0 or 1 this time? or maybe just use AFR%error 'until it works'?

[Kevin Doe]

redhardsupra, are you incinuating that you should move the IAT sensor to the manifold?

[SSpdDmon]

oh yea, and how are you gonna calibrate that? set it to 0 or 1 this time? or maybe just use AFR%error 'until it works'?
Actually, I was going to say leave it stock. It's about heat transfer over time, right? Not sure if there is an easy way to cal that table.


**Please note** I cannot be held responsible for what I may or may not have said in posts older than 2 weeks. I have 'tried' many a things in the loose terms of R&D. The outcome of those events have long since evaded my recent memory. It's the process that I was following that I could tell you about if anything. :)

[redhardsupra]

oh no, quite the opposite, GM went away from having a temp sensor in the manifold for a reason--it wasnt precise enough, once you heat up the engine bay/intake, you ended up measuring the temperature of the intake itself, not the aircharge inside of it.

what i'm saying that moving the IAT around, you're changing the calibration of the aircharge temperature model, and until i figure out a way how to calibrate it properly, we're messing our airmass estimation process some more, instead of making it better. IAT relocation solves one problem, but introduces another, much more complex problem.

think of different approaches. on friend's LS4 (front wheel drive, engine is sideways, so TB is right over a header) we had heat soak issues, and we solved it with 15$ worth of thermal padding from Lowes. the AFR scatter due to imprecise temperature estimation was visibly improved.

[SSpdDmon]

Why wouldn't it be close enough to rely on the charge temp blending table? Like I said before, heat exchange should be a simple factor of flow over time....which is why the table is mapped that way (grams/second). You have a fixed distance the air will travel in the intake tract. The only thing that really changes once the engine is up to temp is how fast the air is drawn in. Flow vs. heat transfer has an inverse relationship. So naturally, as the airflow increases, the amount of temperature increase in the 'charge' should be less because it spends less time in the intake tract. This is what the GM engineers designed that table for...is it not???