IAT heat soak

[samh_08]

Yes boys, I searched . I am having a little bit of an issue keeping my BENs consistent with my IAT sensor getting heat soaked. I know I will always have some heat soak, but Im talking 120*F when ambient is ~70. This happens when I shut it down and let it sit.. Eventutally the IAT comes down to where it should be (also giving me proper BENs) but it takes 5-10 mins of driving.

Now my question is, it there any effective way of tuning this without the use of COS5 table A0014 or reloctating the sensor (I dont believe it would be an accurate example of the air actually entering the throttle body)

Let me know what you guys think,
Thanks!

[SSpdDmon]

Yes boys, I searched . I am having a little bit of an issue keeping my BENs consistent with my IAT sensor getting heat soaked. I know I will always have some heat soak, but Im talking 120*F when ambient is ~70. This happens when I shut it down and let it sit.. Eventutally the IAT comes down to where it should be (also giving me proper BENs) but it takes 5-10 mins of driving.

Now my question is, it there any effective way of tuning this without the use of COS5 table A0014 or reloctating the sensor (I dont believe it would be an accurate example of the air actually entering the throttle body)

Let me know what you guys think,
Thanks!

Re-locate the sensor. `

[GAMEOVER]

Re-locate the sensor. `

I agree...

[ringram]

How do you know thats not just heatsoak to the intake air at low airflow?
Otherwise why not just mount it right out of the airflow behind the bumper etc.
What you want is for it to accurately track intake air temp even if its heatsoaked.

Look at the charge blend factors and tables. In theory thats where it should be modelled. I think the A0014 and similar tables are just an easier way to fudge it.

Hell on the E38 I used Injector flow vs IAT to hack it into shape. But Id much rather figure out the airflow vs charge blend factoring, but it seems there is a table or two missing. Therefore my airmass calculations have gone out the window because Ive hacked injector flow :(

All very sad.

[SSpdDmon]

How do you know thats not just heatsoak to the intake air at low airflow?
Otherwise why not just mount it right out of the airflow behind the bumper etc.
What you want is for it to accurately track intake air temp even if its heatsoaked.

Look at the charge blend factors and tables. In theory thats where it should be modelled. I think the A0014 and similar tables are just an easier way to fudge it.

Hell on the E38 I used Injector flow vs IAT to hack it into shape. But Id much rather figure out the airflow vs charge blend factoring, but it seems there is a table or two missing. Therefore my airmass calculations have gone out the window because Ive hacked injector flow :(

All very sad.

I think the effect is very minor. Air...or should I say the gases that make up our atmosphere...are very poor conductors of heat.

Heat conductance occurs when two atoms or molecules interact with one another. This means that they must be reasonably physically close to one another. Because the density of a gas is so much lower than that of a solid or liquid, it will not conduct heat well.

I don't think that the air coming in the intake is around long enough to be altered much. Stagnant under-hood airflow (like when a car sits at idle) will start to warm up similar to the way your oven operates in your kitchen. Therefore, it's natural to see a slight-moderate increase in intake temps over ambient temps when a car isn't moving. However, other parts get hot too. Granted some are plastic and some are metal…they are still much better conductors of heat. Since these parts typically house the IAT sensor, the sensor is inevitably going to confuse the heat radiating off of them as the incoming air temps getting much warmer.

That being said, it’s also been my experience that the LSx motors respond on the lean side to higher temps and on the rich side to cooler temps. So, if you’re still reading and are waiting for me to get to my point…here it is. If the PCM thinks the air entering the engine is denser than it actually is, that’s fine with me. I’d rather err on the rich side than the lean to avoid detonation, to avoid the PCM from pulling timing, etc. That’s why I suggest (on N/A cars) relocating the sensor to just in front of the filter, suspended by fishing string, and a good 3” away from anything solid.

[SSpdDmon]

The air itself won't be affected much but the sensor will be.
The sensor is sitting in the intake tract and depending on
where it is placed it can constantly see higher temps than ambient.

Example: AN LS1 Fbody IAT is usually about 10° above ambient on a "warm" day
this can sky rocket while idling or after a restart even if ECT has dropped.

My point exactly...the sensor "thinks" the air is warmer when in actuality, it probably isn't much warmer at all. I'd be willing to bet the temp of the air under the hood at idle plus a few degrees is truly what the PCM should be reading. So like I said above, I'd rather err on the safe side, say screw those few degrees, and just go with under hood temps.

[swingtan]

I've done a fair amount of logging with different intakes and have seen some interesting things occurring. All cars are Holdens, but the intakes differ, here's just a few examples.

• OTR with IAT set toward the front of the unit: In this car I have seen the IAT rise when the car is stationary and idling. the hot air rising from the radiator was being picked up by the OTR and showing on the IAT. As soon as the fans turned on, the IAT dropped as the warm air was pulled through the radiator and away from the intake.
• VZ with MAF and stock air box: This is my car and has lots and lots of data logged on it. I also ran a thermometer inside the air box to see the difference between temps. Because of the position of the stock intake, it would easily pick up hot air from in front of the radiator, especially if the air con was on. During summer it was not unusual to see 70'C+ inside the air box. the MAF would always read 5' to 10' warmer while cruising and much more if stuck in traffic. It took a long time to cool everything down once it was hot.
• VZ with modded MAF and stock air box: With some minor mods to the MAF, I found that cool down times reduced and the temp difference between the air box and the MAF also came down. More changes to the MAF saw the MAF reading within 5' to 7' of ambient on a cruise, but it still suffered heat soak when in traffic. It was not too bad though and I'd have been happy to keep it as it was. At times the MAF was reading below the temp in the air box as the thermometer in the air box would get heat soak.
• VZ with OTR and MAF: Current setup, and obviously the best so far. Yes, it still suffers heat soak but the temps drop rapidly once moving. At the moment it's easy to have IAT's within 2' or 3' of ambient on a cruise.

I know many people just hang the IAT out in the air somewhere near the intake, but I prefer to run the IAT in the intake air flow. It the intake is heat shielded, then soak is not as much of an issue and correct fueling should occur. at low speeds, where airflow is minimal, there is much more of a chance of the air warming in the intake pipes. I'd like to know the ECM is measuring this. If you just care about WOT, then of course there is little time for the intake air to heat up.

Finally, if you get your IAT corrections set right and log over a decent time frame, then any heat soak issues should be averaged out. I've found that a single log of 1 or 2 hours driving works much better than 6 to 10 separate 10 min logs.

Simon.

[eficalibrator]

Re-locate the sensor. `

I disagree, see below...

How do you know thats not just heatsoak to the intake air at low airflow?
Otherwise why not just mount it right out of the airflow behind the bumper etc. What you want is for it to accurately track intake air temp even if its heatsoaked.

Look at the charge blend factors and tables. In theory thats where it should be modelled.

...and it is modeled that way from the factory. There is some very real change in the actual aircharge temperature due to warming from underhood components. The factory cal goes a step further and continues to model the heating of the air from the intake port (which is at least at coolant temp) as well. The whole point is to accurately model what's really going into the cylinders, not what's happening outside.

I think the effect is very minor. Air...or should I say the gases that make up our atmosphere...are very poor conductors of heat.

Poor conductor of heat, yes; but they also have very low mass (compared to anything solid around them) which makes them far easier to change the delta-T in short order.

I don't think that the air coming in the intake is around long enough to be altered much. Stagnant under-hood airflow (like when a car sits at idle) will start to warm up similar to the way your oven operates in your kitchen. Therefore, it's natural to see a slight-moderate increase in intake temps over ambient temps when a car isn't moving.

This is precisely what you should see modeled in the factory biasing tables. The effect is smaller at higher flow rates for a reason.

I know many people just hang the IAT out in the air somewhere near the intake, but I prefer to run the IAT in the intake air flow. It the intake is heat shielded, then soak is not as much of an issue and correct fueling should occur. at low speeds, where airflow is minimal, there is much more of a chance of the air warming in the intake pipes. I'd like to know the ECM is measuring this. If you just care about WOT, then of course there is little time for the intake air to heat up.

Agreed. If you're trying to do is keep it from pulling timing because you think you need more, there are other tables to use to increase ignition advance. Although, if the air really is warmer entering the cylinders, you'll find that it won't need as much timing to hit either MBT or the knock limit.

Finally, if you get your IAT corrections set right and log over a decent time frame, then any heat soak issues should be averaged out. I've found that a single log of 1 or 2 hours driving works much better than 6 to 10 separate 10 min logs.

Even better, perform steady state testing at a single temperature and see the effects once things stabilize.

[redhardsupra]

Even better, perform steady state testing at a single temperature and see the effects once things stabilize.

Greg, I agree with everything, except for this last thing--the real world has more than one air temperature, so wouldnt calibrating at that one stable condition be correct only at that one condition, and be wrong at all others?

[eficalibrator]

Greg, I agree with everything, except for this last thing--the real world has more than one air temperature, so wouldn't calibrating at that one stable condition be correct only at that one condition, and be wrong at all others?

Not if you perform most of your testing at the nominal temp that all the VE tables are corrected to. Even better, follow that up with a series of measurements at other temps (which is admittedly not doable in most performance shops on a single day) so that you can check the temp correlation.