IAT Questions

So just a few observations on a car I recently tuned... The IAT sensor is fitted just before the throttle body into steel intercooler piping. Engine was a 2.0L with quite a small turbo and small the medium size FMIC (air to air)

On a 22'c ambient temp day the IAT once the engine was up to temp and idling on the dyno was around 30-35'c I did still notice that If I raised revs a little in netral the IAT did drop so there was some amount of heat soak going on... Is this unavoidable completely? and can just do the best that you can to insulate the cooler piping/sensor itself?

We are told not to put to sensor in the manifold if we have a choice, If it was in the manifold is heat soak ALOT worse? it would it be closer to 60-70'c at idle... or is it mainly only once the car has been switched off and left for 5-10mins that is a real issue?

So originally had the car running at 10psi and during the stabilizing time at the start of the run the temp was 32'c where it pretty much sat until about 3000rpm then it was a pretty much a diagonal line up and reached 46'c by the end of the run (6000rpm) I later raised to boost to 14.5psi and it was pretty much the same except the diagonal line was a little steeper and reached 54degree's

Just curious if this is about normal? Or should you only expect about 5-10 degree's above ambient?

The boost did drop off down to 12.5psi from about 4500rpm onwards so there is no doubt the turbo is too small. But just wondering how you know for sure... ie is the above normal and say if I raised to boost to 17psi and went past the efficientcy limit of the turbo would the IAT skyrocket almost vertically? or am I already well past to efficientcy limit and thats whats causing the temps to get that high? (if 54degree's is considered high in the first place??? I remember andre saying somewhere anything over 40degrees is when you start to loose power?)

And how do you know if its actually the intercooler causing the issues? I guess you would need to have IAT's before and after the turbo and then before and after intercooler?

Or is it also because its on a Dyno and you only have the dyno Fan blowing air, and on the street you would be doing 100km/h etc and IAT would be lower in the real world?

Ok well I think thats well more then enough questions already haha

You have a lot of great questions in there! Let me do my best to address them:

The most important part is to understand what we're trying to achieve - To measure the intake charge temperature. Once we understand this the rest will make more sense.

If you place the IAT in the manifold it will be heavily effected by heat soak as you've identified. For a conventional IAT correction table where we are only looking solely at IAT, or perhaps IAT and load, this will present problems as we can't adequately deal with the heat soak. Hence on an ECU that doesn't include a charge temp approximation table my preference would be to mount the sensor pre throttle body where it is less sensitive, albeit not immune, to heat soak.

With a charge temp approximation table we can have more control over the charge temp based on ECT, IAT, load and rpm. This lets us define a more realistic relationship between the parameters and charge temp. This raises a new problem though - How do you perfectly tune this table? Before you ask, right now I don't have an accurate answer. On my long list is some detailed investigation into this which will hopefully involve logging IAT very close to the intake valve as well as IAT, ECT, rpm and load. I'll post more when I actually have the opportunity to do some testing.

What you are seeing with the IAT during a run is pretty normal. The turbo heats the air as it compresses it. This leads to the rise in IAT. How much this rises through the course of a run will depend on the boost, efficiency of the turbo compressor, and the size/efficiency of your intercooler. When it comes to IAT less is obviously better and certainly 54 degrees is higher than I'd want to see. If you can get IAT within 10 degrees of ambient you're probably doing pretty well.

54 degrees probably indicates that either your intercooler isn't doing its job as well as it could or alternatively your working the turbo pretty hard. There are a few tells that let you know that you're running out of turbo:

1. Excessive IAT

2. Excessive turbine inlet pressure. This will normally be accompanied by the boost dropping off at high rpm.

3. No increase in power with additional boost - This might not be an easy situation to get too as you will be pushing the turbo pretty hard by this point. More often this will be signalled by the increase in power per psi boost dropping off. Lets say from 10-12 psi you pick up 20 kW atw but from 12-14 psi you pick up 12 kW. This is the first hint that you may moving beyond the efficiency of your turbo.

4. Compressor speed - Unlikely you will have the luxury of a turbo speed sensor but this will let you know where on the compressor map you're actually operating.

When your deciding on the performance of the turbo system you will most likely not have access to data like back pressure or turbo speed so IAT, boost curve and power increase are your key metrics for deciding how your turbo is performing.

You can check the effectiveness of your intercooler (and this will also give you another clue as to the efficiency of the turbo) by measuring the IAT pre and post intercooler. I've seen some terrible intercooler cores out of china that are essentially open tubes that do almost nothing for cooling. Swapping one of these for a decent quality core on one car resulted in IAT dropping by about 25 degrees and the car pickng up around 20 kW.

Lastly be aware that what you see on the dyno may not be representative of what you see on the road. It's very hard to replicate the effect of doing 100 + kmh with a dyno fan. I will usually spray water on the intercooler before a dyno run to try and maintain some real world consistency.

There you go. I hope that some of the above will help or add some clarity to your questions :)

I thought that I would throw this bit of logged data out of my car you to help a little with this discussion, The car is a GC8 STi with a EJ207. The turbo is a VF29 sucking through a 34mm restrictor (which is one of the reasons why the post turbo temps rise the way they do). This was done on a hot day in Melbourne, hence the 40 degree ambient temps, and is the third ramp run that had been done back to back to test heat soak. The data for the port turbo temps drops down to the default value as the temp sensor had run out of range and the ECU (M150) had faulted it out of use.

The sensors are fitted into the inlet tract in the following locations;

Ambient - Inlet pipe into the turbo, mounted in a bung in the silicon pipework.

Post Turbo - About 50cm downstream from the turbo outlet, mounted in a bung in the silicon pipework.

Intercooler - 30cm before the throttle body, mounted in a bung in the silicon pipework.

Inlet - In the centre of the standard aluminium Subaru manifold, this is the sensor used by the ECU for Inlet Temp calculations.

As can be seen, there is not a lot of difference in the pre and post throttle body temps off boost, but when the load starts to come up the temps post intercooler do start to rise after an initial dip when the airflow increases through the inlet tract, but no where need in relation to the rise in post turbo temps. The intercooler being used is one of APS's original front mount's for the WRX, with no water spray enabled. The turbo itself has fallen over it's efficiency point, with boost being controlled at this point not by the wastegate but flow restriction through the restrictor. This is seen with the boost level dropping even through the exhaust pressure is still coming up as the ECU is aiming for a higher boost level and not opening the wastegate to bleed off exhaust pressure.

The ECU has started to pull out timing to compensate for the rise in inlet temps, this comes into effect over 40 degrees in the map that I had loaded in the ECU at the time, there are also other compensations going on that I haven't included in the screen grab.

At this point, the torque being generated by the engine has started to drop rapidly, with a higher boost aim actually reducing torque compared to a lower boost aim.

Thanks Andre, So seems this certainly isnt running at its optimum then. It is a Tiny turbo and a Ebay intercooler.

With regards to: 2. Excessive turbine inlet pressure. This will normally be accompanied by the boost dropping off at high rpm.

If you were to fit a External wastegate would this help at all anyway? or just lower TIP but no real gain with anything else?

Hey Stephen, thanks for sharing. I think I suffer from a medical condition where no matter how much data I have, I always feel like I need more. It seems that perhaps you share this condition! Out of interest, what exactly does the 'Hack Jobs' worksheet detail? :)

I'm also interested to know what level of filtering you're applying to the exhaust pressure? I'm currently receiving this data via CAN from an AEM 4 channel UEGO and it's very noisy - Much noisier than I remember it being on my drag car.

Either way your data is a good way of adding some detail to the discussion. A restricter tends to create a fairly unique set of tuning circumstances but the data is a good way of demonstrating what happens to IAT based on turbo efficiency. It's common in my experience with restricted engines that you end up with more power by reducing the boost in the higher revs - You can drive the turbo harder but it's so far off it's efficiency map that it ends up being counter productive.

To answer your questions Clint, an external wastegate won't fix your TIP issue as the turbo requires a certain amount of exhaust 'energy' to reach a specific shaft speed and this in turn results in a certain boost level. Lowering the TIP can be achieved with a smaller wastegate simply by opening it and bypassing more exhaust flow around the turbine but obviously this reduces the energy available to drive the turbo and hence boost pressure.

The only benefit from an external wastegate is that you may be able to fit a heavier spring and hence have the ability to force the gate closed at higher levels of TIP. This can allow you to achieve more boost but this doesn't always translate to more power. It's similar to what Stephen mentioned above. If you are driving the turbo out of it's efficiency then you may actually see the power drop off with more boost pressure.

I don't see anything wrong with having over 1000 channels of engine data logging to go through...

The "Hack Jobs" worksheet is a worksheet that I have that doesn't have any specific layout, I just use it to throw up any channels that I need to look at for the job that I am doing (I think that you have worked out who I am and where I work by now).

With the Exhaust Pressure, that data has been smoothed over 30 samples logged at 50Hz, it is normally quite noisy but I cleaned it (and some of the other channels as well) up a bit for this topic. I have found that there are a number of causes to the amount of movement in the Exhaust Pressure, especially on boosted cars that may have a wastegate that is not optimised for the amount of boost pressure that is being run. The design of the whole of the engines breathing path from airbox to exhaust tip will also effect it, as restrictions and harmonics come into effect.

With the restrictor engines, the M1's ability to calculate the inlet mass flow is a good tuning tool for these engines, as you can tune the boost pressure to ensure that you sit at the restrictor's mass flow rate for as much of the engine speed range as possible, it generates an interesting boost curve when you do it though.