Air temperature compensation for fuel injection

[Robert]

Hello Guy,

This is a general question concerning fuel injection. I want to know the table (or the law) of compensation of the fuel quantity function of the air temperature. I think for 20°C the compensation is 0% for 0°c is +xx%, for 40°C is -yy% etc... Thinking you have the answer.

Thank you and best regards

Robert

[Guy Croft]

You're referring to air density change?

Will this help any?

http://www.engineeringtoolbox.com/air-p ... d_156.html

There is also a calculator at the end of this good article:

http://wahiduddin.net/calc/density_altitude.htm

GC

[chrislandy]

On my ECU the air temp. compensations are as follows (forced induction):

Air temp. (deg.C)/ fuel % / ign %
-20 / 12 / 20
-10 / 9 / 15
00 / 6 / 10
05 / 4 / 5
10 / 3 / 1
15 / 2 / 0
20 / 0 / 0
25 / -2 / 0
30 / -3 / 0
35 / -3 / 0
40 / -6 / 0
50 / -9 / -4
60 / -12 / -10
70 / -15 / -21
80 / -18 / -21
90 / -21 / -21
100 / -24 / -21
110 / -27 / -21

These are the standard fuel % changes but ignition is experimental from my particular engine

Chris

[Robert]

Thank you Chris and Guy.

It was the data I was looking for. This confirms that the correction is function of the density (thus temperature). So an ECU must have the correction like Chris has.

Robert

[Robert]

Hello Guy,

It's me again. What's about the air humidity ? Indeed the air density is function of relative humidity (Hr), so the correction given by Chris are for dry air (~0% Hr). Is it important to use the density for 50% Hr (for example) implying :

Temp Correction
20 0
30 -4
40 -7
50 -11
60 -15
70 -19
80 -24
90 -30
100 -36

Regards

Robert

[Guy Croft]

Interesting point. Oddly enough a given volume of dry air is heavier than the same volume with water vapor in it so the wetter the air the less power it will give.

But you would probably have to put a preset compensation factor in for humidity, the only system (I know of) that can (in effect) cope with this are FI systems with a mass airflow sensor.

GC

[Robert]

If you don't have a mass airflow sensor but only an ECU providing the possibility to correct the A/F taking into account the air temperature, do you use a correction for dry or xx% wet air ?
The main difference is for the high temperatures (> 50°C) and may be it is a "false" problem if you can keep the air cool enough.

The problem is when the heat "comes up" to the intake after the engine is running very hot and if you restart the engine at rest, the air temperature compensation can be not correct and the A/F is too rich or too lean so the engine (re)start is difficult. May be an "average" value for the correction is better than the one for the dry air. Ex: at 80°C the correction is -18% for dry air and -24% for 50% Hr. This 6% difference can be a cause of bad restart of the engine at iddle.
Hoping I'm clear enough, I'm very interrested in your opinion concerning this particular point.

Robert

[chrislandy]

Hi Robert,
Unless you already have a compensation table for humidity and some sort of sensor for it, I wouldn't worry about it too much.

Even if your engine is forced induction you shouldn't be seeing greater than 40ish degrees (C) at the manifold and preferably much lower. Any hot/humid air in the intake will quickly be purged once you start cranking and you have a compensation table for air temperature and probably a start map so that should prove adequate most situations.

Chris

[robert kenney]

Maybe I am being to blunt here and correct me if I am wrong but is a correction table or map the "Rube Goldberg" way around running the mass flow based EFI in closed loop so small corrections are made by an o2 sensor maintaining Stochometricly correct AFR's. Would not this be a more effective way to keep the engine in tune?

On another but perhaps related subject.
We tune our race car using a air quality meter combining density altitude and water grains content. The newest trend it to use water grains content and DA to adjust our mechanical injection. Humidity is not a large factor of late.

[NickRP]

This is a very interesting topic, yet so far there is no theoretical way to calculate the correction factors 100% correctly, based on the inputs available in commercial (i.e. non laboratory) applications.

I am afraid that the uncertainties in determining air density are far greater than (not-so-significant) effect of the air humidity. Of course, air density could be easily calculated, when air pressure, temperature, composition (inc. water vapor contents) would be known. Knowing the composition also implies that the density of oxygen could be calculated, so having a VE table for the given engine, it would be possible to calculate an exact amount (mass) of fuel required to achieve certain mixture ratio. But it gets more complicated when uncertainties are taken into account - air composition is not known (let alone humidity), so averages have to be used. Air temperature readings are not only affected by sensor production tolerances, but also by sensor positioning (not to be underestimated), unavoidable radiant heat sources or sinks, air stream path etc.

As quite correctly noted by Robert, humidity plays much bigger role at higher air temperatures. However!!!, at temperatures that high, of much higher concern is to prevent catastrophic detonation, than to keep the ideal mixture ratio. And since there is no load dependent temperature compensation maps in most aftermarket ECUs, there will be only one correction factor for air at 50°C, no matter if the engine is idling or running under serious load and being at the border of detonations.

So, I would calculate the compensation values using 50% relative humidity, and use that as a starting point. Furthermore, I would add some fuel above 40°C to prevent detonation. How much, depends really on your engine configuration.

Slightly off topic: 6% richer mixture will under no circumstances cause problems during starting or idling (assuming that at 0% we have an ideal starting or idling mix).

Best regards,
Nik

[Guy Croft]

MODEL POST!

Thanks very much Nik for your professional input - it is always highly valued,

GC

[Robert]

Hi all,

thank you for this information sharing. Concerning my practical case, I will monitor the air temperature with my ECU SW and/or with a data logger. If the temperature doesn't exceed 50°C (I hope !) the correction table will be the one of the "dry air" because it is quite the same of the 50% Hr (RH pardon ;o) ) . We can also consider (?) that in an engine bay the air is quite dry.

Robert

[NickRP]

Of course, you can use the dry air as a reference for temperature compensation.

Why do you assume that the air inside the bonnet is dryer than the surrounding air? No water vapor separation is taking place, as far as I know. To make the air dryer, you would have to cool it down, so the water vapor condenses to certain extent. If you considering that the air is hotter inside, that would only change its relative humidity, which is irrelevant in this case.

By the way, could you publish a photo of your intake system?

[Robert]

Nick,

The question mark meant I was not sure the air is drier ;o)
The photos are attached.
OK for all the explanation. My engine was maped on dyno for the load condition and I complained about the low rev and cranking. So I modified the maping (DTA ECU) of the 500- 2500 revs zone (fuel & advance) and the compensation of the throttle speed (enrichment time in engine turns). The result is satisfiying and I crank/start the engine in almost all conditions and the "manoeuvre" (low speed in roundabout, stop, traffic lights...) are possible without sakad/jerky movement due to the "cut off" of the engine. This is true in winter and spring temperature. At the moment, in south of France, the thermometer displays more or less 30°C... I'm facing problem to crank when the engine is hot. So I wondered if my air temperature compensation was correct. Seeking for info I'm on this forum and the air temperature leads to density which leads to humidity ... The loop is closed ;o)
To be exhaustive, the air temp sensor was not the Bosch indicated in the DTA doc and I shifted the values in the table to be coherent with its variation law. Nevertheless I just changed (this evening) this sensor to a Bosch one with the correction table indicated in DTA examples and corresponding to the air density with a little humidity. I will see this week-end the results.

The photos:
1 global intake system of my Lancia Montecarlo: you can see the horms are ended by a filter on a plate. The room is not huge but a duct sends the fresh air from the front of the car to the filter.
2 detail of the sensor position.
3 top view

Sorry to derive to a too much personnal case but I can be "talkative"... Many thanks for your concern.

Robert

[chrislandy]

I'm facing problem to crank when the engine is hot.

Hi Robert,
What do your water temp. compensations look like?
Chris