Increasing EGR flow: gasoline engine

[Teemo]

Originally Posted by razor02097

In a newer vehicle the computer should be able to see changes with the O2 sensors. In theory the ECU should be able to achieve stoich with less fuel...

Not really sure how the 90 VW would do... I have zero experience with mechanical fuel injection systems

My engine has lambda sensor and airflow sensor. I connected the egr directly to the intake manifold -- past airflow sensor and throttle butterfly. The lambda regulation is still working.

For engine with airflow sensor it is possible to add that type of egr (or carb. but that is too history). For engine without airflow sensor(map, or throttle pos/RPM sensor) it is not possible, the ECU will go mad.

[RobertISaar]

a couple of things to keep in mind, since i've researched this a bit before:

temperatures.... they are quite important to consider. the cat WILL heat the exhaust if it's operating correctly, so should you want to draw EGR pre-cat, immediately before the cat will be the coolest location. drawing pre-cat will allow at least some unburned fuel(if any is present) to have a second chance at getting burned in the cylinder to produce any useful work from it. post-cat.... well, if the cat is working well, there should be no unburned hydrocarbons(HC-> CO and H20 reaction), so that advantage will be gone.

also, cool EGR is more dense than hot EGR, so when it's being drawn, it's a higher mass/volume.



there are at least a few US market cars made in the past 20 years that had EGR coolers... the ones that come to mind are Cadillac Northstar powered vehicles. they used engine coolant to drop the temperature of the EGR gasses before they entered the plastic intake manifold. i ASSUME that setup may have been used in other applications with plastic manifolds as well, but i can't think of any off-hand. a side-benefit of this is that if EGR is being utilized early enough, you can actually heat up the coolant faster to get your engine up to full operating temp in less time.



nothing else comes to mind that hasn't been discussed yet other than the desire for intelligent control of EGR... depending on what sensors are already present, you could impliment something as simple as a schmitt trigger(or just a plain comparitor) to only allow EGR above a certain throttle opening. otherwise, you may have to get into a bit more complicated logic.

[UFO]

EGR has been used for a long time. My 1978 Datsun 280Z had an EGR system.

[user removed]

The original purpose of EGR was to reduce NOX emissions which skyrocket when peak combustion camber temperatures rise above around 3300 degrees. In 1976 California required EGR on the 280Z (Nissan-Datsun), while the Federal emissions versions did not reuqire EGR, but the compression ratios on those early EFI engines was about 8.5 to 1.

To a degree adding EGR will increase efficiency, but mapping the amount, to specific load conditions could get tricky. I really requires precision controls and a dedicated ECU response to really cover all of the parameters.

If you can get precise instsntaneous mileage readings and be able to adjust the flow of EGR based on the gains in mileage you may be able to make it work. Making it work automatically may be quite a project.

regards
Mech

[Teemo]

To designin a special electronic control unit is no probelm for me. Writing microcontroller programs is part of my everyday job.
I am considering of using LPG actuator as control valve for EGR:
http://i.ebayimg.com/t/LPG-STEPPER-M...fiNLQ~~_35.JPG

[Teemo]

Finally I got time to make the pictures of current test setup.
Certainly not perfect yet.

[mwebb]

VOLUMETRIC EFFICIENCY is what suffers at part throttle , compression is not changed , there is insufficient air fuel mixture to compress as
VOLUMETRIC EFFICIENCY has been reduced
increasing flow of
EGR improves VOLUMETRIC EFFICIENCY ,until it does'nt there is a limit

there is a closed throttle switch on that system and
there is a WOT switch
you can use inputs from those switches isolated with diodes to your controller to use an actuator to close off EGR flow at idle and WOT
to begin

the DPR controls lambda by changing / regulating upper and lower chamber pressure in the fuel distributor , you can adjust DPR slightly but you should not adjust it without a scope on the 02 / lambda sensor

you can measure DPR current by adding 1 ohm resistor in series to DPR
current should be about 0 to 10 ma engine running hot
voltage will equal current as E=IR voltage equals current times resistance but when resistance equals 1
current equals voltage , measure voltage across the resistor
set co to the middle of the bounce , about 5ma at idle , current will be changing always with 02 sensor input
center the bounce in the middle of the range 0 to 10ma

if current is negative you are in DFCO
current will go high on cold start , shows enrichment open loop
current will go high at WOT , shows enrichment open loop
never frack with CO adjustment without scope on 02 and current measring setup on DPR

installing an upgraded 02 sensor with a heater will improve everything
power up the heater with a relay switched on by ignition

the lambda sensor MUST be switching at cruise for the system to be in fuel control while EGR is functioning , lambda should not change by the addition of post cat inert exhaust as the cat will have used up all remaining 02 in the exhuast stream , in a perfect world , but in reality
you may need to change the CO adjustment slightly

you need a scope to do that
you can increase EGR flow until the 02 sensor starts to show misfire then back off a wee bit
02 sensor waveform at cruise on your car will look like this just as misfire begins 02 sensor is black trace
the 02 sensor waveform with no misfire will appear as a sine wave switching from 150mv to 850mv and back in 100ms or less at 45mph in gear at operating temp

02 sensor in black trace , pressure differential sensor in exhaust shows misfire , blue trace - secondary spark spark wave form in red trace


setup
laptop on front seat


on your system
02 sensor will show misfire before you can feel it , add EGR until 02 sensor shows misfire OR stops switching

[renault_megane_dci]

good project !

[Teemo]

I already have the heated O2 sensor, so that is good

I will take it to plan to set up the ossilloscope. The misfire detection is of cource very important.
Sadly I destroyed my laptop battery last time I did the experiments with engine data logging by leaving it in the car for the night:-15C. Battery only lasts about one minute since then

[arcosine]

Have you noticed any improvment in mileage? I wonder if taking it before the converter would recycle some of the unburnt hyrocarbons.