Dynamic Cylinder Deactivation

[spirilis]

Just read the link Frank posted. I thought about this idea one time, but didn't have a good answer for the O2 sensor problem--looks like you figured it out, using a Wideband O2 sensor and skewing the value as needed to provide a typical O2 sensor output the ECU will understand. Very cool.

I'd love to see some real numbers behind this. I bet adding a port-by-port Water Injection system would substantially improve the workability of the offline cylinders (from steam vaporization). The use of microprocessor is good too, I was dreaming about using an Arduino (8-bit, can have similar # of digital I/O ports if you use the ATmega2560) due to the open-source community surrounding it.

[JasonG]

Wow, guys really good at repeating key search terms multiple times in every post.
Your SEO skills are impressive.

That said, I've also thought a processor skipping every X number of injections would still run smoth at hwy speed but save fuel. An open source / arduino version would be thoroughly ABA tested here.

[t vago]

If you're not deactivating at least the intake valve, you're not going to see much of an increase in fuel economy.

This website has a number of factual mistakes in it.

First of all, traditional cylinder deactivation schemes do contribute to overall engine efficiency in that the firing cylinders in a variable displacement engine have higher peak compression pressures than with a similar engine with no variable displacement scheme in place. This is in contrast to this item: link

Second, pumping losses are less with a traditional cylinder deactivation scheme than with this DCD gimmick because the deactivated cylinders do not pump air. Therefore, they do not contribute to forming economy-robbing engine vacuum, nor do they contribute to economy-robbing exhaust flow.

Third, since there is no airflow through the deactivated cylinders in a traditional cylinder deactivation scheme, the deactivated cylinders will approach average coolant temperature. This is in contrast to this DCD gimmick, where airflow through deactivated cylinders may actually cool the cylinders below average coolant temperature. This is the so-called "thermodynamic expansion" mentioned with DCD.

I'll leave it up to Mr. DCD there to explain how his electronic gadget is going to fool the engine computer from sensing the extra oxygen in the exhaust stream from his DCD algorithm, and fooling the engine computer from applying extra fuel to compensate for that sensed extra oxygen.

You'd be better off checking for correct tire pressure, than with experiementing with cylinder deactivation schemes that do not shut off airflow to the deactivated cylinders.

[Heihetech]

Originally Posted by spirilis

Just read the link Frank posted. I thought about this idea one time, but didn't have a good answer for the O2 sensor problem--looks like you figured it out, using a Wideband O2 sensor and skewing the value as needed to provide a typical O2 sensor output the ECU will understand. Very cool.

I'd love to see some real numbers behind this. I bet adding a port-by-port Water Injection system would substantially improve the workability of the offline cylinders (from steam vaporization). The use of microprocessor is good too, I was dreaming about using an Arduino (8-bit, can have similar # of digital I/O ports if you use the ATmega2560) due to the open-source community surrounding it.

That's great! Your deep understanding has reached the key point -----
Windband oxygen sensor will guide ECU to close the fuel control loop at
higher Lambda points (Normolized air-fuel ratio). This means less fuel
will be burnt when DCD is on, the reason of fuel saving.

Water injection is also a cool idea, but please be careful don't make the
cylinders too cold. Because intake air already has cooling effect to cylinder,
water injection may cause over-cooling. I think the best way to inject
water is to inject into the hot expanded exhaust, absorbing residual heat
energy from hot exhaust, and then, let the mutilple fluids expand in a
secondary compound cylinder ----- this has become my HEIHE invention
which has been patented.

[Heihetech]

Originally Posted by JasonG

Wow, guys really good at repeating key search terms multiple times in every post.
Your SEO skills are impressive.

That said, I've also thought a processor skipping every X number of injections would still run smoth at hwy speed but save fuel. An open source / arduino version would be thoroughly ABA tested here.

Your dream will become true once DCD Controller is installed onto your
vehicle. ABA tests have been done under steady speed mode and stop-
and-go mode around the city block, yielding 8% to 15% fuel savings. At
idle, the savings may go up to 18%.

[jedi_sol]

Originally Posted by JasonG

Wow, guys really good at repeating key search terms multiple times in every post.
Your SEO skills are impressive.

That said, I've also thought a processor skipping every X number of injections would still run smoth at hwy speed but save fuel. An open source / arduino version would be thoroughly ABA tested here.

Yes, its a bit overkill. Give us some hard data (actual numbers) and testing procedures and we may be interested.

[Heihetech]

Just see the data posted above. Steady speed is 30 mph, using triplecate
2-way drivings. So each data is an average of 6 single trips. Stop-and-
go test is driving around a retangular street block, start from one corner
and keep tuning right until you go back to starting point, slow down before
each right turn, and speed up after each right turn, keep steady speed
when it reaches 25 mph.

The guage used for test is DFC-1 digital fuel consumption meter, similar
to MPGuino, directly sampling signal from fuel injectors.

[DonR]

Originally Posted by Heihetech

That's great! Your deep understanding has reached the key point -----
Windband oxygen sensor will guide ECU to close the fuel control loop at
higher Lambda points (Normolized air-fuel ratio). This means less fuel
will be burnt when DCD is on, the reason of fuel saving.

If you have just the one O2 sensor controlling all the cylinders, won't the injectors still be squirting fuel into the deactivated cylinders, just less of it? Wouldn't this would cause the active cylinders to run lean?
Should you not have the computer cut fuel & spark to individual cylinders on a rotating basis?

[Phantom]

DonR you have it backwards if the O2 sensor is seeing extra air it will add fuel making it rich. Shutting the injectors off no fuel will be injected, if a controller is used it can simulate the correct O2 reading allowing for a switch to skew the results for the extra air. It would be easy to calibrate by running with all cylinders firing get an even reading shut off half see the change in O2 adjust then repete a few times till the change in O2 is ~1-2%.

(O2 mentioned above is not actual O2 but the reading the stock O2 sensor gives.)

[Heihetech]

As I have posted, o2 sensor is a WINDBAND OXYGEN SENSOR, it will
cheat ECU to close fuel loop at higher Lambda point, thus less fuel will
be injected. Regular o2 sensor is no longer operable under DCD control,
it must be replaced into WINDBAND OXYGEN SENSOR.