Dynamic Cylinder Deactivation

[christofoo]

Heihetech,

Thanks for posting this. DCD sounds very interesting.

However, much earlier in this post you stated that emissions testing could be passed by deactivating or uninstalling DCD.

What I would like is a definitive answer whether it actually results in significantly higher emissions, regardless of regulations. Have you, or will you, put your invention on an emissions test dynamometer, compare it without DCD, and present the raw data?

Although there is extra oxygen reaching the cat during DCD operation, I wonder if it actually remains efficient - if the closed loop operation of the cylinders that are firing has the same properties as without DCD, then the cat may remain efficient. My understanding is the cat needs to get a small amount of HC/CO as reactants. This should occur if DCD works properly. But I need to see real-world proof.

I would even suggest that DCD should be programmed to be inactive for the first 120 seconds of operation, so that it doesn't slow down cat light-off. I would be further impressed with cold start-up emissions data with and without DCD.

I live in an area where smog is a problem. For example, in the summer, even when the air appears clear, the ground-level ozone affects me when I ride my bike around town. At times I have a chronic cough. Is that fair? Even if I were not personally affected, and only the old people suffered, would that be fair?

Even if you don't have problems with regulations, if DCD results in a tangible tail-pipe smog emission increase, I personally don't want it. And I personally don't want my neighbor to have it, and I would probably write my local legislators and ask them to outlaw it. Saving gas at the expense of others is selfish.

Although it may be an non-issue in areas geographically less prone to smog.

Thanks,
Chris

[christofoo]

Heihetech,

Any updates? From your webpage this appears to have fallen inactive, perhaps from lack of funding. If so I'm sorry as it seemed promising.

I'm also curious whether an exhaust gas test was ever done on a dyno.

... or if you ran into regulatory issues.

Thanks,
Chris

[Jyden]

Originally Posted by Heihetech

Dear All,

"DCD Controller for Fuel Saving Retrofitting", an introduction article with a
photo of the controller kit attached, has been posted online as the entry
of "Create the Future Design Contest". Please help to vote our DCD controller.

Create the Future Design Contest :: DCD Controller for Fuel Saving Retrofitting

Thanks so much!!!

ABOUT THE ENTRANT

Name: Yuanping ZHAO

Type of entry: individual

Software used for this entry:
ORCAD

Patent status: pending

[t vago]

You may want to use this bypass kit for any sort of DCD scheme made by HeiHeiTech, here.

[christofoo]

I realize there is a lot of skepticism here directed toward DCD, however I think it's quite clear that it will improve FE if implemented properly, the principle has been demonstrated previously and you'd expect DCD - or any other valve-open-cylinder-deactivation - to reduce pumping losses almost certainly on first principles. How can I say that?

Try a 'thought experiment' directed at ICE loss mechanisms:

8 cylinder sports car, 60 mph on flat ground (setup for high throttle loss in stock configuration)

Case 1: 8 cylinders fire, closed loop

Case 2: 2 cylinders fire, closed loop (wide-band O2 sensor)

Clearly in both cases the mechanical friction losses and accessory losses are the same due to identical rpm.

But internal fluid dynamics are not identical. In both cases the air volume flow rate as measured in the manifold will be the identical, since the volumetric displacement is identical.

However, since in Case 2 there are 6 cylinders that are not contributing energy via combustion, the 2 cylinders that are contributing must exert at a higher power level per cylinder.

Therefore in Case 2 the density of air in the manifold must be higher than in Case 1 to maintain stoich AFR. Therefore in Case 2 the throttle must be more open than in Case 1.

How then will total fluid (pumping) loss be different between the two cases? In Case 2 there is higher intake mass flow, so it's conceivable for valve, intake and exhaust system losses to compensate for reduced throttle losses, however in practical terms throttle loss always dominates. I.e. if you run out of gas descending a mountain, at what throttle position will the engine braking be most powerful? The answer is always low-throttle, the lower the throttle the higher the air flow power loss, and the more powerful the engine brake.

Case 2 will always have have lower BSFC in practical petrol ICE's, and higher FE.

======================================

So this may be very likely to work for FE, but I think this invention almost certainly dies on emissions.

Somehow I hadn't noticed or forgotten about Old Tele Man's post http://ecomodder.com/forum/showthrea...tml#post326136 It seems the EPA has already ruled on this class of mod and it's considered tampering.

I've thought this through some more since my last post, and I think the EPA is correct and that this will apply to newer generation engines as well. Any type of valve-open-deactivation is necessarily going to suffer from quenching the cat with cold air and also the EGR will be contaminated with O2. So this is now dead in my book. (As a separate but related thought, I'd also considered spoofing the TPS to get DFCO with wide-open throttle as a simplification of P&G/EOC, but it suffers from the same emission problems as DCD (well, give or take the EGR issue).

Earlier Heihetech argued that DCD is similar to DFCO regarding cat temperature, but I think that is incorrect. DFCO produces a much smaller mass of exhaust air due to low throttle and so DFCO will require many more revolutions to quench the cat. This is aside from differing usage profiles.

So I expect DCD to be illegal in the US with no possibility of obtaining EPA approval...

... until proven otherwise. (If I'd thought this all through already I wouldn't have bumped the thread, sorry.)

[serialk11r]

Yea, valves open=cat doesn't treat NOx anymore due to excess air. It seems that OEM emissions is impossible to pass without the cat in tip top shape. However, tailpipe sniffer smog tests can be passed with just a good wideband O2 sensor and no cat apparently, at least for some engines.

[t vago]

The above diagram shows an engine cylinder pressure/volume diagram, showing the amount of work needed and the amount of work generated in that cylinder. The power loop area represents the amount of work generated in the cylinder. The pumping loop area represents the amount of work being taken away from the cylinder, basically in order to maintain an intake vacuum. There's more work being taken away than that, but this pumping loop area is directly related to cylinder output.

Now, take away the power loop area, such as with removing fuel from the cylinder by cutting out a fuel injector (like DCD does), and you still have pumping work to consider. The deactivated cylinder is still pumping air through itself. On a 4-banger pushing a car forward at a given constant speed, the remaining 3 cylinders are doing the same pumping work as the deactivated cylinder. However, the remaining 3 cylinders also have to produce the same amount of power as before. They also have to produce the power that the deactivated cylinder would have made, had it not been starved of fuel. This is so that the 4-banger can still push the car forward at that same given speed, and cover the pumping work that the deactivated cylinder is consuming.

Since DCD only cuts out a cylinder once per firing cycle, intake manifold vacuum never gets a chance to significantly lower, which would otherwise lessen the amount of pumping work needed by the 4-banger. At best, there is no noticeable gain in fuel economy. In real-world conditions, the OBD-II engine computer would notice the misfires being induced by DCD, and would throw the engine into a less-efficient limp-in mode.



Compare that to the above p-v diagram. This represents a cylinder which has its intake and exhaust valves deactivated. The shut-off piston does no work, so it has no power loop area. Notice also that the shut-off piston needs no work, since it is not pumping anything. Therefore, its pumping loop area is also not there. The remaining cylinders still have to increase their work output to compensate for the deactivated cylinder, but now they don't have to also cover the deactivated cylinder's pumping work, because the deactivated cylinder is not pumping anything. This allows for a small, but measurable, savings in fuel.

[serialk11r]

t vago, you're forgetting that in order to produce the same amount of power, the manifold pressure needs to be higher.

[t vago]

Originally Posted by serialk11r

t vago, you're forgetting that in order to produce the same amount of power, the manifold pressure needs to be higher.

Originally Posted by t vago

Since DCD only cuts out a cylinder once per firing cycle, intake manifold vacuum never gets a chance to significantly lower, which would otherwise lessen the amount of pumping work needed by the 4-banger. At best, there is no noticeable gain in fuel economy. In real-world conditions, the OBD-II engine computer would notice the misfires being induced by DCD, and would throw the engine into a less-efficient limp-in mode.

[razor02097]

I don't mean to knock a product but I believe it is a really bad idea. But I see 2 major issues besides causing a check engine light due to misfires...

When you deactivate a cylinder by preventing the injector firing all of the air from that cylinder gets dumped into the exhaust. During normal closed loop operation the O2 sensor will sense a rise in O2 which tells the computer the mixture is too lean. The computer then tries to correct by adding fuel... Not only does it add fuel... it adds it to the whole bank. If the device disables cylinders in both banks you will have an engine constantly running rich. Bummer...

Okay lets say you can trick the O2 sensor somehow... You have other losses too with an engine not designed for cylinder deactivation...not only are the other cylinders fighting the friction of the dead cylinder but they have to work to push that cylinder through the compression stroke. You may say "Ah but after it reaches TDC the pressure pushes the piston back down". You would be correct but nothing is perfect there will be small pressure losses so the volume of air at TDC will not be the same as BDC.... double bummer...



Even before I came to EM.com I have played with the idea of disabling a cylinder at idle. As soon as that cylinder is disabled, the moment the O2 sensor reacts the PW and DC of the injector bank goes up. Even poking the button briefly at 800 RPM you can see the ECU compensate right away. That is an antiquated ECU dinosaur from 1995 I can only imagine what a newer ECU would do.