BMW Group / SGL Group JV opens new Moses Lake carbon fiber plant
Obama Administration withdraws proposed rulemaking on strengthened ozone standards

Volkswagen to implement cylinder deactivation in 4-cylinder 1.4 TSI engines in 2012

Cylinder deactivation (Zylinderabschaltung, ZAS) relies upon two different profiles on the cam pieces: a conventional full profile and a zero-lift cam that is activated (right) under cylinder shut-off. Click to enlarge.

Volkswagen will debut cylinder deactivation (cylinder shut-off) technology in the new 1.4-liter TSI turbocharged, direct-injection engine beginning in 2012. Volkswagen is the first carmaker to implement this technology in a turbocharged four-cylinder engine in large-scale production.

The primary goal of the system is to reduce significantly fuel consumption by temporarily shutting off two of the four cylinders (cylinders 2 and 3) during low to mid loads; Volkswagen says that cylinder shut-off reduces fuel consumption of the 1.4 TSI by 0.4 liter per 100 km in the NEDC driving cycle. When the Stop/Start functionality is integrated, which deactivates the engine in neutral gear, the savings effect adds up to about 0.6 liter per 100 km.

1.4 TSI
The new 1.4 TSI outputs 103 kW (140 PS, 138 hp), with 250 N·m (184 lb-ft) of torque between 1,500 and 4,000 rpm.

The greatest benefits of the technology are realized while driving at constant moderate speeds. At 50 km/h (31 mph), in third or fourth gear, savings amount to nearly one liter per 100 km. Even when running on two cylinders, the 1.4 TSI—with its excellent engine balance—is still very quiet and low in vibration, Volkswagen says.

Cylinder shut-off is active whenever the engine speed of the 1.4 TSI is between 1,400 and 4,000 rpm and its torque is between 25 and 75 N·m (18–55 lb-ft). This applies to nearly 70% of the driving distance in the EU fuel economy driving cycle. As soon as the driver presses the accelerator pedal sufficiently hard, cylinders 2 and 3 are reactivated unnoticed.

Under the deactivation process, the combustion chambers are filled with air—this entrapped fresh air leads to minimal cylinder pressure and therefore to lower energy consumption. Afterwards, the system closes the intake and exhaust valves of cylinders 2 and 3; engine ignition only occurs once per crankshaft revolution. The pistons of the deactivated cylinder are now dragged by the crankshaft. On the other hand, efficiency increases in the two active cylinders, because their operating points are shifted to higher loads.

The valves are closed using a complex set of actuators: on both the intake camshaft and the exhaust camshaft, there are two adjustable sleeves (cam pieces) that are placed on special tooth systems. They are responsible for the eight valves of the second and third cylinders. At the ends of each cam piece, there are two different profiles adjacent to one another: a conventional full profile and a so-called zero-lift cam. The full profiles actuate the roller cam followers, which in turn actuate the valves in four-cylinder operation; that is, they behave like very conventional cams. However, the zero-lift cams rotate over the followers—i.e. they do not actuate them—and the valve springs hold the valves shut. Engine management simultaneously shuts off fuel injection.

Spiral-shaped slots are milled in the outer surfaces of the rotating cam pieces; these slots permit shifting the sleeves a few millimeters along the shafts very quickly; when electromagnetic actuators in the valve cover get a signal from the engine controller, two integrated metal pins engage the slots from outside and move them to their end positions. Finally, the cam pieces are locked in place by spring-loaded balls. As soon as the driver presses the accelerator pedal sufficiently, cylinders 2 and 3 are reactivated.

All mechanical switch-over processes are executed within one-half camshaft revolution; they last between 13 and 36 milliseconds, depending on engine speed. These processes are smoothed by accompanying interventions in ignition and throttle valve control.

Volkswagen utilizes information from the gas pedal sensor to detect the driver’s mode of driving. If the driving exhibits a non-uniform pattern—e.g. while driving in roundabout traffic or in a sporty style on a country road—shut-off functionality is suppressed.

The components for cylinder shut-off weigh a total of just over 3 kilograms. Their actuators, the camshafts and their bearing carriers are integrated in the valve cover. Two roller bearings reduce the friction of the shafts.



".. At 50 km/h (31 mph), in third or fourth gear, savings amount to nearly one liter per 100 km.." ~25% if 60 mpg?


This is one well known way to reduce fuel consumption. It could be done on ALL ICE units.


Let's see if the car will jump like a rabbit when it is running on 2 cylinders. This is one of the main reasons why this technology has not been implemented on 4-cylinder engines before. VW tells us that it will not behave badly. I would like to test drive before I am a believer... For 6 and 8-cylinder engines, this is not such a great problem.

The question is also if this technology has such a big advantage over other options. The combination of fully variable valve technology and a third generation of start-stop, where the engine is shut off also during coasting, is also addressing the low load (and idle) operating area but it does not have the noise, vibration and harshness (NVH) associated with cylinder deactivation. Bosch showed in a recent paper in the MTZ journal that improved start-stop (coasting) has great potential under real-life driving conditions. Most of the technology needed is readily available. Full VVT is more costly but you might note that it is recently used by BMW in combination with both DI and turbocharging, so one could easily figure out that it would work well on the VW TSI engine as well. The combined potential (i.e. not the sum of the measures) of these features amount to roughly the same potential as the one shown by VW but with minimal negative NVH impact.


Variable compression ratio is the best strategy, see MCE5 engine.


Run on 2 cylinders?

What keeps it from shaking like a dog shitting bones?

Roger Pham

As long as the engine has appropriate inertial balance, it will not shake, when the two cylinders fire at regular intervals for the torque damper to do its job. With sufficient torque damping, two-cylinder engine can run very smoothly, except at idling. With a stop-start system, the engine does not have to idle at all.

Let's see if the car will jump like a rabbit when it is running on 2 cylinders.
FTA: "All mechanical switch-over processes are executed within one-half camshaft revolution; they last between 13 and 36 milliseconds, depending on engine speed."

It would be hard to notice, especially if the engine ECU predicts torque demand based on the rate of pedal movement.



    FTA: "All mechanical switch-over processes are executed within one-half camshaft revolution; they last between 13 and 36 milliseconds, depending on engine speed."

IMO, we can consider it as an immediate switch-on of 100% more cylinders. But this is not what is difficult.

Just before the event, the engine is running at a certain state delivering the needed torque with only 2 cylinders at a higher load, from a higher MAP derived from bigger opening of the throttle body.

Just after the activation of those 2 extra cylinders, all the cylinders would have to face a different MAP, to adjust load to achieve nearly the same level of torque as before to avoid the jump like a rabbit effect Peter_XX was cautioning.

Achieving this level of control over a V8 was already difficult and gave bad reputation to the technique long ago, and would be even tougher at an I4 (2:2) than to a V8 (2:6).

Lets hope that harnessing current ECUs higher throughput budget, VW engineers can smooth torque output when switching operating modes.

ECU could a achieve it by coordinating actuation in multiple fronts: throttle, spark timing, and, may be, avoiding fueling the newly activated cylinders for the first cycle(s) while the pressure at the "plumbing" adjusts for the new situation.


With or without "appropriate inertial balance", it WILL shake.

When the two cylinders fire at regular intervals the torque damper absorbs some of the torque variation, but it cannot be well tuned, lest vibration become more severe at all other rpms.

Only one cylinder fires ONCE PER REV.

And the damper cannot be a fluid clutch lest efficiency be really bad.

Meanwhile the motor mounts have a similar problem (trying to keep the passengers from seeing double).

Torque damping, and sophisticated motor mounts reduce the shake somewhat in a two-cylinder engine - like in any good m-m-motor cy-cycle.


Torque modulation via control of ignition timing is very effective. The problem with avoiding fuel is that you'd have a very sudden drop of manifold pressure and increased pumping losses (experienced as hesitation), plus a surge of oxygen in the exhaust which might cause emissions transients.


CelsoS is right - it is a carefully timed dance.

Since the advent of throttle by wire, drivers have had to accept less response but it should be minimized, of course.

Roger Pham

To reassure you, this is a quote from the article:
"Even when running on two cylinders, the 1.4 TSI—with its excellent engine balance—is still very quiet and low in vibration, Volkswagen says."
I'm sure that there will be more tricks in VW's bag than are dreamed of in our imagination.


To simplify this discussion, let´s talk about two things. First, what we could call “mechanical vibration” is not significantly altered (engine balancing is not changed). Second, and more important, the engine torque over the two revolutions of the full cycle (since it is a 4-stroke) will change. If we go from four cylinders to two cylinders, there will be more torque fluctuation, since the power from those two cylinders will have to double. However, it does not help (as ET said) if switching is fast (as it already is as fast for cam-shifting devices). Going fast from a state of smooth operation to a state of high vibration (and back again) will not make things better. What matters is if we can accept - of if VW can reduce - the larger torque fluctuations that inevitably will be the result? For 6 and 8 cylinder engines, this is not a big problem. A V8 that runs on 4 cylinders is not worse than a 4-cylinder engine. Could VW have done anything with engine mounting to overcome the torque fluctuation problem? Do we have experience from similar problems in other applications that could apply here? One answer to the second question is what is done to a diesel engine. This type of engine has higher torque fluctuation (due to high compression, high expansion and no throttling) than a gasoline engine. It is very apparent at idle. The solution to “tame” the diesel engine (more than 20 years ago) was an engine mounting (hydraulic support) where you can change “weakness” and damping by operating a valve. In other words, it is “soft” at idle and “harder” at higher load. Another measure was the double-mass flywheel, which is nowadays used on practically all diesel engines. A new development here is the “pendulum” flywheel, which provide further torque-smoothing. This is used by, e.g. BMW in the “Efficient Dynamics Edition” versions, where the engine can operate at 1000 rpm at full torque with no negative driver perception. One possibility is also to limit operation to low load and somewhat higher speeds but then the potential for reduction of fuel consumption will be limited. VW could have compromised a little here.

What I mentioned above could be a couple of the tricks that Roger Pham is dreaming about. However, these examples are still engines operating on 4 cylinders. On only 2 cylinders, the problem will be worse. Thus, I am still skeptical about how this engine will behave in normal driving or whether I would like it. This is the reason why I would like to test drive first. For example, I do not like the VW 3-cylinder engines. I have not tested the new Fiat 2-cylinder engine but older 2-cylinder engines sounds like lawn movers and shake & rattle. With improvements, I think you could make a 3-cylinder engine acceptable but 2 cylinders is below my threshold limit.

Roger Pham

Thanks, Peter, for your expert explanation.
In my youth, I've riden single-cylinder 4-stroke Honda motorcycle, and it felt acceptable to me...except that the idle was a bit rough, that a Stop-Start system can help. Others may prefer the smoothness of a two or four-cylinder motorbikes, but, when a single-cylinder machine is all one can afford...


I once had a 2-cylinder motor bike. However, it was a 2-stroke, not a 4-stroke. Regarding vibration, it was somewhat better than a 4-stroke parallel twin but I would not mind a smoother ride. I would probably prefer a 4-cylinder engine today. However, for some reason, people tend to like big V2 engines with terrible vibration, i.e. where every bolt and nut tends to come loose after a while. Most of the time, you cannot see anything in the mirrors due to terrible vibration. This also reminds me about the first couple of generations of VW diesels (before TDI). Take off took a while because the steering wheel had so much vibration that you simply had to get a firm grip of it before you could start to drive. VW also conducted a field test with a 2-cylinder engine in one of the small cars but could never handle vibration good enough to launch it. Maybe they learnt something in this process that they now can master.

Roger Pham

In this instance, the very small size of the engine to the mass of the car, at only 1.4 liter, would be of help to minimize torsional vibration. A 1.4 liter engine with 4 cylinders would have a displacement of only 700 cc when running on 2 cylinders. By contrast, a 1.4 liter 2-cylinder engine would have double the vibratory amplitude. So, a micro car having a 2-cylinder engine would have a much more difficult task of controlling vibration than a 4-cylinder engine having the same mass-to-engine-displacement with cylinder deactivation. Furthermore, complete engine inertial balancing in a 2-cylinder inline is a difficult task, unless the 2-cylinder VW engine was 1/2 of the 4-cylinder opposed VW Beetle engine, then, the oppose-cylinder arrangement allows for complete inertial balance even with only 2 cylinders.

The BMW opposed-twin motorcycle with complete mass balancing and regular firing interval is a lot smoother than a Harley-Davidson V-twin, for obvious reasons. However, the vibratory rumbling of the Harley adds to its mystique and image of a brawny tough guy, along with the deep throaty sound of the irregular firing interval further enhancing the "feel" of toughness. The BMW opposed-twin bike sounds just too refined and gentlemanly for many.


Please read my comments once again. It is not about torsional and translational vibration (which does not change); it is about torque fluctuation. If the engine fires on only 2 cylinders instead of 4, the torque must be doubled. Simple!

Roger Pham

Agree, the torque fluctuation will be more noticeable when switching from low-load 4-cylinder firing to high-load 2-cylinder firing.
But in comparison to a car having a 2-cylinder engine of 1.4-liter displacement, running at high load, this 4-cylinder VW engine will feel a lot smoother even when running on only 2 cylinders at high load.
It's all relative, depending on the customer's emphasis on higher comfort level or higher fuel economy.

Perhaps a button on the dashboard to allow the customer to stop this cylinder deactivation can prevent dissatisfaction from customers who value smoothness.


Integrated Starter Generator (ISG), 5-15 kW, if used on engines with cylinder deactivation, could easily make the event of deactivation and re-activation of cylinders almost imperceptible, along with some measures already mentioned. The pancake ISGs usualy have 16-18 poles, so they can very precisely add braking torque within one crankshaft revolution (when going from 2->4 cyls) or an accelerating torque (when going from 4->2 cyls) to make the transition as smooth as possible.
Apparently Honda IMA hybrid system uses exactly that combination, when they operate in (though very limited) all electric mode, where all 4 cylinders get shut-off (in newer Civic Hybrid version), or only 3 cylinders get shut-off (in older Civic Hybrid version).
Also found that in another Honda hybrid model ISG was used to balance the 3-cyl engine at idle to reduce vibrations, cleverly controlling ISG torque, within each crankshaft revolution .
Is it possible that VW will be using an ISG on models with cylinder deactivation?


To my knowledge, VW will not use ISG (or similar) on this car. If so, they would have mentioned it. Deactivation and re-activation is probably not the biggest issue, since 1) switching is fast 2) engine management is sophisticated enough to handle this in a seamless way. In my opinion, torque fluctuation is the issue. In some early development of ISG, it was shown that ISG could be used to smooth torque fluctuations. A starter-alternator would not be suitable to do this trick. So, yes, ISG would be possible but VW is not (yet) using it. One should also note that this kind of smoothing will significantly increase the losses. This might be the reason for that I have seen very little information about it recently.

David Haley

In reading this article, I am set back by the complexity and added weight to accomplish a simple two stage engine. It will take a highly trained mechanic to even decide what a malfunction might be when it breaks. Unfortunately this approach is already obsolete before it is even in production.

The benefits of variable firing are obvious but the design approach has been tried over 60 years that I am aware of. What has not been possible until now is a simple coupling and decoupling of piston pairs. The CV Engine, which is not on the market yet, is a linear piston engine that has eliminated the crankshaft. This allows a hollow center power shaft which permits very simple coupling and decoupling of piston pairs virtually with no limit. A single piston pair engine (A-1) can develop ~ 50 horsepower and weighs 10.2 pounds plus the fuel and ignition system. The CV Engine can be made smaller but the horsepower per pound drops off as you reduce size. The Volkswagon add-on for this feature is 6.6 pounds which is approaching the weight of the entire CV Engine and all those moving parts can not be cheap.

I am certain, knowing Volkswagon, that excellent engineering has gone into making this concept engine ready for market. However in today’s market, innovation can bite you. The CV Engine is out-of-the-box while everyone else is trying to improve what they know which is the same old thing.

Significant details can be found at since it would be inappropriate to take any more of this forums time. As they say “stay tuned”.

Roger Pham

@David Haley,
The Bourke engine did not deliver its promises.

David Haley

The Bourke engine still had a crankshaft and a poor slider crank transfer. This will produce the same non-symmetrical sinusoidal torque curve, have a lot of internal friction and can not have a hollow power shaft. The CV Engine, when compared to any equal displacement crankshaft engine, has more than twice the mechanical torque as well as constant velocity to the power shaft. The Patent has been published and claims a very simple reversing method. Check the web site for some detailed articles.

When looking at other engines, check for some form of the crankshaft to change linear to rotary motion. This generally will cost over 50% of the possible power in mechanical efficiency.


If crankshafts were that lossy they'd have been eliminated long ago. Modern engines don't even use needle bearings (as some 2-cycle engines do because of lubrication issues); they don't have to.

David Haley


I love it. You have just described the acceptance problem by industry and even end users.

Crankshafts are the biggest loss in efficiency in any crankshaft engine which is why 10 years ago we decided to focus on elimination of this loss. Innovation can never happen if one even thinks a solution is easy therefore it would have been done. Innovation often is solving a problem even if no one thinks it is a problem. It may be simple after you do it but not easy. Yes, if it was easy anyone would have done it. That is why the Patent.

Full bearing mounting in an engine is good. However, in a multi piston engine you must either split the bearing or assemble the crankshaft. Now that is not easy. My '56 4 stroke Harley had roller bearing journals.

The comments to this entry are closed.