Atkinson Cycle in F1

[DaveKillens]

Due to the properties of the Atkinson cycle, there are two factors that do limit how high the engine can rev. First off is the extra mass and complexity of any linkages between the crankshaft and piston. That is a physical limitation, determined by mechanical properties. Please remember that reliability has to be considered, since current Formula One engines need to be able to function for more than one or two races. The second is the fuel burn time. Because the Atkinson cycle requires that the exhaust gases be allowed to expand to ambient air pressures, then it takes longer than the typical Otto cycle now practiced in Formula One.
Additionally, since the exhaust gases spend a longer time inside the cylinder than the typical Otto method, then it is logical to assume more heat is passed by the exhaust gases into the cylinder, and engine. This would require redesigning the engine and cooling system to handle the additional load. In other words, because of the Atkinson cycle, larger radiators.
Even if you make the assumption that an Atkinson cycle engine can be made to compete against the current crop, then you need to assess the advantages and disadvantages.
Advantage ...
increased fuel mileage.

Disadvantages ...
Increased complexity and mass of the drivetrain. This added mass is counter to the goal of keeping mass down. The complexity would affect reliability.
Increased inertial mass of the drivetrain. This added mass is also counter to the goal of keeping mass as low as possible for performance.
More heat inside the engine requiring additional cooling. Larger radiators, cooling ducts, pumps, and connecting plumbing would have to be increased. More unwanted mass.

[riff_raff]

ringo asks a valid question. The answer is yes, the Atkinson cycle F1 engine would have better thermal efficiency than a normal F1 engine. But no, an Atkinson cycle engine built to F1 reg's would not have the same power as a normal engine, since the displaced volume of its intake stroke would be much less.

As for the speed capability of an Atkinson linkage mechanism, that's a difficult question to answer. It's not simply the mass of the linkage components that matter, it's the instantaneous dynamic loads resulting from their inertias that is important. And how well the joints in the system can react those loads. That linkage may seem complex and heavy, but who knows? A detailed kinematic and loads analysis may show that it can be made work under the conditions of an F1 engine.

Regards,
Terry

[ringo]

One other consideration is the engine layouts which can be obtained. The crank and eccentric takes up too much room for a V layout. It would be very interesting with a boosted 4 cylinder.

[PlatinumZealot]

So what do you guys think about the Atkinson cycle that uses a Turbine to turn the crankshaft? I think that one will definitely have an advantage.


BTW, I made a solid works Motion model of the one with the linkage just for fun. It was sorta tricky to get it right with all the parts moving without Interfering. I just did the basic movement simulation on it to get the reaction forces. My engine is sorta basic though.. Nasty rod angles,big ugly piston and everything. The inertial force during the extended stroke is much larger than the non extended stroke, but at the same time there are two cranks. (I have to make an otto engine to compare it). You can see the Red tracing of the movement of the connecting rod.



The compression stroke is 78mm and the expansion stroke is 89mm.
Yes.. the strokes are large for my engine. I soon figure out how to make it smaller. (mostly trial and error for this)

I could change the phase of the eccentric to get different compression/expansion ratios. Very interesting.

I realized that the engine has to be basically "flat" so that the linkage can work properly. It hit me that the way the Honda engine looks was no coincidence. This Also rules it out as a V8 (in addition to the space constraints mentioned before).

Very interesting.

[tok-tokkie]

an Atkinson cycle engine built to F1 reg's would not have the same power as a normal engine, since the displaced volume of its intake stroke would be much less.

Is that correct? Is the capacity of the engine not determined by its induction stroke?

The original question was asked with reference to the new regulations where re-fueling is no longer permitted so this car would start with less fuel and thus have a weight and cross sectional area advantage.

[DaveKillens]

an Atkinson cycle engine built to F1 reg's would not have the same power as a normal engine, since the displaced volume of its intake stroke would be much less.

Is that correct? Is the capacity of the engine not determined by its induction stroke?

The original question was asked with reference to the new regulations where re-fueling is no longer permitted so this car would start with less fuel and thus have a weight and cross sectional area advantage.

The theory behind the Atkinson cycle determines that the intake stroke is less than the power stroke. The traditional method of determining displacement goes out the window because the "stroke" is no longer defined. is it the intake or power? Just like calculating the displacement of a Wankel, an alternate method of defining the displacement of an Atkinson cycle needs to be defined.
Maybe a car running with the Atkinson cycle can run with less fuel, but I also believe it will require additional cooling, negating any benefits of reduced cross sectional area and weight.

[PlatinumZealot]

The displacement must be on the intake stroke, because that is where the air enters the engine right.

More cooling? Possibly. I know the piston speed is faster because of the increased stroke, But for the same rpm as an otto engine; I think the cylinder walls are exposed for the same amount of time, and the gas is going to be cooler in that extra exposed area because it expands more. But I don't know what to make of that.

[DaveKillens]

Let's do a simple exercise with your numbers. Of course, they are not definitive, but do illustrate the heat problem.

The compression stroke is 78mm and the expansion stroke is 89mm.

Thus, the expansion stroke is (89/78 = ) 1.141 more than the intake. So the hot exhaust gases spend 14% more time inside the cylinder. Logically, then the cylinders receive more heat compared to a regular Otto cycle.

[tok-tokkie]

Let's do a simple exercise with your numbers. Of course, they are not definitive, but do illustrate the heat problem.

The compression stroke is 78mm and the expansion stroke is 89mm.

Thus, the expansion stroke is (89/78 = ) 1.141 more than the intake. So the hot exhaust gases spend 14% more time inside the cylinder. Logically, then the cylinders receive more heat compared to a regular Otto cycle.

For that maths to be true the crank shaft would have to rotate at different speeds on the two strokes.

[DaveKillens]

For that maths to be true the crank shaft would have to rotate at different speeds on the two strokes.

This is the fundamental concept of the Atkinson cycle, that the exhaust stroke does not stop until the exhaust gas pressure is equal to the ambient air pressure. In order for that to be accomplished mechanically, the intake stroke is shorter (cheap shortcut at present is to keep the intake valve open well past top dead center) or the exhaust stroke is longer.
In order for the mechanicals to fit the theory, the piston movement has to be controlled by something other than just being directly connected to the crankshaft.

[PlatinumZealot]

Yep, my model is modeled after the honda engine on page 1. It has an eccentric shaft that rotates 1/2 speed of the crankshaft (look to the right) But those funky stroke values are just what it ended up to be. I have to change it up a little.

Another thing i just realized. I was wondering if it could be applied to a current F1 engine. I assumed it's a 14:1 CR, and I assumed the peak cylinder pressure is 90bar and bore x stroke is 98*40mm. To get around 1atm the final compression ratio has to be 24! But I guess it can be less in reality because the opening of the exhaust valve. So i used 20.

To get the expansion ratio to be 20 the final stroke required is 58mm! so that's 1.8 cm more stroke! Not very good at all infact it sounds horrible LOL, but i will put in my model and observe it.

On the heat transferred to the coolant, It will be more but not as what it seems.

Remember, the time spent in the cylinder is the same at the same rpms in both types of engines because it is the same number of cycles per second (assuming same cams etc)
It's the piston speed and surface area of exposed cylinder wall that is 14% more at the end of the stroke.
Also at the same time the gas is expanding moving the piston down its temperature drops at every increment.
So the gas is going to be much cooler at the "extra 14% stroke" part of the cylinder wall. The increased piston speed also increased the heat transfer to the coolant.
There are equations out there to calculate these things, but after looking at them I think I will pass for now. But it will be less than the exposed amount.

I still think these issues can be overcome. With some* of the current rules it is going to be extremely difficult. But I wouldn't write off the possibility of the Atkinson cycle in F1 someday.

[malbeare]

n smikle,
I have mucked around with an engine where the expansion stroke change in volume is greater than the intake change in volume and can report that the fuel consumption for the same power output is less. the exhaust and head temperature are down .but peak power is down compared to standard engine, but midrange tourque is up
cheers Malbeare

[PlatinumZealot]

Great input. So it's actually cooler.
I wonder if that type of Power is welcomed in F1? high range vs medium high range?

[riff_raff]

malbeare,

A lower gas temp at EOP would be expected with a greater expansion ratio (ie. an Atkinson cycle), since the process is not adiabatic.

I also agree that the Atkinson cycle would be more efficient, with its more complete expansion process, at least in theory.

But an F1 engine requires, by definition, a stoichiometric combustion condition. So if less oxygen mass is trapped during each intake cycle, then less fuel mass can be combusted. Which in turn, means less heat energy is available for propelling the car forward. Since trapping the greatest mass of an air/fuel mixture within the cylinder is the primary concern of an F1 engine, things like intake mass flow inertias and acoustic pressure wave effects in the intake/exhaust runners have a greater effect on power than expansion ratios.

If you've ever heard an F1 engine at full power, you'd have noticed that it had lots of exhaust noise. The exhaust noise is a function of the pressure energy released when the exhaust valves open. An F1 engine's exhaust valves open with lots of pressure still in the cylinder. That's why they're noisy, but it's still the best for making power.

Regards,
Terry

[PlatinumZealot]

This is a small video of the model I made.