The invention of the internal combustion engine in the 19th century has revolutionized transportation over land, water, and air. Despite their omnipresence in modern day, the operation of an engine may be cryptic. Over the course of this article I’d like to explain the functionality of all the basic engine parts shown in the demonstration below. You can drag it around to see it from other angles:
It’s hard to talk about a mechanical device without visualizing its motion, so many demonstrations in this blog post are animated. By default all animations are enabled, but if you find them distracting, or if you want to save power, you can globally pause them.disabled, but if you’d prefer to have things moving as you read you can globally unpause them.
An engine like this may seem complicated, but we will build it up from first principles. In fact, we’ll start with a significantly simpler way of generating a rotational motion.
Let’s look at a simple crank. It consists of a handle, a crank arm, and a shaft. When a force is applied to the handle the shaft rotates which we can observe by looking at the attached disk:
The force applied at a distance from the shaft generates torque. The harder we push on the handle, the bigger the torque on the shaft. This cranking mechanism is precisely what converts linear force into torque in a manual coffee grinder or a bicycle.
It’s one thing to power something using our own muscles, but the entire point of building an engine is to avoid manual labor and have the device exert the effort instead. To do that we need to find a reliable source of a strong force that is easy to direct. Thankfully, such device was invented hundreds of years ago – a cannon does exactly what we need. In the demonstration below you can observe how a cannon ball is fired from a cannon. The diagonal lines indicate a cross section view – it lets us see what’s going on inside an otherwise obscured region:
As the gun powder is set on fire it quickly produces a huge amount of gases which push the cannon ball down the barrel. Since the ball snugly fits inside it can only go in one direction. While reliable and easy to direct, a cannon ball won’t be very effective at pushing the crank:
We’ve only been able to do a partial turn of the shaft and the cannon ball is long gone. However, with a few modifications we can harness the pushing power of the explosion in a significantly better way.
Firstly, we’ll replace a cannonball with a piston that has a cylindrical shape and a hole drilled in it. We’ll then use a pin to attach to it a rod that can swing freely on a crankshaft:
As the name implies, the crankshaft consists of both the rotating shaft and the crank on which a force is applied. By putting this assembly inside a simplified cannon shell, a cylinder, we’ve managed to solve the problem of the escaping cannon ball, as the piston is limited in its downward movement and will return up as the crankshaft keeps turning:
Notice that the piston has now a minimum and a maximum position it can reach within the cylinder. A single movement over that length in either up or down direction is called a stroke. If we now trigger the explosion, the combustion gases will push the piston down, which turns the crankshaft:
It’s still not a very exciting machine as it only does useful turning work once. To make it more practical we need to keep repeating the cycle of explosions – we have to add in new fuel, trigger a combustion process, and remove the exhaust gases, over and over again.
Solid fuels like black powder are not very practical for an automated machine. It’s much easier to deal with fuels in fluid forms – their intake can be controlled by various valves. We’ll modify the cylinder we’ve built so far by adding new openings at the top of the combustion chamber:
It may be hard to see how the various openings are laid out, so let’s take a look at the cross section view:
Through the first large curved opening we’ll provide a mixture of gasoline and air and through the second one we’ll remove the exhaust gases. Those two openings will be guarded by the intake valve and the exhaust valve. Finally, to light the mixture, we’ll use an electric spark generated by an exposed ends of a wire. Let’s see how all the pieces fit together:
We’re now ready to use this machine to do useful work. At first we’ll open the intake valve while the piston is moving down letting the air with fuel come in which I’ve symbolized using the yellow color. This is the intake stroke:
Once the piston reaches its lowest position the intake valve closes, and the piston starts to move back which compresses the mixture of air and fuel which increases the thermal efficiency of the combustion. This is the compression stroke:
Voltage runs through the open ends of the wire, generating a spark which ignites the air-fuel mixture. The expanding gases created by the combustion push the piston down, creating torque on the crankshaft. This is the power stroke:
Note that the flame propagation inside a cylinder is quite complex, and what you see here is a simplified visualization. The cylinder is now filled with the exhaust gases which we can vent out through another hole by opening the exhaust valve. This is the exhaust stroke:
We’re now back to where we started and the cycle is complete. Let’s look at those four steps together:
Since the piston moves down twice and up twice, it does a total of four strokes and the engine we’ve built is known as a four-stroke engine. Notice that it takes two revolutions of the crankshaft for the piston to do one full cycle of the work as it goes through the four phases: intake, compression, power, and exhaust.
While functional, the engine we’ve built is more of a toy example that doesn’t show a lot of the engineering ingenuity behind many components of real internal combustion engines. Let’s build on the principles we’ve devised so far by constructing a more realistic machine – an engine that one could find in a car.
Let’s start with the biggest and heaviest part of an engine – the engine block. It forms the main body and mounting structure for other parts:
Notice that this block contains four large cylindrical openings that define the four cylinders. Recall that a piston exerts a pushing force on the crankshaft only during the power stroke, so only for about a quarter of time. This uneven action creates a lot of vibration. While it’s often acceptable for smaller engines e.g. in a lawn mower, a typical car engine has more than one cylinder to ensure a more even delivery of power. I’ll discuss these concepts in more depth near the end of the article.
Since the four cylinders are inline, the engine we’ll build is known as an inline four cylinder engine. Other engines may use different arrangements of cylinders, usually in a flat or V-shape configuration.
The sides of the block are reinforced by various ribs to improve the rigidity of the structure – the body has to withhold the power of the explosions inside the cylinders. You may also notice the the top part of the block is perfectly flat – we’ll soon attach another component there. If you look at the cross section of the block you’ll notice that the areas around cylinders are empty:
Those passages are there for the coolant to flow around the cylinders and take the heat of the combustion away. While I’m not going to dive into details of engine cooling, it’s worth noting that engines should run at a specific operating temperature and the coolant pump, thermostats, and radiators make sure that the engine isn’t running too cold or too hot.
Let’s look at the first big part we’ll mount onto the engine – the crankshaft:
Notice that the crankshaft has five main cylindrical parts that define its axis of rotation, they’re called the main journals. There are also four rod journals that are positioned off-axis. All the journals are connected via webs. Note that while the sections have different colors here, the entire crankshaft is made from a single piece of metal.
You may wonder why the two inner rod journals are offset differently than the two outer ones, so let’s pop the piston assemblies on and see how they’ll move on the crankshaft:
Since the rod journals are at different locations each of the four pistons can run at a different phase of the four stroke cycle. Notice that the distance between the center of the main journals and the rod journals defines how far up and down the piston goes in the cylinder.
A real piston and its connecting rod have some mass so they end up creating a weight imbalance on a rotating crankshaft. To counteract that mass, the webs have elongated shape to form a counterweight that helps to even out the inertial forces on the shaft.
One could assume the installation of the crankshaft in the engine block is as simple as putting it directly in a designated spot at the bottom:
Unfortunately, that wouldn’t really work. During engine operation the pistons exert a lot of force on the crankshaft and the main journals would just rub against the housing creating a lot of friction that would wear the parts down. To fix that we need to firstly put in some bearings that will help to make the rotation of the crankshaft smooth:
These strips of metal don’t look like much, but bearings are usually made from a softer material which causes them to wear first which prevents degradation of the crankshaft itself in case any contact occurs. Most of the time, however, the crankshaft doesn’t actually touch the bearings at all. Notice the small hole in the bearing that matches the corresponding hole inside the engine block:
Through that hole the engine pumps oil under pressure. The crankshaft’s diameter is slightly smaller than the bearings’ inner diameter so oil fills the tiny gap between the two surfaces. Presence of oil is critical here as it creates conditions for hydrodynamic lubrication. Oil sticks to the bearings and the crankshaft, but since the crankshaft rotates it creates a variation in velocity of oil between the two surfaces. In the demonstration below the small arrows symbolize the local velocity of the liquid:
The difference of diameters causes a wedge-like shape to develop which then creates an area of increased pressure that lifts the crankshaft journal away from the bearings. Note that the size of the gap in the demonstration is not to scale, but in real running engines the rotating crankshaft should float completely on a very thin surface of oil.
You may have noticed that one half of the bearings also contains a small gutter which creates a small pool of oil under pressure. Moreover, the crankshaft has small holes in it:
Those passages are actually connected inside and the oil from the pool in the bearing travels through the little passages in the crankshaft itself. This brilliant solution distributes the oil from the main journals to the rod journals, which are constantly changing their position inside the engine. The demonstration below shows one of the many typical arrangements of these passages and the presence of oil in an on the crankshaft:
Let’s finally put the crankshaft in. We’ll clamp it down using five end caps that have their corresponding bearings put in and we’ll screw everything together:
Those screws have to be tightened to a precise torque – it has to be high enough so that end caps are able to keep the crankshaft in place despite the force of explosions pushing down on it through the piston rods, but the torque on the screws can’t be too high to avoid any deformation of the circular shape of the final opening in which the crankshaft lies.
The crankshaft itself is there to receive the force from the pistons, so let’s look at at one up close:
Firstly, notice all the empty spaces inside the piston. They’re there to reduce the weight – a piston should be as light as possible to minimize the inertial forces created by its reciprocating motion. In this piston the top part known as the piston crown has a dish-like cavity in it. Other pistons may be flat or have more complicated shapes.
Pistons have actually slightly smaller diameter than the cylinders, otherwise they could seize during movement resulting in a catastrophic engine failure. However, the piston still needs to seal the combustion chamber and prevent gases from leaking around the piston. This problem is solved by piston rings that are placed in the grooves on top of the piston:
On their own piston rings have a fairly big gap, but when placed in the cylinder they’re squeezed into a fitting shape. Note that the fitted ring still has a tiny gap:
While the gap shrinks when a ring gets warmer and expands, the gap should never completely close as the ring may break under pressure. The clearances in the sizing of the rings are very precise. Since the ring is squeezed into a smaller shape, it wants to expand and that tension helps it form a tight seal with the walls of the cylinder. That tension is also reinforced by the pressurized gases getting into the piston grooves and pushing the rings further against the cylinder walls.
The pistons in our engine have three rings with the top two primarily helping to keep the pressure inside the combustion chamber. The third one serves a different role – it’s an oil control ring. The walls of a cylinder under a piston are constantly sprayed with a supply of oil to ensure a smooth movement during strokes. On a downstroke the oil ring scrapes the excessive amount of oil which escapes through the openings in the ring and the groove of the piston:
Top of a piston faces the enormous heat of combustion. The rings are in contact with the piston and the cylinder walls so they heavily participate in heat dissipation. Moreover, since the top part of the piston is in closer contact with the hot gases, it reaches a higher temperature than the lower parts and therefore it expands more. To account for that the pistons are tapered on top so that when the different areas of a piston reach their operating temperatures the shape is more even.
The area of the piston below the rings is called a skirt. Parts of the skirt are, through a thin layer of oil, in contact with cylinder walls during stroke which stabilizes the piston. While they look perfectly round, piston skirts are actually slightly oval.
A piston is attached to its rod via a gudgeon pin. Perhaps you’ve noticed that a piston has tiny grooves near the end of the pin opening. We’ll put snap rings inside them – they’ll prevent the pin from leaving the hole:
Piston rods themselves are very strong as they have to withstand the force of the explosions pressing on the piston during combustion, while also resisting a stretching and pushing forces due to the inertia of the piston changing its direction of movement.
The pistons with the connecting rods and bearings can be slid down the cylinders and attached to the crankshaft:
Recall that the rod bearings are lubricated by the oil coming in through crankshaft passages. Let’s see how the entire thing rotates:
The movement of a piston as the function of the crankshaft angle deserves a closer look. In the demonstration below I marked the maximum, minimum, and a midpoint traveled distance of a piston during its stroke. You can control the rotation of the crankshaft with a slider:
Notice that when the crank arm has done a 90° turn, i.e. halfway through between top to bottom angle, the piston has moved up by less than half of its total stroke distance. It’s a simple geometrical consequence of the length of arms of the triangle formed by the crank arm, the connecting rod, and the vertical baseline.
The maximum piston’s position is known as its top dead center, and its lowest position is known as bottom dead center. As the cylinders move up and down between those two extremes they sweep cylindrical volumes:
The area of a cylinder’s circular cross section A times the stroke length of a piston S define the displacement volume of that cylinder, and the displacement V of the entire engine is the sum of displacements of all of its n cylinders:
If a single cylinder’s displacement is half a liter, then a four cylinder engine would be known as a 2.0-L engine. In the most basic setup, the bigger the engine displacement, the more air the engine can suck in, and the higher its peak power.
With the parts we’ve assembled so far we’re getting close to completing the combustion chamber. We’ve created the walls with the engine block and the movable bottom with the pistons. We just need to seal it from the top – this is where the cylinder head comes in:
There are a lot of openings there. Firstly, notice four large sections at the bottom – these dome-like cavities will form the top of the combustion chamber. Each of those four sections is the same, so let’s look at the individual segment up close:
In this engine each cylinder has four major openings – through two of them the intake air is sucked in and through the other two the exhaust gases are let out. If you look at the head from a side you’ll notice that each pair of the openings is joined into one elliptical hole that exits on the side of the head. Those passages are known as intake and exhaust ports. Additionally, there is a smaller hole drilled centrally through the axis of each opening – they’re there for the intake and exhaust valves:
Modern engines typically use more than one intake and exhaust valves per cylinder as it increases the flow of gases in and out of the combustion chamber. Moreover, the intake valves are usually a little bigger:
When the piston moves down on an intake stroke, the pressure difference it creates is no larger than that of the incoming air, which, for traditional engines, is roughly equal to the atmospheric pressure. However, at the end of the combustion stroke the pressure inside the cylinder is many times higher than the atmospheric pressure, so it’s significantly easier to expel the combustion gases than to suck the intake air in. For this reason the intake openings and valves are larger.
Let’s look at the operation of those valves up close. Firstly, the seal they form has to be very tight so that the only pathway for the expanding gases created in the combustion is to push the piston down. The edges of a valve and its seat have a conical section so that the seal becomes tighter as the valve is pressed up:
In our toy engine the valves magically opened and closed on their own, so let’s see how it can be actually done in practice. To keep a valve shut we’re going to use a spring to keep tension on a closed valve. We can then lock the spring against the valve using simple locking mechanism that consists of two valve keepers, a spring retainer, and an inverted bucket:
Notice that the top part of a valve has a little groove in it so that the keepers can lock in place. The keepers themselves form a section of a cone that the retainer wedges against:
Since the keepers are locked in the grooves, the retainer can’t move which holds the spring under tension. The bucket provides a big and smooth surface for a force to transfer onto a valve – now whenever we push on the bucket the spring will push it back in place:
We’ve got the return mechanism all figured out, but this still leaves the problem of actually pressing the valves – they need to be opened at a certain cadence which depends on the movement of the piston inside the cylinder. That periodic nature of the operation implies that we should use some sort of rotary motion to push the valves.
We can shape a piece of metal so that it pushes on the valve at different offsets as it rotates on a shaft – it’s known as a cam. The spring ensures that the bucket is tightly pressed against the cam and follows its shape:
The shape of the cam defines when, for how long, and how much the valve is opened. In the demonstration below you can control the height and angular span of the section of the cam profile that deviates from a circle and see how it affects the position of the bucket and thus valve lift at different angles. The plot in the upper part shows the offset of the bucket relative to its normal position as the function of cam rotation angle:
The shape of the cam is critical for defining the way the engine operates. All cams for a set of intake or exhaust valves are usually placed on a single camshaft:
Most modern engines use two camshafts, one for intake valves and another for exhaust valves. Let’s see how the camshafts open and close valves during a typical engine operation:
All cylinders go through four stages in a predefined order. The timing of the valves is actually not as straightforward as it was in our toy engine so let’s look at it up close:
Firstly, notice that the intake valves close after the piston reaches the bottom end of the intake stroke – the air coming through the valve has some inertia, which, especially at high engine speeds, makes it pile up in the cylinder despite the opposite movement of the piston.
Similarly, the exhaust valves open before the piston reaches the bottom of the power stroke, as the majority of the useful work has already been done by the gases and the pressure surplus in the cylinder should be minimized so that the exhaust stroke doesn’t have to actively compress the exhaust gases.
The intake valves open slightly before the piston reaches the top end of the exhaust stroke. When combustion gases escape through the exhaust valves they help to create a “scavenging” effect that helps to pull the intake air in. For that reason the exhaust valves close after the piston reaches the top of the exhaust stroke.
An engine running at low speed may have a different ideal parameters than an engine running at full speed, so many modern engines use a few methods to vary both the timing and lift of the valves during operation.
We can now assemble the pieces together. While the top surface of the engine block and the bottom surface of the cylinder head are very flat and smooth, they need to form a perfect seal as the have to prevent the combustion gases from leaking through. A gasket, often made from a softer, compressible metal, is placed in-between the two and the head is bolted to the block:
The bolts are tightened to the predetermined torque in multiple steps and in a specific order to ensure that the head doesn’t deform during assembly. The sturdiness of the head bolts is critical as they literally have to contain the force of the explosions inside the cylinder.
The camshafts are then installed – they’re held in place by their caps and bolts:
The only remaining piece of the puzzle is to how to turn the camshafts while ensuring they’re synchronized to the movement of the pistons. To achieve this, most engines use a rubber timing belt that is driven by a gear mounted to the crankshaft itself. The timing belt is teethed so that it locks in the notches on the timing gears, the rollers keep the belt in tension:
Recall that in a four stroke engine a single cycle of operation requires two full revolutions of the crankshaft as the piston goes through intake, compression, power, and exhaust phases. However, the intake and exhaust valves open just once during that cycle so the camshafts should do just a single revolution during that time.
To fulfill this requirement the crankshaft gear is twice as small as the camshaft gears – this ensures that the camshafts rotate just once when the crankshaft rotates twice which you can verify by observing a small black dots at the perimeter of the gears:
In the examples so far we’ve assumed that the intake valve let in a mixture of air and fuel, and that indeed was the case in older engines that used a carburetor to create the mix.
Modern engines, however, use a fuel injection system where the amount and timing of fuel injection is controlled by an electronic Engine Control Unit, often abbreviated as ECU. At appropriate time the solenoid inside the injector is energized which electromagnetically pulls the needle up, which in turns lets the pressurized fuel escape through the tiny outlet holes. When the power to the solenoid is cut, the spring pushes the needle back to seal the nozzle:
In some engines the injection happens in the intake port, very close to the intake valve itself, but in our engine we’ll use a direct injection system in which the fuel is put directly into the cylinder itself.
The fuel and air mixture in the combustion chamber is lit by a spark plug. In a simplified form a spark plug consists of two pieces of metal separated by a ceramic insulator. The outer shell is connected to the engine body which acts as a ground, and the central electrode is connected to a source of voltage high enough to bridge the gap to the tip of the outer shell which creates a spark:
That high voltage is generated by an ignition coil. In older engines there was a single coil that sequentially provided high voltage to individual spark plugs, but in modern engines each spark plug will usually have its own coil with the discharge controlled by the ECU.
With a set of injectors and spark plugs in hand we can finally fill the remaining holes in the engine head. I’ll hide the rest of the engine so that we get a better view as to where they fit:
Let’s see how an injector and a spark plug work together during the strokes. Note that normally the injector is attached to a fuel rail feeding it highly pressurized gasoline and a spark plug is connected to its coil, but for the sake of clarity they’re not being shown here. The air is depicted as a blue gas that turns yellow when mixed with fuel:
You may notice that the spark plug fires before the piston reaches the end of the compression phase – it’s done on purpose since it takes a moment for the burning of the air-fuel mixture to begin. Let’s take a look at all four cylinders firing in sequence:
All the demonstrations in this article run at very slow speeds, but it’s worth mentioning how fast things happen in an actual car. For an engine running at a modest speed of 1500 revolutions per minute there are 25 revolutions of the crankshaft every second. Each cylinder fires once per two revolutions of the crankshaft, but since we have four cylinders, there are actually around 50 explosions per second happening in an engine running at that speed.
The ratio of air to fuel in the combustion chamber is important as simply adding more fuel doesn’t necessarily make the cylinder pressure bigger as we’ll just end up with incompletely burnt gasoline. To actually increase the speed of the engine we need to increase the supply of fuel and air.
Let’s look at the pressure inside the cylinder a bit closer. During the intake stroke the piston creates a negative pressure difference which sucks the air in. During the compression stroke the pressure increases due to shrinking volume, only to increase even more due to combustion. It’s finally reduced by the expanding volume of the chamber as the piston goes down during the power stroke:
The pushing force generated on the piston is proportional to the pressure in the cylinder, however, the torque generated by that force is also affected by other factors. Firstly, observe that during the compression stroke, the pressure inside the chamber pushes on the piston and through the rod against the rotation of the crankshaft:
Secondly, the magnitude of torque depends on the effective length of the arm of force, but that length changes during crankshaft rotation. For instance, when the piston is at the top dead center the gases merely push the crankshaft down, but they don’t turn it because the force arm has the length of zero. In the demonstration below the red dashed line shows the direction of the rod force and the black line show the arm of force on the crank:
If we account for these effects we can calculate that the torque generated by the pressure inside one cylinder is roughly as follows:
However, these are not all the forces that affect the crankshaft. Both piston and its connecting rod have some mass and during crankshaft’s rotation they keep changing their direction of motion. Let’s look at the plot of the velocity of a piston as it moves up and down the cylinder when the crankshaft rotates with a constant angular velocity:
You may be surprised that the plot isn’t symmetrical, but it’s just a consequence of the already discussed behavior of the crank mechanism taking more time to move through the lower half of the stroke compared to the upper half of the stroke.
Let’s consider a piston in a top dead center position – as the piston reaches that point its velocity is 0, so the crankshaft has to drag the piston down. Roughly halfway through its stroke the piston reaches its maximum velocity, and now the crankshaft has to actually slow the piston down so that it stops moving by the time it reaches the bottom dead center position. The piston, however, wants to keep going, so it exerts a pushing force. These inertial forces keep oscillating back and forth creating an inertial torque on the crankshaft:
The magnitude of the inertial forces depends on the speed of the piston and thus on the crankshaft’s rotational velocity. That system is very dynamic since crankshaft’s rotational velocity in turn is affected by the inertial forces acting on it.
The resulting output crankshaft’s torque is a sum of these pressure-based and inertia-based torques. The cumulative diagram of torque on the crankshaft from a single piston looks roughly like this:
As you can see, the torque created by a single piston is very varied. Even when we overlap the resulting torque from all four pistons the total torque is still fairly uneven:
A torque T acting on a shaft creates an angular acceleration α that is proportional to that torque:
This is a rotational equivalent of traditional linear equation that ties force F to mass m and linear acceleration a:
In rotational motion the equivalent of mass m is moment of inertia I. Similarly to how velocity is affected by acceleration, angular velocity is affected by angular acceleration. Let’s see how angular velocity of the crankshaft varies over time under a constant load:
As you can see the value fluctuates a lot. To reduce the angular acceleration and thus variation in angular velocity we have to increase the rotational inertia I of the system. For that purpose a heavy flywheel is attached to the crankshaft with a bunch of bolts:
Because the flywheel is heavy and has a large moment of inertia the variation in angular velocity of the crankshaft is reduced which makes the engine work more evenly:
Notice that the flywheel has gear teeth cut on its perimeter. Those teeth mesh with a pinion gear that is powered by an electric starter motor when the engine is being turned on.
Once the flywheel starts moving its inertia helps to keep the crankshaft going which in turn lets the engine continue to operate on its own. Note that while the high inertia of the flywheel helps to smooth out the variation in angular velocity, it also comes at a cost – a very heavy flywheel is difficult to spin up so the engine becomes less responsive to throttle input.
In cars with manual transmission an engaged clutch presses against the flywheel to transfer the rotational motion to the transmission and further down to the wheels. Cars with automatic transmission don’t have a flywheel but instead they use a flexplate which is connected to a torque converter which serves as a source of large inertia to smooth out the engine’s efforts.
The journey through our engine ends here, but there are still many components needed to fully embed an engine in a car. The previously mentioned cooling system ensures the engine is kept at an appropriate operating temperature. The intake and exhaust manifolds direct the flow of gases into and out of the cylinders. An oil pump, oil filter, and oil passages in the engine block and cylinder head ensure all components are properly lubricated. Many modern engines also employ a turbocharger that uses the exhaust gases to increase the amount of air that gets pushed into the cylinders. All those components make sure that the engine can operate at expected level of power and efficiency.
Engineering Explained is one of the most popular channels dedicated to car technologies enthusiasts. Over the years Jason Fenske has covered a breadth of topics including a comparisons of injection systems, intake manifold design, and even rotary engines.
How a Car Works is a fantastic video course on the operation of a car. In high quality episodes the presenter goes into a lot more details than what I’ve described here. While the course is paid, there are some free preview videos available on YouTube.
Unchallenged for decades, the internal combustion engines in cars are slowly being supplanted by their electric equivalents, which are simpler, quieter, and more environmentally friendly.
Despite their drawbacks, classic engines still have something mythical about them – their intricate mechanisms are synchronized together to create carefully controlled conditions for harnessing fire in a truly Promethean way.