One Piston in This Engine Weighs More Than a Car — And a Man Can Stand Inside the Cylinder

A single piston out of this engine weighs five and a half tons. Not the engine, not an assembly of parts. One piston and it weighs more than three family cars stacked on top of each other. The piston assembly stands six meters tall, which is 20 feet, or roughly the height of a two-story house. The crankshaft it drives weighs 300 tons, 660,000 pounds, and it is the single heaviest component in the machine. The bore of each cylinder is 960 millimeters, which is over three feet across. The stroke is two and a half meters, so each cylinder is a steel tube more than eight feet deep and wide enough that a person can stand inside it with room to spare. Each of those cylinders displaces 1,820 liters. For comparison, a large family car engine displaces about 2 liters. One cylinder of this engine is 900 of them. Fourteen of those cylinders in a straight line produce 107,390 horsepower at a hundred and two revolutions per minute. The whole assembly weighs 2,300 tons, stands 13 and a half meters high, and runs 27 meters long. This is the Wärtsilä-Sulzer RTA 96C, and this video is about what is actually inside it, because the outside is a photograph everybody has seen. The inside contains engineering that does not exist in any engine most people have ever worked on, and the single most important piece of it is a component that a car engine does not have at all. In your car, the connecting rod runs from the crankshaft directly up to the piston, and it is pinned to that piston with a gudgeon pin. That arrangement has a consequence, and every engine that uses it lives with the consequence, whether the designer likes it or not. As the crankshaft turns, the bottom of the connecting rod swings in a circle. The top of it is attached to the piston, which can only move up and down inside a cylinder. So the rod has to lean, first one way and then the other, twice per revolution. And a leaning rod does not push straight down. It pushes down and sideways. That sideways force is called piston thrust, and it is transmitted through the piston skirt into the cylinder wall. It is why the piston has a skirt at all, and it is why cylinder bores wear oval rather than evenly round over the life of an engine, worn most on the two sides where the thrust acts. In a car, it is entirely manageable. The forces are modest. The piston weighs a few hundred grams, and the engine is expected to last a few thousand hours before anybody minds. In an engine with a five-and-a-half-ton piston running continuously for years, it is not acceptable at all. The thrust would be enormous, the liner would wear rapidly, and replacing a cylinder liner on this machine is not a weekend job. So the RTA-96C does not attach the connecting rod to the piston. The top of the connecting rod attaches instead to a component called a crosshead, which rides in guide channels machined into the engine structure. Those channels take the entire sideways force, transmitting it into the frame of the engine rather than into the cylinder liner. A separate long piston rod then runs upward from the crosshead to the piston itself. Because the crosshead is constrained by its guides and can only move up and down, that piston rod is always perfectly vertical. It never leans, not by a fraction of a degree, which means the piston is pushed straight down with no sideways component at all. The cylinder liner is therefore doing only one job. It contains the pressure and guides the piston. It is not being used as a bearing surface to absorb thrust loads, which is what a car engine's bore is doing whether anybody intends it or not. That is the crosshead engine, and it is why these machines can run for years between overhauls when the pistons weigh as much as a lorry. It is not a new idea. Crossheads were standard on large stationary steam engines in the 19th century for exactly the same reason. Anywhere the reciprocating masses are enormous and the machine is expected to run for decades, the crosshead reappears. It costs height, which is why no car uses one. This engine is 13 and a half meters tall and can afford it. There is a second benefit that follows directly from it, and it solves a problem specific to marine engines. Because the piston rod is always vertical and never moves sideways, it can pass through a tight seal in a diaphragm plate below the piston. A rod that leans cannot be sealed that way. Any seal would be worn open within hours by the sideways movement. It is only the crosshead's rigid guidance that makes a tight seal possible at all. That seal separates the cylinder space above from the crankcase below completely. In a car engine, there is no such separation. The underside of the piston is open to the crankcase, and that has two consequences that everybody simply accepts. The first is that combustion products blow past the piston rings and contaminate the oil. That is why engine oil turns black, and it is the main reason it has to be changed. The second is that the same oil is doing two entirely different jobs. It is lubricating the crankshaft bearings, where it needs to last and stay clean, and it is lubricating the cylinder walls, where it is exposed to combustion heat and acids and gets burnt. One fluid, two incompatible requirements, and the formulation is a compromise between them. In this engine, those two spaces are sealed apart, so the lubrication is split, and each oil can be formulated for exactly one job. The crankcase gets one oil, designed to look after bearings, resist oxidation, and last a very long time. It is filtered and recirculated in the normal way, and because combustion products cannot reach it, it stays clean. The cylinders get a completely different oil, injected continuously and in timed doses onto the liner walls, and that oil is consumed as it works. It is not recirculated at all. It does its job once and is burned, and it is doing something a car engine's oil never has to do. Heavy fuel oil contains high levels of sulfur. When sulfur burns, it produces sulfur oxides, and those combine with water vapor from combustion to form sulfuric acid inside the cylinder on the liner wall while the engine is running. That acid attacks metal. Left alone, it would eat the cylinder liner and the piston rings. So the cylinder lubricant is formulated to neutralize acid as well as to lubricate, and it is dosed continuously, precisely because it is being consumed by that chemistry as fast as it is being applied. Two oils, two circuits, two entirely different jobs, made possible by a seal that only works because the piston rod never leans. Now the scavenging, because this is a two-stroke, and two-strokes have a breathing problem. A two-stroke fires every revolution rather than every second revolution, which is why it produces so much power for its size. Twice as many power strokes for the same number of turns. But it pays for that with the hardest problem in engine design. It has to get the exhaust out and the fresh air in during the brief moment near the bottom of the stroke, rather than having an entire stroke dedicated to each. Get it wrong, and burnt gas stays behind, diluting the next charge, or fresh air blows straight through and out of the exhaust, wasting the work of compressing it. The RTA-96C uses uniflow scavenging. Air enters through ports around the bottom of the cylinder liner, uncovered by the descending piston, and exhaust leaves through a valve in the cylinder head, at the top. Everything moves in one direction, upward, with the incoming air column pushing the burnt gas ahead of it, and out through the valve like a plunger. That is the best behaved scavenging arrangement there is, because nothing has to turn around, and nothing gets left in a corner. It is why every serious two-stroke diesel uses it, from truck engines up to this. And here is the detail that shows how completely this engine has been thought through. The descending piston is used to compress incoming combustion air for the adjacent cylinders. The piston is doing two jobs at once. On its way down, it is producing power, and it is simultaneously acting as an air pump for its neighbors, using the underside of itself and the sealed space below. That is only possible because of the crosshead arrangement. The space beneath the piston is sealed off from the crankcase, so it can be used as a pumping chamber. In a conventional engine, that space is simply open to the sump, and cannot do anything useful. And there is a third effect from the same arrangement. Compressing that air cushions the piston as it approaches the bottom of its stroke, which removes some of the load from the bearings. Consider what that means. A five-and-a-half-ton piston arriving at the bottom of a two-and-a-half meter stroke has an enormous amount of momentum to reverse, and every bit of that has to be absorbed by the connecting rod and the bearings. Trapping air underneath it turns some of that violence into a gentle spring. One component, three functions, and the third one is essentially free. The exhaust valves are worth a note of their own, because of what is not there alongside them. They are operated by electronically controlled common rail hydraulics. There is no camshaft in this engine, no chain gear, no mechanical fuel pumps, no hydraulic actuators of the conventional kind. The common rail system replaced all of it. That deserves a moment, because a camshaft has been part of essentially every four-stroke and most two-stroke engines since the beginning. A camshaft is a rotating bar with lobes ground into it, and each lobe opens a valve at a fixed point in the cycle for a fixed duration. The shape of the lobe is the timing, permanently, which means the timing is a compromise from the day the shaft is ground. It is correct at one operating point and approximately correct at every other speed and load the engine will ever see. Removing it and controlling everything electronically means the valve timing and the fuel injection timing are no longer set in metal. They are set in software, and they can be varied continuously according to load, speed, fuel quality, and operating conditions. On an engine that spends weeks at a time at a steady output and then maneuvers in a harbor at a fraction of it, that flexibility is worth a great deal. The results Wärtsilä quote are better performance at low revolutions, lower fuel consumption, and lower harmful emissions, all from deleting a shaft that had been standard practice for a century. Now the fuel figures, which are the numbers that put the scale of this machine into human terms, each cylinder receives approximately a hundred and sixty grams of fuel per cycle at full load. A hundred and sixty grams, that is roughly the weight of an apple, injected into a cylinder that holds one thousand eight hundred and twenty liters. That sounds impossibly small until you follow it through. It happens a hundred and two times a minute in fourteen cylinders, and the total consumption reaches up to 250 tons of fuel per day. 250 tons per day, that is the mass of roughly 200 family cars burned every 24 hours to move one ship. Each individual cylinder produces 7,780 horsepower. One cylinder, 7,780 horsepower. A Formula One car produces about a tenth of that from its entire engine, and a large lorry produces about one twentieth. The torque figure is 5,608,312 pound-feet at a hundred and two revolutions per minute. Now, building and maintaining one of these, because the practical reality is as remarkable as the specifications, the components are so large that they are moved on rail systems inside the works, not cranes lifting parts across a factory floor. Rails laid in the ground because the pieces are too heavy and too awkward to handle in any other way. Installing the cylinders takes a team of 25 specialists approximately three months. Three months for a team of 25 people to install the cylinders in one engine. Anybody who has spent a Saturday afternoon fitting a cylinder head will find that figure difficult to absorb. And once running, the maintenance interval is measured differently from anything on land. Wärtsilä's own technical material describes design measures that extended the time between overhauls for the cylinder components, including piston ring renewal, to at least three years. Three years between piston ring changes on an engine running continuously, not three years of occasional weekend use. Three years of turning hour after hour, day and night, crossing oceans. The way that was achieved is instructive, and it is the opposite of brute force. It came from careful attention to what engineers call piston running behavior, which is the interaction between the piston rings, the cylinder liner surface, and the lubricant film between them. Get that interface right and wear collapses. Get it wrong and no amount of oil will save it. Improving it allowed the lubricating oil feed rate to be reduced at the same time as the overhaul interval was extended. Better wear behavior and less oil consumed from understanding one interface properly. That is engineering at its most satisfying, and it is invisible from outside the machine. There is a detail about scale that I think conveys this better than any of the weights. These engines cost in the region of 25 million dollars each, and the reason a ship owner pays that is straightforward. A single engine driving a single propeller is what they want, because it is more efficient and simpler to operate than two smaller plants. As container ships grew larger, the engine had to grow with them, so the engine grew until it reached this. The 14-cylinder version entered service in September 2006 aboard the Emma Maersk, which was the largest container ship in the world when it was built, and it is worth understanding what that engine is actually doing on a working ship, because the operating pattern is unlike anything on land. A container ship crossing an ocean runs at a steady output for days or weeks continuously. There is no stopping, no idling in traffic, no cold starts every morning. The engine is brought up to load and then held there. That is the ideal condition for a machine like this, and it explains a great deal about the design. Nothing about the RTA-96C is optimized for flexibility, quick response, or frequent starting, because it does not need to be. It is optimized for running at one output efficiently for an extremely long time without being touched. An engine that spends its life at a steady state can be tuned far more aggressively for efficiency than one that has to be good across a range, which is a large part of why this is described as not only the most powerful, but the most efficient prime mover in the world. Now the thing I keep coming back to about this engine, almost every remarkable feature of it exists because of the size, and would be pointless at any smaller scale. The crosshead exists because a five and a half ton piston generates side thrust that would destroy a cylinder liner. In a car, the side thrust is small enough to live with, so nobody bothers with a crosshead and the extra height and complexity it costs. The split lubrication exists because the crosshead made it possible, and because burning high sulfur fuel makes it necessary. Neither of those applies to a car, and neither would justify the plumbing. The three-year overhaul interval exists because taking this engine apart is a three-month operation with 25 people, so anything that extends the interval is worth an enormous amount of engineering effort. On a car, an oil change is a 20-minute job, and nobody spends a research budget avoiding it. And the low speed exists because a propeller works most efficiently turning slowly, so the engine is designed around a hundred and two revolutions per minute, rather than running fast and fitting a reduction gearbox that would itself weigh hundreds of tons. Every one of those is a solution that only makes sense at this size. Scale each of them down, and it becomes an unnecessary complication that a smaller engine would be better off without. That is what makes very large machinery genuinely different, rather than simply bigger, and it is the thing photographs never convey. The engineering does not scale up. It changes character entirely. Forces that were negligible become dominant. Wear rates that took years now take weeks. Components that could be lifted by hand now need rail systems. And solutions that were unthinkable on a small engine, like adding an entire extra mechanism between the rod and the piston, become obviously correct. A machine ten times the size of another is not the same machine ten times larger. It is a different machine that happens to work on the same principle. A five and a half ton piston, six meters tall, a 300 ton crankshaft, a cylinder over three feet across and eight feet deep, wide enough to stand inside, displacing 1,820 liters on its own, a connecting rod that never touches the piston, and a piston rod that never leans so that the liner is never used as a bearing, and the crankcase can be sealed off entirely from the combustion above it. Two separate oils doing two separate jobs, one of them consumed as it neutralizes the acid produced by its own fuel. No camshaft anywhere, with the valves and injection controlled electronically instead. 160 grams of fuel per cylinder per cycle and 250 tons a day. 7,780 horsepower from each cylinder and 107,000 from all 14. Three months to install the cylinders and three years before the piston rings need renewing. If you enjoyed this, subscribe, because there is a great deal more engineering worth telling properly. And tell me in the comments which component of this engine you would most want to see in person, because the photographs genuinely do not convey the size of any of them. Thanks for watching and I will see you in the next one.