Why do steam and diesel engines have differing cylinder configurations?

Consider a fixed steam engine, such as in a Lancashire cotton mill or pumping in a Cornish mine. Then consider a fixed diesel engine such as a large generator or ship's engine.

Why does the steam engine have one or two large cylinders with long stokes, but the diesel has many cylinders with relatively short strokes?

• I think it has mostly to do with power density. Mar 28 '16 at 13:31
• Perhaps in brief: So long as you keep pumping steam in from a high-pressure reservoir, you can keep pushing the cylinder head. But the expansion pressure from igniting a bit of fuel has no "reservoir" of high-pressure gas, so a long cylinder throw will gain no power for you. Mar 28 '16 at 19:40

Steam engines can expand the steam only so much before the water vapor starts to condense. In addition, the volume of the vapor increases significantly as it expands, making it difficult to obtain further work, as the cylinder size must increase to accommodate the volume of vapor passing through it.

Most steam engines use three or four cylinders to expand the steam, and are referred to as "triple expansion" engines. The steam enters the smallest, or high pressure cylinder, first out of the boiler. Then, it leaves the HP cylinder and enters an intermediate pressure cylinder next, which is physically larger than the HP cylinder. Finally, the steam enters the low pressure cylinder, which may be split into two cylinders because of the large volume of steam which must be handled.

On the other hand, a diesel engine does not expand the combustion gases in multiple stages like a steam engine. The fuel is injected and burned all in one cylinder. Large diesel engines can therefore have their power output increased by adding additional cylinders. These engines are usually designed in a modular fashion to make it relatively easy to increase the number of cylinders during construction.

The maximum efficiency of either the Diesel or a steam engine cycle is (assuming you can treat steam as an ideal gas): $$\eta_{Diesel} = 1 - \frac{1}{(\frac{V_{Max}}{V_{Min}})^{k-1}} \left[\frac{r_c^k-1}{k(r_c-1)}\right]$$ Where $k$ is the heat capacity ratio, and $r_c$ is the cutoff volume ratio - i.e. the ratio of volumes after and before the combustion process.

A few things to note in this formula - Clearly it would seem that having a long cylinder stroke would be advantageous either way. However, this would also require a very large engine, as the piston rod has to be longer. Going to the second part, for a standard diesel combustion, the volume may expand to double at the most (i.e. $r_c = 2$,making the bracket = 1.17) between the onset of combustion and the end of combustion. For steam, this isn't really the case. As can be seen from this handy diagram:

The cutoff ratio approaches 20 before constant pressure injection is finished. This makes the bracket about 2.0, really impeding efficiency. However, if you've got the space for a boiler, condenser, reservoir and pump, you've got space for a really long piston rod to bump up the bottom term and bring the efficiency back to a reasonable number.

However, what does this mean in terms of raw power? $$\mathscr P = \eta*Q_{in} = \eta * \dot m C_v (T_{expand} - T_{comp})$$ Where $\dot m$ is the mass flow rate, $C_V$ is a constant based upon the gas used, and the $T$'s are the temperatures at the end of compression and the beginning of expansion, respectively.

Once we have a similar efficiency, now the hard part is temperature. A steam engine can't reach 1000K like a diesel can. To makeup for the lack of power, it needs a lot of mass flow - big cylinders and double action really help. For the diesel, simply adding more of the already compact, high power, and fairly efficient engines will be easier than making a single big cylinder and dealing with the difficulties (such as the engine losing isoentropic compression / expansion), and removing that excessive heat from many distributed points (rather than a single massive point) makes a lot more sense due to the square cube law.

So, to summarize:

1. Higher cutoff ratio means steam engines need longer piston rods to maintain a higher efficency
2. Since the temperature is higher, the diesel pistons need to be smaller diameter to allow for proper cooling to avoid damaging the materials.
3. Since each small diesel piston is able to perform efficiently and powerfully, it makes more sense to add a bunch of them then dealing with a single large chamber with inefficiencies.

The difference of few large cylinders or many small cylinders has less to do with running on steam or diesel, but more with the era in which the engines were built. Early engines (both steam and diesel) had few, large cylinders, probably because that is how engineers tried to make more powerful engines: make everything bigger. The realization that multiple smaller cylinders gives a smoother running engine and more power output (because small pistons can run at higher rpm) came later, and by that time steam cylinder engines were being replaced with either diesels or steam turbines. But smaller cylinder steam motors also existed. And if you search for videos of people running antique diesel engines, those are also often single or few cylindered.

Length of the stroke is not really relevant, what matters is the ratio between the minimum and maximum volume, and you can increase that by having a long stroke with a large maximum volume, or a very small minimum volume. Modern engines use short strokes because those allow for higher rpms.

addition: Another factor is that a single cylinder engine is easier to build than a multi cylinder engine. In the 19th century the technology for building engines (both steam or internal combustion) was less advanced, and the simpler designs were preferred. As tools and construction methods improved, and as more uses for light high speed engines became apparent, engine designs moved from single large cylinders to multiple smaller cylinders. This too has little to do with the engine type, but much more with the era in which the engine was built.

Early steam engines weren't actually steam engines at all but were rather atmospheric engines and the steam was a convenient way of generating a partial vacuum in the cylinder by condensation. This is typically the case for the really huge Cornish beam engines. As they were driven by less than one atmosphere of pressure differential they needed very large cylinders to do a useful amount of work. Also they did work on what would conventionally be called the compression stroke which is why they use an overhead rocker beam rather than a connecting rod/crank arrangement.

Once you get into the realms of high pressure steam engines the difference in scale compared to IC engines is less pronounced.

Another aspect is that steam engines are typically driven by a fairly large reservoir of pressurised steam (ie a boiler). By contrast the hot working fluid for an IC engine is, by definition, generated by discrete combustion within the cylinder itself so there is a limited amount of energy available to do work per cycle. By contrast with steam (or compressed air) you can continue to do work on a piston for an indeterminate length of cylinder, limited only by the volume of your pressurised gas reservoir.

In practice there are limits as boring very long cylinders starts to get technically difficult and expensive.

Having said that there is a crossover in scale between steam and IC engines indeed it is possible to convert a IC engine to run on steam. Air cooled early VW engines have been converted in this way.

https://youtu.be/LJq2Hc_mXFI

• Power transmission and gearing tech are concomitant reasons. Walking beam engines had massive inertia and drove loads with massive inertia. As soon as crank and flywheel designs became feasible, they took over. As you point out, steam and ICE aren't that different for crank engines. A triple expansion steam engine with a double acting HP cylinder, a double acting MP cylinder, and a pair of single acting LP cylinders looks a lot like a 12 cylinder four stroke ICE from the flywheel's point of view. The real issue is cylinder speed and heat loss associated with slow speeds. Aug 16 '19 at 1:54