Why do steam engines produce maximum torque at 0 rpm, unlike internal combustion engines which have minimum torque at 0 rpm? Just as it says in the title , the title is self explanatory. Can anyone explain the reason for that ?
 A: The two types of engine run on very different principles, so it is not surprising that their torque curves are very different too - it would be more surprising if their torque curves were at all similar.
An unloaded internal combustion engine has a natural or "idling" RPM which is determined by the force which the ignition of the air/fuel mixture in the cylinder exerts on the piston. Although this can be varied somewhat by controlling ignition timing and the ratio of fuel to air, it is fairly obvious that running an IC engine well below or well above its idling RPM will be inefficient. Standard petrol engines in cars, for example, are designed to idle at between 600 and 1,000 RPM and to deliver maximum torque between 1,000 and 2,000 RPM. Lower or higher output RPM is produced using gears.
A simple steam engine has no compression or ignition, and its RPM is determined by how fast steam is introduced into the cylinder. With a large bore cylinder and slow steam intake, RPM can be as low as you like. The RPM of a steam locomotive, for example, is controlled by varying the amount of steam introduced on each stroke and the length of the piston cycle. This allows a steam locomotive to operate efficiently over a wide range of RPM values, and so it does not need gears.
A: In a steam engine, the source of the power is external- you can build up a huge head of steam and feed it at maximum pressure into the cylinder when the the engine is stationary.
With an internal combustion engine, you can't build up a source of power externally when the motor is not turning- you have to turn the motor to compress the fuel-air mix and generate the spark. When a cylinder fires, the instantaneous torque is actually very high- much higher than the mean torque, so you could argue that an engine does indeed produce huge torque with a single spark. The problem is that the torque is delivered as a series of pulses as the cylinders spark. The torque from a single cylinder is actually negative on the compression stroke (in the sense that it actually needs a torque applied to the crankshaft to compress the file mixture, rather than generating a torque). The instantaneous torque delivered by a car engine is a series of peaks and troughs, with the troughs dipping below zero in some engine designs. Car engines have flywheels that help smooth the delivery, but even so, with insufficient RPM there just isn't enough angular momentum built up, so it is possible for the engine to stall during the troughs in the instantaneous torque delivery.
I found a nice write-up of the effects here http://www.epi-eng.com/piston_engine_technology/torsional_excitation_from_piston_engines.htm
A: A steam engine is simply a specific case of what I'd call a "gas expansion engine" (don't know if that's a "standard" term or not).  It'll run just as well on compressed air, and that's actually been done for some recovery operations without "steaming up".  (let it help itself back onto some good rails and then tow it from there)
The principle is even simpler than for an internal combustion engine:

*

*You have a reservoir of (relatively) high-pressure gas of some kind.

*You connect that reservoir to one side of the cylinder.

*The difference in pressure pushes the piston.

*Done!

Of course, that only works for 1/2 cycle at most, so you need a set of valves to then connect the reservoir to the other side of the cylinder to push the piston back the other way, and simultaneously release the spent charge of gas.*
Then you add a second cylinder and its set of valves at 90deg to the first, so that it can provide power while the first is at dead center.  Some engines actually have three cylinders at 120deg: one visible on either side and one in the middle.  You can tell by counting the chuffs per revolution: 4 or 6.  That smooths out the power pulses, but the overall dead spots are already eliminated with just two.
Then you add a mechanism to adjust the valve timing while running, to provide both better efficiency at speed and the ability to reverse the motion.**

*That release/exhaust both makes the iconic sound (large low-speed ICE's sound like that too, by the way; it's just a volume of still-pressurized gas escaping, and thus a sign of inefficiency) and pulls additional air through the fire, making it burn hotter, and then through an array of small tubes (lots of surface area) to transfer that heat to the boiler.

**Just a spritz at high speed and let it expand from there, vs. filling the entire stroke at low speed because it could be anywhere in the cycle and still needs to work.  (can't rely on inertia yet)
A locomotive just starting out will have the valve gear at full stroke and the throttle just barely cracked open.  That controls the flow rate, not the pressure, so you'll eventually get maximum force at any throttle position as the cylinders "fill up", provided the wheels don't slip first.
In practice though, the cylinder cocks (low-point drains) provide a constricted escape path, giving some control of pressure via the throttle, and it's critical to have them open when starting to avoid hydrolock from condensed steam in a cold cylinder.  You might not be able to rely on inertia to pull the train yet, but it's already enough to blow a head off!
Then cruising down the line, it'll probably have the throttle wide-open and the valves for each cylinder almost at neutral, or barely moving at all.  (control it at speed using the valve gear, not the throttle)

