The problem with short commutes is that the engine cools down a lot before each journey, wasting a lot of fuel at the start of each journey. Even if you have a block heater, the return journey home will often consume a lot more fuel than the outgoing journey.
I have had an idea to improve this - as follows (assuming that your engine gets up to temperature at some point during the commute - it would happen after 3-4 miles, usually).
(1) In between the radiator and engine thermostat, insert an insulated water container, of about 10 litres capacity. Then insulate the pipe leading from the engine into this container (which will contain engine coolant). Once the thermostat opens, the hot water will flow out, at full temperature, into this container. If you design baffles inside this container, then the water will flow through the circuit fairly well. Once going through the box, it will then go to the radiator, cool down, and go back into the engine. After maybe 5 miles, you will have 10 litres engine coolant, in an insulated container.
(2) Have two pipes coming from this box, and inserted into a place in the cooling circuit, together with a small 12v pump in one of the pipes, and 12v solenoids in each pipe (such that, when there is no power to the solenoids, the solenoids are closed). They should be set up so that, when water flows out through one pipe, it will then go through the engine, before going back into the other pipe. Hint : the air heater circuit is a good place for this, if the heater is set to cold.
(3) To use the system - when you get back into your car, flick a switch in the car, which will open the two 12v solenoids, and begin circulating the water. The baffles in the 10 litre tank will mean that the cold water will enter 'at the back' while the hot water flows into the engine.
This should make the engine reach near operating-temperature, prior to starting up and driving off, in a few minutes.
The concerns would be heat shock - however, if a slow 12v pump is used, the heating would occur over e.g. 3 minutes, and, the thermal shock is going to be much less than with a cold engine, which warms up to full temperature in a few minutes of driving, and always has a huge temperature gradient between the coolant and the combustion chamber anyway.
Any thoughts on this? It is something I intend to make soon, particularly as it will be quite cheap to make, and could give a real benefit to fuel consumption
__________________