SFC is the mass of fuel that is used per hour to produce each unit of thrust (for a jet engine) ot power (for a piston/prop or turboprop engine). So if SFC increases (at constant thrust or constant power output), the fuel flow will increase and the endurance will decrease.
SFC is affected by a number of factors including the following:
ENGINE RPM
...
AIR PRESSURE AND DENSITY
...
AIR TEMPERATURE
...
What about speed?!
The "jet approximation" has always troubled me. The idea of using a measure such as SFC is that the ratio should be relatively unaffected by other parameters, allowing us to see the strong relationship between numerator (power or thrust) and denominator (fuel flow).
For a prop, the constant PSFC approximation is pretty good. You'd expect it to be, wouldn't you? A given mass of fuel holds a well determined chemical energy. Divide by time and it's pretty intuitive that burning a given mass per unit time transforms the chemical energy into work at well determined rate, i.e. constant power. The power
should, to a good first approximation, be proportional to the rate at which we burn fuel. And it is, with a bit of an exception at speeds close to zero where the engine is doing no useful work but still needs fuel to turn. Then we look for inefficiencies, and look at the dependence of those on other parameters, like propeller efficiency and pumping losses with their dependence on throttle setting.
But what about jets? The fundamental physics doesn't change. We're still turning hydrocarbon into energy, and to a good first approximation, the
power, not the thrust, should be proportional to the rate at which we burn fuel. Why should we expect to get anything like constant
thrust from burning fuel at a constant rate? Of course there may be greater efficiency at high speed, but would we really expect a ton of kerosene to produce twice the power simply because we're flying twice as fast?
And if you look at a cutting-edge turbofan like the RB211 (oh well, it's only 40 years old), thrust specific fuel consumption is indeed vastly higher at higher speeds. That doesn't mean that power specific fuel consumption is constant either, it's substantially lower at higher speeds. But we're nowhere near the "ideal jet approximation" of constant TFSC.
Does it matter? Isn't TSFC just a useful ratio to measure with no real presumption that it's independent of speed? Well perhaps, but when we start suggesting e.g. that best endurance on a jet is at Vmd "because TSFC is constant", aren't we in danger of getting the real best-endurance speed quite a long way out?
(I'm not suggesting that Keith is saying anything incorrect or misleading in his post, just that the ideal jet approximation is a somewhat odd place to start when considering aircraft performance in 2010.)