The following is from Bob Tait's blog:
Impact ice is really just an example of normal airframe ice that has formed over the air intake to the induction system. It requires sub-zero temperatures and the presence of supercooled water droplets. The application of carburettor heat will do nothing to remedy impact ice. If the aircraft is not equipped with alternate induction system air, the only other option is to descend below the freezing level. Impact ice is not at all likely for a VFR aircraft in Australian latitudes.
Throttle ice refers to the additional adiabatic cooling caused when the throttle is partly or completely closed. For example, in a glide passing through 5000 feet, the pressure upstream of the throttle is about 25" while the pressure immediately downstream of the closed throttle is about 10". This sudden pressure drop is accompanied by a significant drop in temperature in accordance with the gas laws.
The evaporation of fuel in the vicinity of the throttle adds an additional cooling effect due to the absorption of latent heat during the process of evaporation. This is sometimes called refrigeration ice - not a very appropriate name actually.
By the way, I once had a complete engine failure just after lift off in a Cessna Cardinal. In this case the problem wasn't ice, it was a collapse of the scat hose that carried induction air to the carburettor. When I applied carburettor heat, power was restored simply because the carburettor heat was an alternate source of air to the carburettor. Just goes to show that the application of carburettor heat after a power loss is always a good option - even if the problem is not ice.
Bob
Note that many aero engines, like Continentals and Lycomings, do have alternate air and application may keep the engine running if there is impact icing only. But of course we will not be flying VFR in conditions conducive to impact icing, anyway.
Kaz