If I guess correctly it’s is based on the cooling effect of expanding gas (reversed of the compressor heat development).
I’ve done some back of the envelope calculations and the reason the water has to be removed from the air is exactly why this setup will only work with low loads. Water would adsorb too much of the ‘cold’ and therefore limit the temperature drop. But what we’ve left after drying is a gas with a very low heat adsorption capacity and so a very rapid heating when a load is applied. Also, the heat transport over the gas/solid interface will be problematic since the exchange will be severely limited by the not-optimal heat adsorption capacity of dry air.
But I might be wrong and I'm happy to help with some more detailed calculations if more info on the system is posted (like total air flow (in mol/sec or another mass & time related unit), pressure drop of the expanding air, temperature in pressure chamber and preferably CPU die temperatures under different loads). I'm sure some users in here can help in adapting an phase-change evaporator for optimum heat exchange for such a setup.
|