This article presents a transient model for a two-bed, silica gel-water ads
orption chiller. Compared with our previous cyclic-steady-state model, we f
ound very good agreement between our model prediction and experimental data
. We found that, regardless of the initial mass distribution, the chiller w
as able to achieve the same cyclic-steady-state within four cycles or 1800
a. We also demonstrated that the manufacturer had empirically evolved to se
lect a fixed switching and cycle time such that the best peak evaporator te
mperature suppression, and near maximum cooling capacity can be achieved by
a compact design. (C) 1999 Elsevier Science Ltd and IIR. All rights reserv
ed.