Keey and Wee (1985) designed an apparatus to test precisely the problem of interest in the present project. Such an apparatus was used in their later experiments (Wee, Keey & Cunningham 1989; §3.3.10).
The greatest difficulty is the provision of steady uniform vapour mass fractions on the hot and cold walls. Their idea was to construct these walls from sintered nonhygroscopic polymer (high density polythene) beads. The outsides of these walls would be exposed to air vigorously stirred in the presence of saturated salt solutions (the use of saturated salt solutions to control humidity is well known--see, for example, Leopold & Johnston 1927 or Stokes & Robinson 1949). Their preliminary tests showed that the mass transfer `resistance' of the porous walls was about five times that of the same thickness of stagnant air or about half that of the quiescent test cavity, although it varied with the imposed mass fraction difference. The humidity boundary conditions so imposed on the cavity would not be known with much accuracy. This turned out to be the case (see §3.3.10).
The apparatus of Keey and Wee is incapable of measuring the energy transfer associated with condensation and evaporation of the vapour, since the hot and cold walls allow the vapour to pass through without condensing. This is unfortunate, since this may often be the dominant energy transfer mechanism (McBain 1995; McBain, Harris, Close & Suehrcke 1998; Suehrcke & McBain 1998; §4.5).