Journal of Building Physics

The discussion of the computer communications revolution to date has tended to dwell more on the methods than the content. The technology makes trans mission easy, but it does not relieve people of the challenge of expressing themselves comprehensibly. Let's not let the machinery get in the way of com municating as well as we can ...

In a previous paper by the same author [1], a simplified analysis of the heat and moisture response of protected membrane roofs (PMR) was discussed. The analysis included a ΔU-correction to the thermal transmission coefficient U of the equivalent, conventional roof to account for the effects of percolation, evapora tion, and butt joint losses between the insulation boards. This second paper presents experimental results.

The hygrothermal performances of five Protected Membrane Roof systems (PMR) were monitored to a varying degree of completeness for a number of years in the period between 1981-1992. The systems included extruded polystyrene boards, mineral fibre boards, a composite foam-concrete board manufactured with expanded polystyrene, and a composite foam-concrete board manufactured with extruded polystyrene.

The monitoring indicated that the heat and moisture response of a protected mem brane roof is substantially reduced when ponding water is present either below or above the insulation. The reduction increases when the vapour diffusion resistance of the insulation material is low. Both conclusions supported the ΔU-correction, derived from the simple model. They have prompted changes in the design of PMRs and modifications to the corresponding design rules.

Cell size and shape and closed or open configurations were found to greatly influence the compressive strength, density, water absorption, and insulative properties of two sintered fly ash/acid materials. In one case, finely ground fly ash, mixed with a water/phosphoric acid solution and alumina, is sintered to produce an open microstructure consisting of a large number of small pores dispersed relatively uniformly throughout the material. The material has a density of 0.8 gm/cm³ (800 kg/m³), very low bending strength, a compressive strength of under 800 psi (5000 KPa), and thermal conductivity of .17 W/mK. In the second case, the material had a lower density of 0.52 gm/cm³ (520 kg/m³) a higher compressive strength of 800-1000 psi (5000-7000 KPa), and thermal conductivity of .11 W/mK. This con sisted of larger closed spherical cells mixed within the other material. In a third case, theoretical studies lead us to predict values of density (696 kg/m³), improved com pressive strength (6500-9100 KPa), thermal conductivity (.099 W/mK), and less water absorption for a material consisting of closed cell/close packing structure ob tained by a different sintering process, one using a vacuum to expand the spheres to close pack them. These materials are for use as exterior applied insulation on build ings.

The effect of the initial moisture distribution on the transient heat transfer and moisture redistribution processes in fiberglass insulations is studied experimentally and numerically. Moisture evaporation starts from the hot side of the insulation and the drying-front sweeps across the slab. Condensation occurs in the wet side of the slab and also at the cold plate. The heat flux decreases progressively as the drying-front moves through the slab. The numerical model predicts the trends for the heat flux and temperature well. However, the measured transient variations occur at a slower rate than predicted by the model.

This paper examines thermal and hygric aspects of the performance of ventilated air spaces in both walls and sloped roofs. A ventilated air space between roofing tiles and the insulated roof deck is required by most European construction codes and standards. The upper side of the roof deck is covered with a membrane (called here, a roof deck membrane). Since this membrane performs the same func tions as the roofing felt or the asphalt impregnated paper used behind the air gap in the frame wall systems of North American houses, one may call it a weather barrier. Results from experimental tests on different roof-deck membranes available on the French market revealed that migration of water vapor in the roof is mostly caused by convection through small slits in the structure. The main functional contribution of the ventilated air space is, therefore, to reduce the amount of condensed moisture that remains in the structure, whether it is a wall or a sloped roof.

It is important to underline that similarly to a modern wall construction, the roof ing deck must also be constructed as the airtight structure. Both through the numeri cal modelling presented in this paper, as well as through the results of in-situ mea surements reported in the literature, it is shown that for an airtight roof deck structure and air velocities observed in ventilated cavities, the effect of air ingress into the insulation is insignificant. This study also shows that ventilation below the roof deck membrane is unnecessary when vapor permeable or capillary roof deck mem branes are used.