Improved convective heat transfer and air infiltration models for building thermal simulation
10 Intermediate-level'o computer codes are advocated as being the most appropriate for meeting the requirements of dynamic building thermal models. Such codes may be developed via the .4 computer-generalizationA Of analytical solutions and data correlations, which are then verified using higher-level ccoputational procedures and/or experimental measurements. Two intermediate-level ccniputer codes are described: one to model the convective heat exchange at the external facades of a building (WIND-CHT program), and the other to calculate the hourly mean rates of air infiltration into buildings (FLOW program). These codes take into account most of the key parameters such as wind speed and direction, the change in shape and height of the atmospheric boundary-layer over different terrains, the relative dimensions of the building, the indoor-outdoor temperature difference and the leakage characteristics of the building. Both the WIND-CHT and FLOW programs are carpared with field experimental data, and good agreement is shown. The sensitivity of two dynamic building thermal models to the external convection and air infiltration input data are then assessed. The NBSLD (National Bureau of Standards Load Determination) 'response factor' program (1981) and the BM (British Research Establishment) 'admittance procedure' program (1984) were chosen for this purpose. The sensitivity of these models to the internal convection input data was also assessed. In this case the ROOM-CHT program, developed by Alamdari and Hammond (1982) was employed. Both models displayed a considerable variation in their results when the 'traditional' input data were replaced by the 'improved' values, although the extend of the impact of the convection and infiltration models is likely to depend on the conditions prevailing in and around the particular building being simulated.