RESEARCH OF A BETTER CONVECTIVE MODEL FOR THE THERMAL SIMULATION OF BUILDINGS.

THE EXAMPLES PRESENTED ARE EXTRACTED FROM EXPERIMENTAL STUDIES ON DIFFERENT HEATING SYSTEMS IN A CLIMATIC ROOM. A FIRST ATTEMPT CONSISTS OF ESTABLISHING A CONTINUOUS MONO-DIMENSIONAL MODEL OF THE MAIN INTERCONNECTED JETS, BOUNDARY LAYERS AND CORE FLOWS. A SECOND ATTEMPT IS A DISCONTINUOUS MODEL, ESTABLISHED BY IDENTIFICATION OF THE MAIN AIR STREAMS IN THE ROOM AND DEFINITION OF A LIMITED NUMBER OF INTERCONNECTED < PERFECT MIXING > ZONES. A TWO-ZONE MODEL COULD BE SIGNIFICANTLY RICHER THAN THE CLASSICAL METHODS WHICH CONSIDER ONLY ONE AIR TEMPERATURE AS A REFERENCE. EVEN IF THE NETWORK HAS TO BE REDUCED, IT IS STILL POSSIBLE TO MAKE IT MORE REALISTIC BY INJECTING SOME OF THE CONVECTIVE HEAT DIRECTLY ON CERTAIN SURFACES.