Modeling of air inlets in CFD prediction of airflow in ventilated animal houses

This study investigates different methods to model wall inlets in computational fluid dynamics (CFD) simulations of airflow in livestock rooms. The experiments were carried out in an 8.5 in long, 3 in high and 10.14 in wide test room equipped with a forced ventilation system. Four wall inlets were distributed symmetrically along an end wall 0.5 in beneath the ceiling. To obtain uniform and easily modeled boundary conditions the inlets were designed as rectangular frames with an elliptic profile in the contraction section following the ISO standard for a long-radius nozzle. Vertical and horizontal air speed profiles in the jets were measured with thermistor speed sensors at four distances from the inlets and an ultrasonic sensor was used for measurement of air velocity in the occupied zone close to the floor. CFD-simulations with the k-epsilon turbulence model were carried out with a number of different grid constructions. Both measurement and CFD simulations showed that two different airflow patterns occurred in the test room. In airflow pattern in the jets beneath the ceiling turned towards the symmetry plane of the room and above the floor the air flowed away from the symmetry plane. In air flow pattern 2 the jets turned away from the symmetry plane and above the floor the air flowed towards the symmetry plane. The findings in this study indicate that assuming two dimensional (2-D) inlet conditions might be a useful way to simplify inlet boundary conditions and grid constructions for prediction of air flow in the occupied zone of livestock rooms with many wall inlets. However, more work must be done to evaluate this statement in other arrangements, including changed orientation and locations of inlets, unattached jets and non-isothermal conditions. (C) 2002 Elsevier Science B.V. All rights reserved.