Predictive models of air temperatures inside a naturally ventilated vehicle transporting weaner pigs

Abstract

Maintaining proper interior thermal condition during transportation is vital for animal welfare and sustainability of livestock supply chain. This study investigated the air temperatures inside a multi-deck naturally ventilated vehicle when transporting weaner pigs under warmer weather condition by using computational fluid dynamics (CFD). Predictive models of interior air temperatures were developed by using response surface methodology (RSM) and gradient boosting machine (GBM) with the inputs of exterior air temperature, vehicle speed, wind speed, incident wind angle and opening height of shutter based on the dataset generated from CFD simulations and validated as well. The results showed that predictive models developed by RSM were sufficient for predicting the interior air temperatures of moving naturally ventilated livestock vehicle, and GMB could improve the prediction accuracy moderately. RSM models indicated that the interior temperatures increased linearly with the increase in exterior air temperature, opening height and wind speed while insensitive to vehicle speed. GMB model indicated that the plane-average air temperature of front compartments was 2.2 °C higher than those of the other two compartments at the same deck, and the air temperature increased slightly from the bottom to the upper deck. High spatial variations in air temperature were observed inside the moving livestock vehicle, which poses a challenge on monitoring interior air temperatures. The developed models are expected to predict the interior air temperatures and provide suggestion on regulating ventilation systems in advance. Further study could be conducted to investigate the optimum control of opening for improving the natural ventilation potential.