Abstract
This study investigates the accuracy of the frame and divider object in Energy Plus for predicting energy losses and gains. Additionally, it explores the use of a standard surface object to model the window frame. The two simulation approaches are compared against experimental data collected at the Façade System Interaction Lab of Eurac Research. Key performance indicators (KPIs) such as internal surface temperature, external surface temperature, and heat flux of the frame in a closed cavity façade (CCF) system were measured. The frame’s conductance, calculated from the experimental data, was used as input for both modelling approaches. Subsequently, models’ calibration was performed to replicate the real behaviour of the frame accurately. The analysis was done for both nighttime and daytime trends for the KPIs. The results of the Energy Plus simulations using the frame and divider object did not align well with the measured data, showing significant divergence in trendlines and notable underestimation. In contrast, simulations that modelled the frame as a surface displayed more aligned trendlines, with peak values closely matching the measured data, thus validating this approach for the given window system. The error rate for the frame and divider object simulation was 64% (underestimation), while the frame as equivalent surface approach had a lower error rate of 23%. Additionally, modelling the frame as a surface enables more refined calibration due to the greater number of adjustable parameters available. These findings underscore the importance of selecting an appropriate window frame modelling approach in Energy Plus to achieve more accurate predictions.