9.2. Model and workflow

A representation of the entire workflow and model is presented in Figure 150. In the first step to creating the window’s energy model, all SIOPs relevant to the energy balance are generated. In the analysed case, the absorption-SIOPs for the three window panes, the SIOP for the solar direct transmission to the interior and the absorption SIOP of the blinds are required. The latter is used to model the blind’s surface temperature, which is considered in the longwave radiative exchange with the first glazing surface. As mentioned, all SIOPs can be generated within a single raytracing scan of the object (see section 7.10).

The fenestrations’ dynamic thermal simulation model is established in the next step. In order to have full control over the simulation parameters and avoid any third-party-related issues, the author has implemented a simple nodal dynamic thermal simulation platform that is used for this application.
It allows setting up models based on thermal nodes and heat transfers based on a library containing a variety of classes. The simulation platform features thermal mass and heat input modelling. The latter is, of course, essential to model the solar inputs to the specific nodes. The dynamic simulation platform has been validated against other validated tools based on simple reference cases.
The simulation model for the presented application case can consider the following boundary conditions: air temperature, sky and ground temperature (for longwave radiative heat exchange), wind speed (for convective heat exchange), and interior temperature. The model comprises seven variable thermal nodes. The first one represents the blind temperature, whereas the other six represent the internal and external surfaces of the three glass panes. While the solar inputs and view factors are based on detailed three-dimensional raytracing data, the energy model is set up in one dimension, i.e. one temperature is assigned to each layer. Due to the relatively high thermal conductivity of the glass panes, this approach is reasonable and commonly applied. However, the same approach can be used to create a spatially refined model, either by vertical stratification or two-dimensional segmentation.
All radiative, conductive and convective heat exchanges were modelled according to the specifications provided by the standards EN 673 (EN 673, 2011) and ISO 6946 (EN ISO 6946, 2015). In order to determine the heat transfers of the IGU accurately, all relevant parameters, i.e. surface temperatures, gas temperatures and gas properties, were calculated dynamically for each time step. This leads to a significant variation in the heat flows within the IGU and, consequently, in the thermal transmittance of the entire glazing. In order to be able to model transient effects, thermal inertia is considered for each node based on the component’s relevant mass and specific heat capacity. An overview of the simulation model with all relevant nodes and heat transfers is depicted at the bottom of Figure 150. The SIOPs are evaluated using the Perez irradiance model as specified in section 6.2.3. The generic Perez incident profiles are created with irradiance data provided by weather data time series. As outlined in section 4.5, the local solar position is calculated using the SG2 (Blanc et al., 2012) algorithm in the implemented module (section 4.5.2). Subsequently, the Perez irradiance components are determined based on the measured irradiance data (see section 6.2.5). Finally, the SIOPs can be evaluated to provide the power inputs to the specific nodes with high temporal resolution.
The derived solar power input time series and additional time-series data relevant to the air body, i.e. air temperature and wind speed, are used to update the model’s boundary conditions on each time step. Finally, the equation system is solved on every time step, and the resulting node temperatures are used to derive all relevant heat flows. The entire time series is processed twice, with all solar inputs set to zero in the first run. The resulting heat flows contained in this time series that excludes any solar inputs are further used as a reference to identify heat flows resulting from solar gains. This approach, therefore, follows the principles expressed in equation (2).