Inverse design of indoor environment using an adjointRNGk‐ε turbulence model

Abstract

The adjoint method can determine design variables of an indoor environment according to the optimal design objective, such as minimal predicted mean vote (PMV) for thermal comfort. The method calculates the gradient of the objective function over the design variables so that the objective function can be minimized along the fastest direction using an optimization algorithm. Since the objective function is controlled by the Reynolds‐averaged Navier‐Stokes (RANS) equations with the RNG k‐ε model during the optimization process, all the corresponding adjoint equations should be solved, rather than the “frozen turbulence” assumption used in previous studies. This investigation developed adjoint equations for the RNG k‐ε turbulence model and applied it to a two‐dimensional ventilated cavity and a three‐dimensional, two‐person office. Design processes with the adjoint RNG k‐ε turbulence model led to a near‐zero design function for the two cases, while those with the “frozen turbulence” assumption did not. This investigation has successfully used the new method to design a two‐person office with optimal thermal comfort level around the two occupants.

Inverse design of an indoor environment using a filter‐based topology method with experimental verification

“…Since the objective function implicitly contains the design variables, we cannot directly calculate the gradient of the objective function over the design variables. Therefore, we have used the CFD‐based adjoint method with the adjoint RNG k‐ε turbulence model to calculate the gradients. A previous study provides detailed information about the adjoint equations, gradient equations, and adjoint boundary conditions of the CFD‐based adjoint method with the adjoint RNG k‐ε turbulence model.…”

“…Therefore, we have used the CFD‐based adjoint method with the adjoint RNG k‐ε turbulence model to calculate the gradients. A previous study provides detailed information about the adjoint equations, gradient equations, and adjoint boundary conditions of the CFD‐based adjoint method with the adjoint RNG k‐ε turbulence model. For identifying the number, size, location, and shape of air supply inlets and the air supply velocity, temperature, and angle, the area‐constrained topology method was the best approach.…”

“…If the objective function did not satisfy the convergence criterion, the method calculated the derivative of the objective function J ( ξ ) over the design variables ξ , so that the method could find the direction in which to adjust the design variables and further minimize the objective function. Since the objective function was controlled by the RANS equations with the RNG k‐ε model during the optimization process, this study used the CFD‐based adjoint method with the adjoint RNG k‐ε turbulence model to calculate the gradients. Next, the gradient filter was used to smooth the gradients.…”

Microventilation system improves the ageing conditions in existent wine cellars

“…For example, Geyrhofer et al (2011) indicates normal values, ensured by conventional natural ventilation, of air velocity in the range of 0.3–0.4 m/s. Ventilation efficiency, however, could be insufficient to maintain constant and optimal conditions inside the cellar because it is strictly dependent on the external climatic conditions (Santolini et al 2019, Zhao and Chen 2019). Ventilation in a wine cellar should be able to ensure homogeneous conditions of temperature and RH, avoiding areas of stagnant air in order to limit wine losses and to avoid mould growth.…”

Thermal performance analysis of a naturally ventilated system using PMV models for different roof inclinations in composite climatic conditions

“…Another comparative study was carried out to assess which model is appropriate for the isothermal analysis of natural ventilation in a crossventilated building [47] between standard k-epsilon, RNG model and Shear-Stress-Transport (k-ω SST) turbulence models, which concluded that k-ω SST model provided a good agreement with experimental data for the isothermal analysis. Likewise, several other studies were also undertaken to compare which model gives the best results when the effects of buoyancy [48] and wind-driven forces are considered, which further furnished CFD model as the best tool for the studied cases [31,44,49,50].…”