Elsevier

Applied Thermal Engineering

Volume 37, May 2012, Pages 267-274
Applied Thermal Engineering

Evaluation of various CFD modelling strategies in predicting airflow and temperature in a naturally ventilated double skin façade

https://doi.org/10.1016/j.applthermaleng.2011.11.028Get rights and content

Abstract

Demands for energy savings, thermal and visual comfort and a high-tech image for new building envelopes can be met with a Double Skin Façade (DSF). These kinds of building envelopes are widely encouraged, proposed and increasingly designed by architects. Naturally ventilated DSFs seem very interesting from an energy point of view, but a good design is crucial to improve the energy savings and the proper operation of the system. Computational Fluid Dynamics (CFD) can play an important role in evaluating and improving the thermal behaviour of a DSF. This paper shows, through a sensitivity analysis, a good strategy for carrying out a CFD simulation of this special building envelope. In this work the validations of the results are based on experimental data from the literature.

The paper provides a discussion that highlights which factors are important in the simulation, and which increase model complexity without improving the prediction capacity. The results show that, for a DSF characterized by a prevalent bidirectional flow, the additional effort required to make a 3D model is not justified by a significant improvement of the results. This work shows also that the accuracy can be improved by modelling outdoor ambient. The performance of k-ε and k-ω, the two most commonly used turbulent models for simulating the naturally ventilated DSF is evaluated.

Highlights

► A CFD model for the natural ventilated double skin façade was developed. ► The model was validated using experimental data. ► The paper provides a discussion that highlights which factors are important to the simulation. ► The results of a 2D and 3D models were compared. ► The accuracy can be improved by modelling outdoor ambient.

Introduction

In the last years, new building envelope systems have been developed in order to improve thermal insulation, to shade solar radiation and to provide suitable thermal and visual comfort conditions. One of these special types of envelopes is “Double Skin Façade” (DSF). DSF are made with two layers of glass separated by a significant amount of air space. The space between the glasses can be ventilated with three different strategies: mechanical ventilation, natural ventilation or hybrid. The ventilation of the air gap contributes to saving energy both during the summer and the winter time. In fact, during the winter time, the air between the glass is heated by the sun rays (greenhouse effect [1]), thus improving the thermal performance of the façade with a consequent reduction of heating costs. With hybrid ventilation systems, during the winter, the fresh air can be pre-heated in the DSF gap before entering in the HVAC system. During the summer, the air flow through the DSF (mechanical or natural) can help to decrease the temperature in the gap.

A blind for solar control is usually installed in the DSF gap. In addiction to reducing heat gain during the summer, this blind increases airflow through the gap with a strong buoyancy effect. In mild seasons, stack effect occurring in the intermediate space can be used as driving force to promote natural ventilation of the whole building [2].

The correct behaviour of a DSF is the key to increasing energy savings, but correct behaviour requires the structure to be designed correctly. One of the weakest spots of this kind of envelope is the design, especially for naturally ventilated façades, where the thermal process and the airflow mechanism influence each other. The magnitude and extent of this interaction depend on the geometric features of system, and the thermal and optical properties of various components.

Ventilated facades are already a common feature of architectural competitions in Europe; but there are still relatively few buildings in which they have actually been realized, and there is still too little experience of their behaviour in operation [1], [3], [4]. For this reason the CFD analysis could be one of the most important tools to predict the behaviour of DSF and help architects make decisions during the design process.

In the literature there are several examples of using CFD to study the behaviour, features and energy consumption of a DSF [5], [6], [7]. The advances in computing power and commercial CFD software available to building mechanical engineers make it possible to use this tool [8]. Using CFD does not necessarily ensure accurate results [9] and it requires engineering judgment [8], [10]. Thus the steps of validation, verification, and reporting results described by Chen and Srebric [8] are of great importance.

This research discusses the primary parameters that can influence CFD results during a modelization of natural ventilated DSF. This was carried out through an accurate sensitivity analysis.

The model was compared by using Mei’ measurements [11]. These measures were used for two reasons:

  • they were carried out in a laboratory, so they were not influenced by wind. The instability of wind can strongly influence the DSF behaviour and make the comparison between CFD and experimental results problematic;

  • the velocity and temperature fields inside the gap are presented in the Mei’ paper. These can be compared with the CFD velocity and thermal fields to better understand the impact of a different user’s choices;

The key points of a sensitive analysis are:

  • air property definitions, as constants or as a function of temperature;

  • turbulence model;

  • presence or absence of external environment;

  • 2D or 3D model.

The scope of this work is to show the effects that principal simulation parameters have on CFD results. The scope was not to validate the model. This is because of the other actions, like modifying the CFD model dimensions [5], were not performed in order to improve the results agreement.

The model was realized with the commercial software Fluent [15]. Fluent was used in other similar works [10] [6].

Section snippets

Case description

In this work a typical single-story commercial façade was modelized. The main dimensions of a double skin were drawn from an article by Mei [11]. The CFD model was realized with the following dimensions: the outer skin of the façade is a single 12 mm thick clear glass pane, which is 144 cm wide and 206 cm high comprising an aluminium frame. The glass area is 128 cm wide and 191 cm high. In the Mei’ case study both the air intake and exhaust of the DSF are designed as a commercial grille

Boundary conditions and numerical methods

The inside air temperature (temperature inside the room) was the same as outside, 293.15 K. The air ingress and egress are modelled as a pressure inlet and pressure outlet with the same gauge total pressure equal to 0. In Fig. 1a simplified section of the double skin is shown.

The solar radiation was not directly simulated but the surface temperatures measured by Mai et al. were used as boundary conditions. This was advantageous considering the next step of this work, coupling of CFD program

Sensitivity analysis

An accurate sensitivity analysis was carried out for every parameter that could influence the quality of the results. In this section the different parameters are discussed separately.

Results

The predicted results were compared with the experimental data. The model accuracy criteria used to evaluate the difference between experimental data and CFD results are quite similar to that used by Zhang et al. [16]. This model quantifies the relative error between prediction and measured points. If this error is less than 10% the rating is A, between 10% and 30% the rating is B, from 30% to 50% is C, while greater than 50% the rating is D. The scale was improved in order to evaluate

Discussion

The velocity field prediction is slightly improved by the addition of external environment in the case using the RNG turbulent model. In the case using the SST model, adding the external environment worsens the prediction for two points and improves it only for one. The thermal field prediction is noticeably improved in both cases. For both cases the presence of external environment improves the velocity prediction in the DSF part between the venetian blind and the inner glass. More than 60% of

Conclusions

A CFD model for the natural ventilated double skin façade was developed. The model can be used to predict the airflow patterns, air temperature and air velocity distributions, and heat flux from gap into the room. The model was validated using experimental data collected in a full-scale double skin module test facility by Mai et al. The computed air temperature, and velocity generally agree well with the measured data. Furthermore the simulation results show some interesting aspects:

  • in the

References (20)

There are more references available in the full text version of this article.

Cited by (119)

View all citing articles on Scopus
View full text