# DESIGN WITH MODELING TECHNIQUES INTRODUCTION

“How does an air curtain respond to a temperature change of, say, 10°?” “How does the air sheet react to an increased pressure difference across it?” Ventilation engineers want answers to such questions, and they want them immediately!

The most popular approach to solve these problems is to use experience and good engineering judgment. A quick experiment may be another solution. A computer simulation is a third option. All those approaches may eventually lead to success. This chapter presents various methods of computer simulation for industrial ventilation design.

Computer simulation involves three stages of simplification:

• Description of the physical situation by mathematical equations

• Transformation of these equations into a form that can be solved by a computer

• Numerical solution of the system of equations

To start the numerical solution, initial and boundary conditions must be specified.

The quality of the computed results is only as good as the quality of the boundary conditions supplied by the user.

Other key issues include the selection of the appropriate model and the abstraction of the physical problem for that model.

Presenting the results graphically completes the simulation. Before a simu­lation tool is handed to the design engineer, it must be thoroughly validated by computation of cases for which measurements have been performed and by comparison of the calculated results with the measured data. A basic rule:

The computer model is valid only for cases within the parameter range of

Successful verification.

Example: An airflow model that has been validated at air velocities up to 10 m/s (Mach number below 0.05) cannot be expected to perform reliably at a Mach number of 0.5.

Example: An airflow model that has been validated for temperature dif­ferences of 15 K cannot be expected to predict smoke movement accurately in a building that is on fire with temperature differences that are 10 times larger.

The engineer decides when to do a numerical simulation. He or she will consider numerical modeling when one or several of the following apply:

• Measured data for the particular case is unavailable (e. g., during design).

• The structure of the problem involves a large number of interacting zones connected by a complicated network.

• Different physical effects interact (e. g., buoyancy forces acting on an air curtain).

• Boundary conditions and thermophysical properties, etc. are clearly specified (e. g., indoor air quality can be predicted only if the location and strength of the sources are known).

• The influence of parameter variation is to be investigated (e. g., to optimize a local extract system).

Experiment or simulation? A decision may be based on the following con­siderations, but the final choice rests with the ventilation engineer:

Design or troubleshooting: During design, an on-site experiment with the actual equipment is usually not possible. Numerical prediction may then be easier. However, to investigate comfort complaints, field measurements are quicker, perhaps in combination with simulations.

Availability of facilities and expertise; Are test facilities, instrumentation, and data-acquisition systems available? Are simulation tools, hardware, and specialized staff for numerical simulation available? The decision depends on the answers to these questions.

Scaling laws: If a full-scale test is not possible, reduced-scale experiments are a good alternative. However, certain scaling laws must be observed (see Section 12.4, “Scale model experiments”). Correct scaling for isothermal flows is usually possible. However, scaling of buoyant flows in large rooms may be difficult, if not impossible. Then numerical simulation is the better choice.

Ventilation components with small geometric detail: Numerical modeling of diffusers with complex geometry is difficult. Therefore, it is more reliable to measure airflow around such devices at full scale.

Accident simulation: Large-scale accidents normally cannot be staged for experimental purposes. It is impossible to put a building on fire in order to see how smoke spreads. In this case, numerical simulation is the sensible approach.

Parameter variation and optimization: If parameter sensitivity is investi­gated, simulation is cheaper and quicker than experiments. Trends are usually well predicted by computer models.

Time available: It takes some time to set up a simulation case. After a ba­sic configuration has run successfully, it is easy and quick to do additional computer simulations with different boundary conditions. Experiments, on the other hand, are generally time-consuming.

Cost: If simulation software is in place and trained staff available, compu­tation is considered faster and less expensive than measurement.

Ideally, experiments and simulations should be performed simultaneously, and the results obtained should complement each other. The computer model for a basic configuration can be verified by comparison with experiment and then applied to more complex configurations.

Simple design methods may be found in Chapters 7 and 8 of the Fundamen­tals volume. Methods reported in this chapter are generally powerful and yield detailed information. However, there is a trade-off between simple and complex methods. This is illustrated in Fig. 11.1. More information provided is at the ex­pense of a higher cost and the requirements for more time and skilled staff.

Four methods for industrial air technology design are presented in this chapter: computational fluid dynamics (CFD), thermal building-dynamics sim­ulation, multizone airflow models, and integrated airflow and thermal model­ing. In addition to the basic physics of the problem, the methods, purpose, recommended applications, limitations, cost and effort, and examples are pro vided.

Depending on the purpose of the computer calculations, different tools are selected. The modeling methods described in this chapter and their pri­mary application are listed in Table 11.1.

Inefficient

Effort to obtain results (cost of simulation)

| FIGURE 11.1 Trade-off between the complexity of the method and information content of results. Simple methods are at the lower left; complex methods, such as computational fluid dynamics, are near the upper right corner of the graph.

 1 hermal building-dynaintcs simula­tion (Section 11.3 )

Computational fluid dynamics fCFD) (Section 11.2)

Multizone airflow models (Section 11.4)

Integrated airflow and thermal modeling (Section 11.5)

Airflow in space, air quality at all points in space, local age of air, entrainment of jets, heating and cooling by airstreams (heat transfer), buoyancy of warm and cold jets, thermal comfort at arbitrary points

Long-term (e. g., I 2 months) thermal response of a building, simulation of thermal storage of building mass and its effect on comfort, energy, and heating and cooling loads. In addition, plant and controls modeling is used to estimate annual energy consumption

Air infiltration and exfiltration, air exchange between rooms, mean indoor air quality in a room natural ventilation

Air temperature in rooms imposed by building dynamics and ventilation, nighttime cooling, cooling by natural ventilation

In an actual design, thermal modeling (Section II.3) for different seasons will come first to set temperature boundary conditions. Multizone airflow simulation (Section 11.4) will follow’ to define ventilation needs in each zone. For large en­closed spaces, for natural ventilation, and for a variety of other special problems, CFD (Section 11.2) and integrated modeling (Section 11.5) are applied.