The Application of CFD to IUS Buildings Analysis

In this study, wind characteristics and energy potential of IUS Campus region is investigated. Assessment of the site is carried out by using turbulent flow models. Commercial software STAR CCM+ was used to solve turbulent flow equations. STAR CCM+ is a computational fluid dynamics (CFD) software package that solves Navier-Stokes equations that define fluid flow, by using numerical techniques.


Introduction
Computational Fluid Dynamics (CFD) can play an important role in building design.For all aspects and stages of building design, CFD can be used to provide more accurate and rapid predictions of building performance with regard to air flow, pressure, temperature, and similar parameters.
Computational Fluid Dynamics (CFD) can play an important role in building design.For all aspects and stages of building design, CFD can be used to provide more accurate and rapid predictions of building performance with regard to air flow, pressure, temperature, and similar parameters.
Computational fluid dynamics (CFD) is used in science andengineering for numerical simulation of steady or unsteady flowfields.At Fraunhofer IBP we provide CFD competences in terms ofmodelling, simulation and analysis of complexphenomena relatedto indoor environmental quality.CFD enables in depth analysisof thermodynamic systems in all relevant application domainslike buildings and construction, as well as theautomotive andaviation sector.The CFD simulations consider heat transfer modesof convection, radiation and conduction incombination with multi-physics approaches.The CFD results are used to evaluate thermalcomfort of humans, quality of trading goods, ventilation efficiencyin buildings or thermal management strategies in vehicles.
Also the purpose of the case study is to compute flow around A and B buildings of IUS campus using CDF.The main focus of the case study is how heat is transferred by the forced convection which is natural wind.The computation is used to compute a heat transfer coefficient at the buildings wall for the winter season.For that purpose, velocity field, heat transfer coefficient distribution along the buildings, pressure distribution, wall y+, turbulent kinetic energy, and temperature distribution will be shown in scalar, vector and graph forms.The temperature of outside the buildings is declared as -2℃, and the wind flows at 21℃ with the velocity 5 m/s.STAR-CCM+ includes the latest numerical algorithms such as advanced solvers,raspredelennny (segregated solver) and connected (coupled solver), moreover, it is created using the most advanced programming techniques.All this allows STAR-CCM+ provideunprecedented accuracy, reliability and flexibility in solving problems of computational continuum mechanics.Powerful new tools for building grids: from restoring the integrity of the surface (surface wrapping) to create grids of polyhedral cells.Using these new tools for building grids can reduce the clock time for solving the problem.A large set of physical models: laminar and turbulent flow of Newtonian and non-Newtonian fluids, multiphase flow, cavitation, radiation, combustion, development of boundary layer flows with high Mach numbers, the conjugate heat transfer, as well as new models for the calculation of heat exchangers and fans.

Theoretical Background
The flow module determines the velocity and pressure fields by solving the three-dimensional momentum equations and the pressure correction equations, respectively.These equations are guided by the laws of conservation of mass and momentum, which lead to the use of the Navier-Stokes equations to iteratively resolve the flow solutions.The following sections describe the governing flow equations [2].
Applying the mass, momentum and energy conservation, the continuity equation, momentum equation and energy equation can bederived as follows.

General Form of Navier-Stokes Equation
To simplify the Navier-Stokes equations, they can be rewritten as the general form.
When Φ=1, Uj, T, and they can be respectively get as continuity equation, momentum equation and energy equation.

Finite Volume Method
The Navier-Stokes equations are analytical equations.Human can understand and solve them, but if we want to solve them by computer, we have to transfer them into discretized form.This process is discretization.The typical discretization methods are finite difference, finite element and finite volume methods.Here we introduce finite volume method, because finite volume methods are generally used in CFD [3].

The Approach of Finite Volume Method
Integrate

Computational Details
A 3-D finite-volume approach is adopted by the software STAR CCM+because of its capability of conserving solution quantities.The software solves the conservation equations for continuity, momentum, and energy as well as the equations for turbulent kinetic energy and its dissipation rate.The two-equation k -ε turbulence model is used.
A three-dimensional model of the IUS buildings in the shape of a rectangular prism was developed.The physical dimensions were set to be 48 m wide by 51 long by 20 m tall.
RANS solver is used to solve the governing equations of continuity and momentum in a viscous flow around the buildings model.The Navier-Stokes Equations (NSE) are solved with a segregated, algebraic multigrid solver using Gauss-Seidel iterations.
In this study case, for the three-dimensional case, implicit unsteady simulations were carried out at step intervals of 0.01 seconds for 200 time steps which is equivalent to 2seconds.A description of the conditions and assumptions is given in the following sections.

Computationaldomain
Generally, the size of the entire computational domain depends on the targeted area and theboundary conditions.For a single building, the distance from the top of the building to the top ofthe computational domain should be about20H, where H is thebuilding height and is defined as 20m.For the laterals boundary, 20H is required between the buildings and the computational domain.In the figure 2, the computational domain and dimensions are shown, and dimensions' units are in cm.

Computational Grids
The hexahedra cell has been chosen for meshing purpose.Also prism layer mesher and surface remesher options were used.All computational domains of this study were undertaken for generating mesh by means of relative cell size.The base size value is defined as 5m.Boundary growth rate is selected as very slow, and default growth rate is as medium.Also the mesh values in the buildings are selected relative size is 1m.Thus, relative maximum cell size of 5m and relative minimum size of 1m were used as limits for generating local mesh of all domains and the buildings.Meshing of this computational domain was carried out with 1,243,390 cells and 3737529 faces, covering the whole volume of the domain.

Turbulence Models
The CFD program was set to use a standard k-ε model for turbulence.The required input variables were the turbulent kinetic energy, k, and the turbulent dissipation rate, ε.The inlet flow velocity was taken to be 5 m/s.This inlet flow velocity can be directly inserted into equations in section 2.5, to result in a turbulent kinetic energy value, k, of 0.25 J/kg.Also, the turbulent dissipation rate, ε, can be calculated using the equation in the section 2.5, and the result is calculated as, ε, of 17 m 2 /s 3 .

Initial Conditions
Initial conditions describe the state of the system at the initial instant of time through the values of a solution variable, and have to be defined in the whole domain: f0 =f (t0).If the equations to be solved include high-order time derivatives, the initial condition include also the specification of the correspondingly lower derivatives of solution variables at the initial instant of time.The initial parameters are set as calculated before, shown in table 1.

BoundaryConditions
A boundary condition describes the state of a system, expressed through conditions on the solution variables holding on all boundaries of the system.

Figure 4: Boundaries of all faces
In figure 4, each surface is described as boundary conditions, and their types of boundaries are defined.Also the physical values of boundaries are given in table 2. Heat transfer coefficient and thermal resistance were as constant typical values [5].Asymmetry plane boundary represents an imaginary plane ofsymmetry in the simulation.The solution that is obtained with asymmetry plane boundary is identical to the solution that would be obtained by mirroring the mesh about thesymmetry plane.

Results
The results are shown in figures below.In Figure 5, the velocity magnitude of streamlines around the building for turbulence model is presented.In figure 6, velocity is shown as vector of a created plane from the middle of the computational domain, and as it is seen in the figure, when wind direction is defined from left to right, there is impinging on the up side of the wall where the wall has the first impact of the wind.Monitoring the vector and contour plots of velocity during calculation of the flow analysis are shown in figure 6 and 7. Also while figure 6 shows velocity vector and figure 7 shows velocity scalar, they have the same sense of velocity as expected.
In figure 8, a probe line was created along the flow direction in the middle of the computational domain.The purpose of the probe line is to plot some graphs along it.Also the probe line was located in 10 m height.Temperature field around the buildings are represented in figure 11, and the graph of temperature along the probe line is plotted in figure 12. Also in figure 11, the highest temperature appears at back side of the first building when the flow is on -x direction.Heat transfer coefficient around the buildings is shown in figure 13, and it is plotted along the probe line in figure 14.Heat transfer coefficient can be shown in formula as h=q/ΔT, where q is amount of heat required (heat flux), h is heat transfer coefficient, and ΔT is difference in temperature between the solid and surrounding fluid area.The temperature differences bottom side of the buildings' wall is less, and thus heat transfer coefficient is high at that region as shown in figure13 and 14.In addition, scalar of pressure distribution on the buildings is shown in figure 15, and also pressure drop is plotted as a graph along the probe line in figure 16. Figure 18 shows the turbulent kinetic energy, and figure 17 shows the non-dimensional wall distance.Also although their scalar fields are not the same, their senses are very similar as shown in those figure.Also their graphs can be compared.Figure 19 shows the graph of turbulent kinetic energy, and figure 20 shows the graph of wall y+.Turbulenceis a very important phenomenon that affects all atmospheric processes, but it is more important near the surfaces.Turbulence represents the irregularity or randomness of the flow.The kεturbulence model is used.The most unique measure for turbulence is kinetic energy of turbulent part of the flow.This variable depends on a variety of mechanisms.The most important are certainly buoyant production or consumption and shear production of turbulence kinetic energy (TKE).Both of these vary significantly in time and in space (especially in height).The third very important term is dissipation.This term provides the reduction of turbulence with time which is the primary purpose of turbulence.This causes that the energy is dissipated from large vortices to small one.
Also the purpose of the case study was to compute flow around A and B buildings of IUS campus using CDF.The main focus of the case study was how heat is transferred by the forced convection which is natural wind.The computation was used to compute a heat transfer coefficient at the buildings wall for the winter season.For that purpose, velocity field, heat transfer coefficient distribution along the buildings, pressure distribution, wall y+, turbulent kinetic energy, and temperature distribution will be shown in scalar, vector and graph forms.The temperature of outside the buildings is declared as -2℃, and the wind flows at 21℃ with the velocity 5 m/s.Also it can be said that as a result of the analysis, the IUS buildings are located in the optimum position, actually it was the most important information of this study case.

Figure
Figure 1describes the wind direction in the region in the winter season as defined.The wind direction

Figure 1 :
Figure 1: Direction of wind

Figure 2 :
Figure 2: The Computational domain and dimensions.

Figure 3 :
Figure 3: computational grids of the buildings.

Figure 9 :
Figure 9:velocity graph along the probe line.

Figure 10 :
Figure 10: Flow direction with velocity magnitude; (a) Real direction (b) Imaginary direction.In figure9, velocity changes are plotted along the probe line which was created in figure8.Also in figure10, real flow direction with velocity magnitude is shown on the part A, and imaginary flow direction is shown on the part B. when they are compared according to flow directions.In part B, there is more and more turbulent and it needs to less and less.Thus the real flow can be considered as the optimum flow direction.

Figure 11 :
Figure 11: temperature field around the buildings.

Figure 12 :
Figure 12: plotted graph of temperature along the probe line.

Figure 13 :
Figure 13: Heat transfer coefficient around the buildings.

Figure 14 :
Figure 14: graph of heat transfer coefficient around the buildings.

Figure 19 :
Figure 19: graph of turbulent kinetic energy [4]the above equations, Vbis the bulk velocity, ρ is the fluid density, μand μtare the laminar and turbulent dynamic viscosities respectively.The turbulence model constant used in the aboveequations are Cμ= 0.09[4].

Table 2 :
Boundary conditions and physical values.