sustainability
Article The Numerical Diffusion Effect on the CFD Simulation Accuracy of Velocity and Temperature Field for the Application of Sustainable Architecture Methodology
Vladimíra Michalcová 1 and Kamila Kotrasová 2,*
1 Department of Structural Mechanics, Faculty of Civil Engineering, VSB—Technical University of Ostrava, Ludvíka Podéštˇe1875/17, 708 33 Ostrava-Poruba, Czech Republic; [email protected] 2 Institute of Structural Engineering, Faculty of Civil Engineering, The Technical University of Košice, Vysokoškolská 4, 042 00 Košice, Slovakia * Correspondence: [email protected]
Received: 9 November 2020; Accepted: 3 December 2020; Published: 5 December 2020
Abstract: Numerical simulation of fluid flow and heat or mass transfer phenomenon requires numerical solution of Navier–Stokes and energy-conservation equations, together with the continuity equation. The basic problem of solving general transport equations by the Finite Volume Method (FVM) is the exact calculation of the transport quantity. Numerical or false diffusion is a phenomenon of inserting errors in calculations that threaten the accuracy of the computational solution. The paper compares the physical accuracy of the calculation in the Computational Fluid Dynamics (CFD) code in Ansys Fluent using the offered discretization calculation schemes, methods of solving the gradients of the transport quantity on the cell walls, and the influence of the mesh type. The paper offers possibilities on how to reduce numerical errors. In the calculation area, the sharp boundary of two areas with different temperatures is created in the flow direction. The three-dimensional (3D) stationary flow of the fictitious gas is simulated using FVM so that only advective transfer, in terms of momentum and heat, arises. The subject of the study is to determine the level of numerical diffusion (temperature field scattering) and to evaluate the values of the transport quantity (temperature), which are outside the range of specified boundary conditions at variously set calculation parameters.
Keywords: CFD; discretization scheme; numerical diffusion; transport equation
1. Introduction Aerodynamics deals with the movement of the air and the interaction between airflow and solid objects. Aerodynamics of buildings study the physical problems of airflow effects on buildings and their surroundings. The motion of the air—the wind—affects not only the design of the load-bearing parts of the building structures, but the dimensioning and construction of their non-load-bearing parts. One of the important areas of sustainable architecture is the knowledge of airflow effects on the surrounding objects [1]. The wind significantly affects the energy efficiency of buildings [2], associated with the general phenomenon of air filtration [3] (see Figure1), the details, elements, and systems of the packaging structures [4].
Sustainability 2020, 12, 10173; doi:10.3390/su122310173 www.mdpi.com/journal/sustainability Sustainability 2020, 12, 10173 2 of 19 Sustainability 2020, 12, x FOR PEER REVIEW 2 of 18 Sustainability 2020, 12, x FOR PEER REVIEW 2 of 18
Figure 1. Illustration of air filtration filtration and numerical simulation [[3].3]. Figure 1. Illustration of air filtration and numerical simulation [3]. TheThe aerodynamics aerodynamics of of a a building building examines the effect effect of the windwind onon thethe structurestructure itselfitself [[5],5], thethe airair velocityvelocityThe near nearaerodynamics the the structure structure of a [6–10] [building6–10] (see examines Figure 22), ),the theth effecte pressurepressure of the on onwind the the structureon structure the structure [ [5,11],5,11], itself thethe turbulenceturbulence [5], the air aroundvelocityaround the thenear building building the structure [12], [12], and and [6–10] the the influence (see influence Figure of of meteorological2), meteorological the pressure conditions conditionson the structure [13]. [13 ].The The [5,11], mutual mutual the grouping turbulence grouping of thearoundof thebuildings buildingsthe building modifies modifies [12], the and theairflow, the airflow, influence which which createsof meteorological creates the thewindy windy conditionsclimate climate in [13].their in their Thesurroundings mutual surroundings grouping [14], [ 14and of], affectstheand buildings aff theects human–wind the modifies human–wind the interactio airflow, interaction,n, which human humancreates safety, safety,the and windy thermal and thermalclimate comfort in comfort their [15]. surroundings [By15 ].studying By studying the[14], wind theand movementaffectswind movementthe human–windin relation in relation to the interactio wider to the topographicaln, wider human topographical safety, units, and it units,isthermal possible it is comfort possibleto positively [15]. to positively Byregulate studying the regulate efficiencythe wind the ofmovementeffi theciency ventilation of in therelation ventilation of tothe the urban wider of the units, topographical urban the units, scatte the units,ring scattering itof is thepossible of exhaust the to exhaust positively fumes, fumes, and regulate and the the theformation formation efficiency of snowdriftsofof snowdriftsthe ventilation around around theof thebuildings buildingsurban andunits, and th ethe line linescatte transport transportring structuresof structuresthe exhaust [16–18]. [16 fumes,–18 ]. and the formation of snowdrifts around the buildings and the line transport structures [16–18].
Figure 2. The Computational Fluid Dynamics (CFD) simulationsimulation of the wind around buildings [[10].10]. Figure 2. The Computational Fluid Dynamics (CFD) simulation of the wind around buildings [10]. TheThe results results of of the interaction of air movement and constructionconstruction cancan bebe obtainedobtained fromfrom realreal The results of the interaction of air movement and construction can be obtained from real measurementsmeasurements [19], [19], from from building building scale scale models models in the wind tunnel [[20–22],20–22], asas wellwell asas byby usingusing measurements [19], from building scale models in the wind tunnel [20–22], as well as by using computercomputer simulations. simulations. Computational Computational Fluid Fluid Dynamics Dynamics (CFD), (CFD), also also called called CFD CFD flow flow analysis, analysis, is isone one of computer simulations. Computational Fluid Dynamics (CFD), also called CFD flow analysis, is one of theof thebasic basic methods methods in innumerical numerical modeling modeling concerning concerning fluid fluid flow, flow, and and it it is is used as aa solutionsolution toto the basic methods in numerical modeling concerning fluid flow, and it is used as a solution to engineeringengineering problems problems in in almost almost all all industries industries [2 [233,24],,24], especially especially in in building building construction [4,10]. [4,10]. engineering problems in almost all industries [23,24], especially in building construction [4,10]. TheThe numerical numerical solution solution of of the general transport equations in thethe AnsysAnsys FluentFluent softwaresoftware byby thethe The numerical solution of the general transport equations in the Ansys Fluent software by the FiniteFinite Volume MethodMethod (FVM) (FVM) uses uses the the discretization discretization process, process, in which in which the basic the problem basic problem is the accurate is the Finite Volume Method (FVM) uses the discretization process, in which the basic problem is the accuratecalculation calculation of the transport of the transport quantity throughquantity thethrough walls the of the walls particular of the volumeparticular and volume its advective and its accurate calculation of the transport quantity through the walls of the particular volume and its advectiveflow across flow these across boundaries these boundaries [25]. When [25]. calculating, When calculating, it is necessary it tois countnecessary with to the count occurrence with the of advective flow across these boundaries [25]. When calculating, it is necessary to count with the occurrencethe so-called of “numericalthe so-called di “numericalffusion”, often diffusion”, referred often to in referred literature to asin “diliteratureffusion as error” “diffusion or “numerical error” or occurrence of the so-called “numerical diffusion”, often referred to in literature as “diffusion error” or “numericalviscosity” [ 26viscosity”], and with [26], the and occurrence with the of occurrence values that of are values outside that the are range outside of the the correct range of solution the correct [27]. “numerical viscosity” [26], and with the occurrence of values that are outside the range of the correct solutionThis non-physical [27]. This non-physical CFD artifact impairsCFD artifact the accuracy impairs ofthe the accuracy discrete of solutions the discrete of the solutions equations of the in solution [27]. This non-physical CFD artifact impairs the accuracy of the discrete solutions of the equationsdescribing in the describing advection the transport advection of the transport scalar [ 28of]. the It is scalar known [28]. that It the is numericalknown that di fftheusion numerical occurs equations in describing the advection transport of the scalar [28]. It is known that the numerical diffusionmainly in occurs the case mainly where in the the flow case directionwhere the is flow not parallel direction to is the not grid parallel walls to [29 the]. grid walls [29]. diffusion occurs mainly in the case where the flow direction is not parallel to the grid walls [29]. However,However, the the optimal states (parallel flow)flow) can only bebe achievedachieved byby calculatingcalculating straightstraight sectionssections However, the optimal states (parallel flow) can only be achieved by calculating straight sections ofof pipeline pipeline without without the the obstacles obstacles using using the the hexahedral hexahedral cells cells [30]. [30]. The The direction direction of of the the flow flow is isalways always in of pipeline without the obstacles using the hexahedral cells [30]. The direction of the flow is always in thein thegeneral general direction direction with with respect respect to the to cell the walls cell (hexahedral, walls (hexahedral, tetrahedral, tetrahedral, and polyhedral) and polyhedral) in the most in the general direction with respect to the cell walls (hexahedral, tetrahedral, and polyhedral) in the most flowsthe most [31]. flows When [31 evaluating]. When evaluating the advective the advective state, it state,is necessary it is necessary to consider to consider the numerical the numerical error flows [31]. When evaluating the advective state, it is necessary to consider the numerical error (numericalerror (numerical diffusion) diffusion) [32]. [This32]. Thiscauses causes considerab considerablele problems problems in inthe the numerical numerical solving solving of severalseveral (numerical diffusion) [32]. This causes considerable problems in the numerical solving of several technicaltechnical problems, problems, including including sustainable sustainable architecture. architecture. technical problems, including sustainable architecture. ThereThere are aa number number of of studies studies improving improving the numericalthe numerical solution. solution. Total Total Variation Variation Diminishing Diminishing (TVD) There are a number of studies improving the numerical solution. Total Variation Diminishing (TVD)schemes schemes have been have [33 been] a widely [33] applieda widely group applied of monotonicity-preserving group of monotonicity-preserving advection di ffadvectionerencing differencing(TVD) schemes schemes have for been partial [33] di fferentiala widely equations applied ingroup numerical of monotonicity-preserving computational fluid dynamics advection and heatdifferencing transfer schemes since the for last partial century. differential Many scientequationsific teams in numerical continue computational to develop this fluid TVD dynamics method. and It heat transfer since the last century. Many scientific teams continue to develop this TVD method. It Sustainability 2020, 12, 10173 3 of 19
Sustainabilityschemes for 2020 partial, 12, x FOR diff erentialPEER REVIEW equations in numerical computational fluid dynamics and heat transfer3 of 18 since the last century. Many scientific teams continue to develop this TVD method. It allows the allowsimplementation the implementation of the whole of the spectrum whole ofspectrum TVD schemes of TVD into schemes unstructured into unstructured networks, whilenetworks, their while exact theirformulation exact formulation was restored was on restored structured on networksstructured [34 networks]. The authors [34]. The [35] authors analyzed [35] the analyzed TVD diff erencingthe TVD differencingon unstructured on three-dimensionalunstructured three-dimensional meshes, focusing meshes, on the non-linearityfocusing on ofth TVDe non-linearity differencing of and TVD the differencingextrapolation and of thethe virtual extrapolation upwind node.of the Furthermore,virtual upwind they node. proposed Furthermore, a novel monotonicity-preserving they proposed a novel monotonicity-preservingcorrection method for the correction TVD schemes method that for significantly the TVD reducesschemes thethat numerical significantly diffusion reduces caused the numericalby mesh skewness.diffusion caused The authors by mesh [36 skewness.] analyzed The the authors causes [36] of the analyzed numerical the causes errors, of in the terms numerical of the errors,numerical in terms diffusion of the and numerical the compression diffusion arising and fromthe compression the use of the arising explicit from second-order the use totalof the variation explicit second-orderdiminishing schemestotal variation in the diminishing one-dimensional schemes advection in the simulation.one-dimensional advection simulation. TheThe presented presented work work offers offers possibilities possibilities on on how how to reduce the numerical errors inin AnsysAnsys FluentFluent softwaresoftware using using the the correct correct calculation settings. It It uses uses the available discretization schemes inin thethe software,software, together together with with the the available available solutions solutions methods of transport quantityquantity gradients.gradients. TheThe physicalphysical accuracyaccuracy of of the the calculations, calculations, for for the the various various combination-listed combination-listed parameters, is monitored on three typestypes ofof mesh. mesh. TheThe conclusions conclusions will will recommend recommend suitable suitable variants variants of calculations and, thus, contribute toto betterbetter numericalnumerical simulations simulations in in th thee field field of of construction. construction.
2.2. Method Method TheThe three-dimensional three-dimensional stationary stationary virtual virtual gas gas flow flow of the computational domain withwith dimensiondimension 11 × 11 × 0.250.25 m m (x( x× y ×y z) zis) issimulated simulated by by the the Finite Finite Volume Volume method method (FVM). (FVM). The The fictitious fictitious gas gas density density is × × 3 × × 1 1 is ρ = 1 kg−3 m . The values of the thermal conductivity λ −(W1 −m1 K ) and the dynamic viscosity ρ = 1 kg·m ·. The− values of the thermal conductivity λ (W·m ·K· ) −and· −the dynamic viscosity μ (Pa·s) µ (Pa s) of the gas are close to zero. The pressure-velocity coupling algorithm is determined by of the gas· are close to zero. The pressure-velocity coupling algorithm is determined by the segregated SIMPLEthe segregated method, SIMPLE which is method, suitable which for steady-state is suitable ca forlculations steady-state [37]. calculationsThe boundary [37 conditions]. The boundary are set soconditions that the identical are set sovectors that the enter identical at the two vectors mutually enter perpendicular at the two mutually walls with perpendicular the velocity walls vx and with vy. Onethe velocityof these wallsvx and hasvy .the One temperature of these walls T1 = has300 theK and temperature the secondT T12 == 300400 K. K andThe thepressure second outletT2 = is400 on theK. Thetwo pressureopposite outletwalls (the is on static the two pressure opposite p = 0 walls Pa). (theAs at static the entrance, pressure thep = outlet0 Pa). temperature As at the entrance, is also onethe wall outlet T temperature1 = 300 K and is the also second one wall wallT1 T=2 300= 400 K andK, see the Figure second 3. wall Thereby,T2 = 400 the K, sharp see Figure boundary3. Thereby, of the twothe sharpdomains boundary with the of temperature the two domains difference with the∆T temperature= 100 K is created difference in the∆T calculation= 100 K is createddomain. in This the boundarycalculation is domain.in the direction This boundary of the isflow in the (in direction the domain of the at flowthe (inangle the of domain 45°). Only at the the angle advective of 45◦). transmission,Only the advective in terms transmission, of the momentum in terms of and the momentum the heat, should and the occur heat, shouldfor the occur accurate for the numerical accurate calculation.numericalcalculation. The diffuse Thetransfer diffuse should transfer not shouldoccur. The not occur.output The on outputthe two on side the opposite two side walls opposite (in the walls xy plane)(in the isxy theplane) zero is flow the zeroof all flow quantities of all quantities across the across border the (symmetry border (symmetry boundary boundary condition, condition, normal velocitynormal is velocity zero). These is zero). two These sides two of the sides opposite of the oppositewalls are walls not shown are not in shown Figure in 3. Figure 3.
Figure 3. The calculation domain scheme 1 × 11 × 0.250.25 m m and and the the boundary boundary conditions. conditions. × × The solution of the problem is independent of the domain dimensions, gas density, and velocity. It was verified on pilot calculations, whereby gradually changing all of these values in the thousands. These test tasks were also performed for porous domain, with a wide range of permeability values corresponding to building materials. The transport quantity does not have to be only the temperature, Sustainability 2020, 12, 10173 4 of 19
The solution of the problem is independent of the domain dimensions, gas density, and velocity. It was verified on pilot calculations, whereby gradually changing all of these values in the thousands. These test tasks were also performed for porous domain, with a wide range of permeability values corresponding to building materials. The transport quantity does not have to be only the temperature, but also, for example, the concentration of substances. In the presented paper, the size of the calculation area is set so that the calculations did not have the problem with above-standard number of cells and the scattering of the temperature field (numerical error—diffusion) was obvious at the same time. The air density is close to the air density. The velocity is chosen so that the calculations converged in an acceptable time. The Ansys Fluent software uses the FVM to convert the general transport equations to the system of linear equations that are solved numerically by the Gauss–Seidel iteration method. This solution consists in integrating the equations in each control volume (cell), where the result is the discrete equations presented the flow equilibrium (the conservation laws of the transport quantity Φ in given volume). Its mathematical description for the stationary flow in the integral form is: Z Z Z ρ →v Φd→A = Γ Φd→A + S dV, (1) · · Φ·∇ Φ A A V where →A (m s 1) is the surface vector, V (m3) is the control volume (cell volume), ρ (kg m 3) is the · − · − density of the flowing medium, →v (m s 1) is the velocity vector, Φ (K) is the value of the transport · − quantity (temperature), Φ (K m 1) the gradient of the transport quantity Φ, Γ is the diffusion ∇ · − Φ coefficient of the transport quantity Φ (K kg m 1 s 1), S (K kg m 3 s 1) is the source term of the · · − · − Φ − − quantity Φ per unit of the volume. Equation (1) is applied to all control volumes of the calculation area. Equation (2) is obtained by the discretization of the Equation (1) in the given cell:
N N Xf aces Xf aces ρ →v Φ →A = Γ Φ →A + S V, (2) f · f · f · f Φ·∇ f · f Φ· f n
→ 1 where N s is the number of the faces surrounding of the cell, ρ →v A (kg s ) is the mass flow over face f · f · f · − → 2 the surface f, A f (m ) is the surface vector f, Φf (K) is the value of the transport quantity flowing over the surface f (the face value), Φ (K m 1) is the gradient of the transport quantity Φ on the surface f. ∇ f · − The left side in both equations represents the advective transfer of the quantity Φ, the right side expresses the diffuse transfer and the source term of the transport quantity Φ (its decrease or increase). The basic problem in the discretization of the advective term is the exact calculation of the transport quantity on the face of the specific volume Φ and its gradient Φ . The diffusion process f ∇ f affects the transfer of the transport quantity along its gradient in all directions, while the advective transfer pervades only in the direction of the flow. It is very difficult to find the exact discretization computational scheme for the solving of the advective term in the Equation (2). The software Ansys Fluent stores the discrete values of the scalar quantity Φ in the center of the cell. The values of the scalar quantity Φf on the cell face are required for the calculation of the advective term in the equations and they are determined by the interpolation from the values in the centers of the adjacent cells. The number of the surrounding cells depends on the type of the grid, but, in most cases, the amount is the same as the number of the faces forming of the interest cell. The discretization “upwind” schemes are used for this process; it means that the value Φf is derived from the value of the next cell in the flow direction. The most upwind schemes require the determination of the transport quantity gradient Φ ∇ for their solution. The gradients are necessary for the calculation of the scalar values on the cells faces not only for discretization the advective but also for the diffusion term in the Equation (2). Sustainability 2020, 12, 10173 5 of 19
3. The Parameters Influencing the Accuracy of the Calculation The level of the physical accuracy of the numerical calculation is influenced by the mesh type, the choice of discretization schemes for the conversion of the general transport equations to the linear equations, and also the choice of the calculating method of the transport quantity gradient Φ ∇ (here temperatures). Sustainability 2020, 12, x FOR PEER REVIEW 5 of 18 3.1. The Mesh Type and the Mesh Density Three types of the grids: hexahedral, tetrahedral, and polyhedral are used for solving the problems,Three see types Figure of the 4. grids: All grid hexahedral, types have tetrahedral, double anddensity. polyhedral The coarse are used hexahedral for solving and the tetrahedral problems, meshsee Figure were4 formed. All grid from types 25 havecells doublewith the density. length Theof 1 coarsem at all hexahedral longitudinal and and tetrahedral vertical edges mesh wereof the domainformed ( fromx, y-axis 25 direction), cells with thesee Figure length 5a. of 1The m 6 at cells all longitudinalper 0.25 m are and formed vertical on the edges domain of the edges domain in z- axis(x, y direction.-axis direction), The fine see hexahedral Figure5a. and The tetrahedral 6 cells per 0.25mesh m were are formed formed on from the 100 domain cells per edges 1 m in inz the-axis x, ydirection.-axis directions, The fine see hexahedral Figure 5b. and The tetrahedral 25 cells pe meshr 0.25 were m formedare formed from on 100 the cells domain per 1 m edges in the inx, yz-axis-axis direction.directions, The see polyhedral Figure5b. Themesh 25 was cells formed per 0.25 directly m are formedin the Ansys on the Fluent domain from edges the in tetraz-axis cells. direction. The six calculationThe polyhedral areas mesh with wasthe identical formed directly dimensions in the (Figure Ansys Fluent3) and from the parameters the tetra cells. listed The in six Table calculation 1 were created.areas with the identical dimensions (Figure3) and the parameters listed in Table1 were created.
(a) (b) (c)
FigureFigure 4. 4. ThreeThree mesh mesh types types and and axis scheme onon whichwhich thethe transporttransport quantities quantities are are evaluated; evaluated; the coarse themesh: coarse (a) mesh: hexahedral; (a) hexahedral; (b) tetrahedral; (b) tetrahedral; (c) polyhedral. (c) polyhedral.
(a) (b)
Figure 5. TwoTwo different different densities of the tetrahedral mesh: ( a) coarse; (b) fine.fine.
Table 1. Number of cells in computational areas. Table 1. Number of cells in computational areas. Mesh Type Hexahedral Tetrahedral Polyhedral Mesh type Hexahedral Tetrahedral Polyhedral Values in Values in Thousands25 Cells 25 Cells 100 100 Cells Cells 25 25 Cells Cells 100 CellsCells 25 25 Cells Cellstetra 100100 Cells Cellstetra Thousands tetra tetra Cells number 3.8 250 28.4 1660 5.9 294 CellsFaces number number 3.8 12.2 250 765 28.4 58.9 16603355 38.8 5.9 2023 294 Faces number 12.2 765 58.9 3355 38.8 2023 Nodes number 4.7 265 5.8 294 32.9 1729 NodesNodes number number 4.7 4.7 265 265 5.8 5.8 294 32.9 32.9 1729 1729
3.2. The Discretization Scheme 3.2.3.2. The The Discretization Discretization Scheme Scheme It is possible that the choice from the five upwind discretization schemes for the given problem. ItIt is is possible possible that that the the choice choice from from the the five five upwind discretization schemes for the given problem.problem. Their brief description and the scheme are in Table 2. Their brief description and the scheme are in Table2.
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 18
Sustainability 2020, 12, x TableFOR PEER 2. The REVIEW discretization schemes for the adventive flow calculation. 6 of 18
SustainabilityFlow Direction 2020, 12, x Table→ FOR PEER 2. The REVIEW discretization Description schemes for the of adventiveDiscretization flow calculation. Schemes 6 of 18 Sustainability 2020, 12, 10173 6 of 19 Sustainability 2020, 12, x FOR PEERFirst-order REVIEW upwind scheme is based on the assumption that the value6 ofof 18 Flow Φ(x)Direction interpolated →Table 2. The discretizationDescription schemes for the of adventiveDiscretization flow calculation. Schemes ΦP value the quantity in the cell center corresponds to the average value in the Φ ef First-order upwind scheme is based on the assumption that the value of whole control volume and the face value Φf is set equal to the cell-center Flow Φ(x)Direction interpolatedTableΦE 2. Table→The discretization2. The discretization schemes Descriptionschemes for the adventivefor the of adventiveDiscretization flow calculation. flow calculation. Schemes W P E the quantity in the cell center corresponds to the average value in the ΦP e value value of Φ in the upstream cell. The first-order upwind scheme is not Φ ef First-order upwind scheme is based on the assumption that the value of Flow Direction → whole control volumeDescription and the offace Discretization value Φf is set Schemes equal to the cell-center Flow Φ(x) Direction interpolatedΦE depended on the gradientDescription ∇Φ. of Discretization Schemes W Φ P P e value E→ the quantity in the cell center corresponds to the average value in the Φ First-ordervalue of Φ inupwind the upstream scheme cell. is based The first-orderon the assumption upwind schemethat the isvalue not of Φ(x) interpolated ef First-order upwind scheme is based on the assumption that the Φ whole control volume and the face value Φf is set equal to the cell-center Φ Evalue thedependedPower quantity law on scheme in the the gradient cellis based center ∇ onΦ corre. the analyticalsponds to solutionthe average the valueone-dimensional in the W P P e E value of the quantity in the cell center corresponds to the average Φef value of Φ in the upstream cell. The first-order upwind scheme is not advection-diffusion equation. The face valuef Φf is determined from the Φ(x) interpolatedΦE whole controlvalue in volume the whole and control the volumeface value and Φ the is face set valueequalΦ tof is the set cell-center equal W P E Powerdepended law onscheme the gradient is based ∇ onΦ. the analytical solution the one-dimensional ΦP e value valueexponential ofto Φ the in profile cell-center the upstream by valueusing cell. of theΦ The invalues the first-order upstream of the cell cell.upwind in The their first-orderscheme center. is The not advection-diffusion equation. The face value Φf is determined from the Φ(x) interpolated Φef ΦE upwind scheme is not depended on the gradient Φ. dependedexponentialPower law on scheme profile the gradient is baseddepended ∇ onΦ. the on analytical the Peclet solution number∇ thePe (ratioone-dimensional of the W Φ P e value E P exponentialconvectionPower heat profile law flow scheme by (advection) usingis based the onvalues and the the analytical of heat the cellflow solution in by their the the center. mechanism The of advection-diffusion equation. The face value Φf is determined from the Φ(x) interpolated Φef ΦE one-dimensional advection-diffusion equation. The face value Φf is Powerexponentialthe convection law scheme profile (diffusion)). is baseddepended Theon the solutionon analytical the Peclet is the solution number same with thePe (ratioone-dimensionalthe first-order of the W Φ P P e value E exponentialdetermined profile from by theusing exponential the values profile of the by cell using in thetheir values center. of the The advection-diffusionconvection heat flow equation. (advection) The and face the value heat Φ flowf is determined by the mechanism from the of Φ(x) interpolated upwind scheme for |Pe| ≥ 10. Φef ΦE exponentialcell in profile their center. is depended The exponential on the profile Peclet is number depended Pe on (ratio the Peclet of the ΦP value W P e E theexponential convectionnumber profile Pe (diffusion)). (ratio by ofusing the convectionThe the solutionvalues heat of is flowthe the cell (advection)same in theirwith and thecenter. the first-order heat The convection heat flow (advection) and the heat flow by the mechanism of Φef ΦE upwindexponentialSecond-orderflow scheme by profile theupwind for mechanism is|Pe| depended scheme: ≥ 10. of the the on convection face the Pecletvalue (di Φnumberffusion)).f is determined Pe The (ratio solution from of the is the Φ W interpolatede W P E cellthe convectionvaluesthe same in the (diffusion)). with two the cells first-order upstream The solution upwind of the schemeis face.the same forThis|Pe withscheme| 10. the requiresfirst-order the Φ(x) value convection heat flow (advection) and the heat flow by ≥the mechanism of Second-orderupwindSecond-order scheme upwind for upwind|Pe| scheme: ≥ 10. scheme the :face the facevalue value Φf isΦ determinedis determined from the Φ interpolated thedetermination convection of(diffusion)). the gradient The ∇ Φsolution in each is cell. the Thissamef is with more the accurate first-order than W ΦP Φef ΦE cell valuesfrom in the the cell two values cells in upstream the two cells of upstreamthe face. ofThis the scheme face. This requires the W Φ(x)P e value E upwindthe first-order scheme upwind for |Pe| scheme, ≥ 10. but in the regions with the strong determinationSecond-orderscheme requiresupwindof the gradient the scheme: determination ∇ theΦ in face each of value the cell. gradient Φ Thisf is determined is moreΦ in each accurate cell.from thethan Φ W interpolated ∇ ΦP Φef ΦE gradients,cell valuesThis it in is can morethe result two accurate cells in the thanupstream face the va first-orderlues of the that face. upwind are This outside scheme, scheme of butthe requires in range the of the Φ(x)e value Second-orderthe first-order upwindupwind scscheme,heme: thebut face in the value regions Φf is withdetermined the strong from the W P E the correctregions cell with values. the strong gradients,∇ it can result in the face values that Φ W interpolated cellgradients,determination values itin can the of result twothe gradientcells in the upstream face Φ va inlues ofeach the that cell. face. are This This outside is schememore of accuratethe requires range than ofthe Φ(x)ΦP Φ evaluef ΦE are outside of the range of the correct cell values. e the first-order upwind scheme, but in the regions with the strong W P E determinationtheQuadratic correctQuadratic Upwindcell ofvalues. Upwindthe Interpolation gradient Interpolation ∇Φ for in eachConvective for Convective cell. This Kinetics is Kinetics more (QUICK) accurate than ΦP Φef ΦE scheme.gradients, The it canquadratic result incurve the faceis fitted values with that two are upstream outside ofnodes the rangeand one of W P e E the first-order(QUICK) upwind scheme .scheme, The quadratic but in curve the regions is fitted withwith twothe upstreamstrong ΦW interpolated Quadraticdownstreamthe correctnodes Upwind cell andnode. values. one ThisInterpolation downstream scheme node.requires for Convective This the scheme determination Kinetics requires (QUICK) the of the gradient Φ(x) value gradients, it can result in the face values that are outside of the range of scheme.determination The quadratic of the curve gradient is fittedΦ inwith each two cell. upstream This is a very nodes accurate and one ΦP the∇Φ incorrect each cellcell. values. This is a very accurate∇ scheme, but it can lead to stability ΦW interpolatedΦef ΦE downstreamQuadraticscheme, Upwind node. but it ThisInterpolation can lead scheme to stability requires for Convective problems the determination in theKinetics calculation (QUICK) of the in the gradient Φ(x) value problems in the calculation in the regions with strong gradients. W P e E scheme.regions The quadratic with strong curve gradients. is fitted QUICK with scheme, two upstream documented nodes in [and37], one ΦP Quadratic∇Φ in each Upwind cell. This Interpolation is a very accurate for Convective scheme, but Kinetics it can lead(QUICK) to stability ΦW interpolatedΦef ΦE QUICKdownstreamis scheme, mainly node. suitable documented This for scheme the proin requires[37], structured is mainly the hexahedral determination suitable meshes, for the of but thepro it gradient Φ(x) value problemsscheme. The in thequadratic calculation curve in is the fitted regions with with two upstreamstrong gradients. nodes and one W P e E structuredcan behexahedral used for the meshes, unstructured but it orcan the be hybrid used meshes.for the unstructured or Φ W interpolated ΦP ∇Φ in each cell. This is a very accurate scheme, but it can lead to stability Φef ΦE QUICKdownstream scheme, node. documented This scheme in requires[37], is mainly the determination suitable for the of thepro gradient Φ(x) value the hybridThird-order meshes. Monotone Upstream-centered Scheme for W P e E problems in the calculation in the regions with strong gradients. ΦP structured∇Φ in eachConservation hexahedralcell. This Laws is meshes,a very (MUSCL). accurate but This it can scheme, third-order be used but convectionfor it thecan unstructuredlead scheme to stability or Φef ΦE Third-orderQUICKwas scheme, created Monotone documented from theUpstream-centered original in [37], MUSCL is mainly [28 Scheme], by suitable mixing for Conservation thefor centralthe pro Laws W P- e E theproblems hybrid in meshes. the calculation in the regions with strong gradients. differentiation scheme and the second-order winding scheme. This QUICK(MUSCL).structured scheme, Thishexahedral third-order documented meshes, convection in but [37], it canis scheme mainly be used wassuitable for created the for unstructured thefrom pro the or Third-orderscheme Monotone requires the Upstream-centered determination of the Scheme gradient forΦ Conservationin each cell. It Laws - originalthe hybrid MUSCL meshes. [28], by mixing the central differentiation∇ scheme and the structuredis usable hexahedral for all grid meshes, types. but it can be used for the unstructured or (MUSCL).second-order This winding third-order scheme. convection This sc hemescheme requires was created the determination from the of - theThird-order hybrid meshes. Monotone Upstream-centered Scheme for Conservation Laws originalthe gradient MUSCL ∇Φ in[28], each by cell. mixing It is the usable central for alldifferentiation grid types. scheme and the 3.3. The Solution of the Transport Quantitysecond-orderThird-order(MUSCL). Gradients This Monotone winding third-order scheme.Upstream-centered convection This scheme scheme Scheme requires was forcreated the Conservation determination from the Laws of - the(MUSCL).original gradient MUSCL This ∇Φ third-order in[28], each by cell. mixing convection It is theusable central scheme for alldifferentiation gridwas createdtypes. schemefrom the and the The3.3. gradients The Solution are of needed, the Transport not only Quantity for the Gradients constructing values of the scalar at the cell faces, but - originalsecond-order MUSCL winding [28], by scheme. mixing This the centralscheme differentiation requires the determination scheme and the of also for computing the secondary diffusion terms∇ and the velocity derivatives. The gradient Φ of 3.3. TheThe Solution gradients of the are Transport needed,second-orderthe gradient Quantitynot only windingΦ Gradientsfor in theeach constructischeme. cell. It isThisng usable valuesscheme for of allrequires the grid scalar types. the at determination the cell∇ faces, of but a given the variable Φ is used to discretization of the advection and the diffusion terms in the flow also for computing the secondarthe gradienty diffusion ∇Φ in termseach cell. and It the is usablevelocity for derivatives. all grid types. The gradient ∇Φ of a conservation equations. The gradients are computed in Ansys Fluent, according to the three following given3.3. TheThe the Solution gradients variable of theareΦ isTransport needed, used to Quantitynot discretization only Gradientsfor the of constructi the advengction values and of the the diffusion scalar at termsthe cell in faces, the flow but methods:also for computing the secondary diffusion terms and the velocity derivatives. The gradient ∇Φ of a 3.3.conservation The Solution equations. of the Transport The gradients Quantity are Gradients computed in Ansys Fluent, according to the three following givenThe the gradients variable Φare is needed, used to not discretization only for the of constructi the advengction values and of the the diffusion scalar at termsthe cell in faces, the flow but 1. Green–Gaussmethods: cell-based; ∇ conservationalso forThe computing gradients equations. arethe needed, secondarThe gradients noty diffusion only are for computed theterms constructi and in Ansystheng velocity values Fluent, derivatives.of according the scalar toThe at the the gradient three cell followingfaces, Φ ofbut a 2. Green–Gaussalsomethods:1.given forGreen–Gauss the computing variable node-based; cell-based; Φthe is secondarused to ydiscretization diffusion terms of theand adve the ctionvelocity and derivatives. the diffusion The terms gradient in the ∇Φ flowof a 3. Leastgiven2.conservation Green–Gauss Squares the variable cell-based.equations. node-based;Φ is Theused gradients to discretization are computed of the in adve Ansysction Fluent, and theaccording diffusion to the terms three in following the flow 1. Green–Gauss cell-based; conservation3.methods: Least Squares equations. cell-based. The gradients are computed in Ansys Fluent, according to the three following The2. firstGreen–Gauss two methods node-based; use the Green–Gauss theorem to compute the Φ at the cell center c0: methods: ∇ 1. TheGreen–Gauss first two methods cell-based; use the Green–Gauss theorem to compute the ∇Φ at the cell center c0: 3. Least Squares cell-based. X 2. Green–Gauss node-based; 1 → 1. Green–Gauss cell-based; ( Φ)c0 = Φ f A f , (3) 3. TheLeast first Squares two methods cell-based. use the∇ Green–GaussV theorem1 · to compute the ∇Φ at the cell center c0: 2. Green–Gauss node-based; ∇Φ f = Φ ∙ 𝐴⃗ , 𝑉 (3) 3. LeastThe first Squares two methods cell-based. use the Green–Gauss theorem1 to compute the ∇Φ at the cell center c0: The face value Φ by default of the Green–Gauss cell-based is⃗ taken from the arithmetic average f ∇Φ = Φ ∙ 𝐴 , (3) 𝑉 ∇ of the valuesThe at first the neighboringtwo methods cell use centers: the Green–Gauss theorem1 to compute the Φ at the cell center c0: ∇Φ = Φ ∙ 𝐴⃗ , 𝑉 (3) Φ + Φ1 Φ = ∇Φc 0 = ci Φ.∙ 𝐴⃗ , (4) f 2 𝑉 (3) Sustainability 2020, 12, x FOR PEER REVIEW 7 of 18
The face value Φf by default of the Green–Gauss cell-based is taken from the arithmetic average of the values at the neighboring cell centers:
Φ +Φ Φ = . (4) 2 Sustainability 2020, 12, 10173 7 of 19 The face value Φf by default of the Green–Gauss node-based is taken by the arithmetic average of the nodal values on the face: The face value Φf by default of the Green–Gauss node-based is taken by the arithmetic average of the nodal values on the face: , N 1 Xf ,nodes Φ = 1 Φ , (5) Φ f = 𝑁 , Φn, (5) N f ,nodes n where Nf,nodes are the number of the nodes on the face. The nodal values Φn are determined from the where Nf,nodes are the number of the nodes on the face. The nodal values Φn are determined from the weightedweighted average average of of the the cell values surrounding the nodes. The node-basednode-based gradientgradient methodmethod isis notnot availableavailable with with the the polyhedral polyhedral meshes. meshes. TheThe third third least least squares squares cell-based cell-based method method is is based on assumption the change inin thethe cellcell valuesvalues between the cell c0 and ci along the vector 𝑟⃗ from the center of the cell c0 to ci is expressed as: between the cell c0 and ci along the vector →r i from the center of the cell c0 to ci is expressed as: