FLOW3D铸造_进阶培训_物理选项.ppt

1、新近增加了的模型(2007年和2008年版本)。 常规模式 特别输出,物理工具和模型参数设置详解,目录,Inlet BC-体积流量网格边界条件 - 离心铸造入口逆流旋转counter-rotating inlet flow in centrifugal casting -质量/动力源 Filling-卷气和表面缺陷跟踪 -紊流、涡流粘度 turbulence - open prpplt plot file in Analyze; - look at the graphs of angular velocities and accelerations vs. time.,new in 9.3,1

2、,2,Example with and without Counter-rotating Flow,Without the counter-rotating flow component the incoming metal is dispersed by the centrifugal force,With the counter-rotating flow component the incoming metal jet is much more coherent,counter-rotation,质量/动力源,Unlike mass sources that are associated

3、 with components and characterized only by flow rate, mass/momentum sources are not associated with geometry and can be defined anywhere in the open space. These sources are characterized by time-dependent flow rate and direction. Mass/momentum sources can be used to define inlets inside the domain

4、as an alternative to using mesh boundary conditions.,Both sources have the same mass flow rate. Mass/momentum sources have a distinct direction to the emerging flow, characterized by the fluid velocity, u = Q/An.,Mass/Momentum Sources vs. Component Sources,mass/momentum source,mass source,A mass/mom

5、entum source approximates a flat, two-dimensional orifice out of which emerges liquid metal, e.g., a cross-section of a pipe.,质量/动力源 : 解释,Mass/Momentum Sources: 定义,shape,Up to 100 mass/momentum sources can be defined, each characterized by:,size,location and orientation,mass flow rate,source motion,

6、shape size location and orientation flow rate,new in 9.3,Mass/momentum sources are represented in the computational domain with a special type of particles. These particles can be used to visualize the sources on 2D and 3D plots.,Mass/Momentum Sources: Visualization目测查看,充填期间的表面缺陷跟踪 : 氧化和夹渣,Oxidation

7、 and other inclusions forming on liquid metal free surface can cause structural defects in the final casting. The defect tracking model predicts the location and the likelihood of the potential defects during filling,maximum bending stress of an aluminum plate as a function of filling rate (from Joh

8、n Campbell),micrograph of a crack surface of a motorcycle break pedal; reportedly, the pedal broke during a ride,充填期间的表面缺陷跟踪 : 模型界面,Calculates a concentration of oxides, based on exposure time to air. The value for oxide generation Rate is constant and arbitrary (0.0). The result is interpreted as a

9、 probability of a defect the higher the value, the higher the probability. C(x,t) - oxide film concentration; t time.,at free surface in the bulk,Similar model applies to foam residue in the lost foam process. The residue is formed at the metal/foam interface as a fraction of the burnt foam.,likely

10、location of defects,充填期间的表面缺陷跟踪 : 范例,When analyzing the results, look for “hot spots” of the defect concentration localized peaks in concentration.,Aluminum die casting,充填期间的表面缺陷跟踪 : 更多范例,Magnesium die casting,Aluminum die casting,Lost foam (aluminum),充填期间的卷气,Simple entrainment of air due to turbule

11、nce during filling may cause defects like porosity, leakage and fracture.,Air entrainment occurs as a result of the balance of three factors: - turbulence creates perturbations on the free surfaces, - gravity and surface tension act as stabilizing forces. Entrained air is recorded as a volume fracti

12、on Cair of the air/metal mixture.,a,卷气: 模型界面,s,Limitations: - pressure effect on entrained air volume is not included. As a result, pressurization of the cavity does has no effect on the amount of air in the metal; - the size of entrained air bubbles is not predicted.,Note: It is not necessary to us

13、e turbulence for this model,卷气: 范例,entrained air,defect tracking model,porosity in an aluminum die casting,As with the defect tracking model, look for “hot spots” of the air volume fraction highly localized peaks in concentration.,卷气: 范例 #2,The air entrainment model can be used in two modes: Passive

14、 mode : The entrained air does not alter the dynamics of the metal. Once entrained, the air stays in the metal. Active mode: In the active model the following phenomena are added: metal/air density is computed as a weighted average (buoyancy) the bulk volume of the metal increases (bulking) entraine

15、d air bubbles drift though the metal and escape back into the atmosphere at free surface (de-aeration) The active mode should be used when: air entrainment is significant time is long enough for the air to separate and escape; this may happen in gravity and low pressure filling processes, but not in

16、 pressure die casting.,卷气: 模型选项,- activate variable density model (no heat transfer can be used then) - set air density - activate drift-flux model with the escape option set average air bubble size,bottle filling,激活卷气模型,variable density model,air density,average bubble size,air escape at free surfa

17、ce,activate drift-flux model,紊流,flow inside mold during filling is usually at high Reynolds numbers ( 20,000) and is, therefore, turbulent; out of the five available models, the RNG model is the best choice; RNG CPU load is not excessive, 20%; but typically, better mesh resolution is needed to captu

18、re turbulence than for laminar flows; do not use turbulence when the flow is not turbulent.,available turbulence models,When using a turbulence model, Turbulent mixing length (TML) should be set. TML is related to the average scale of turbulent eddies in the flow. In casting, it is related to the sm

19、allest of the average wall thickness, d, and runner and gate width, h: TML=0.07*min(d,h) Hint: the larger the value of TML, the higher the level of turbulence.,紊流:紊流混合长度,turbulent mixing length,Generally, turbulence models are not been validated for filling analysis in great detail, especially for h

20、igh pressure die casting. The main experimental parameter to match is the filling time, as in gravity casting (hard to measure anything else). If filling rate is fixed, as in high pressure die casting, then modeling turbulence is not usually necessary. In many cases of complex geometries, it is hard

21、 to achieve sufficient resolution to benefit from the inclusion of turbulence in the filling model, especially in thin-wall castings.,紊流: Is It Worth the Price?,A simpler approach, Eddy Viscosity, can be used to approximate turbulent behavior during filling: - laminar flow option; - increased consta

22、nt viscosity coefficient, meddy, to represent a uniform level of turbulence The value meddy is computed from the Reynolds number at a location in the flow where velocity is known, e.g., at the inlet: meddy = rUL / 2000 U = average speed at the selected location r = metal density L = average inlet, r

23、unner or gate width,紊流:涡流粘度的近似法 Eddy Viscosity Approach,laminar flow,eddy viscosity coefficient,紊流: 涡流粘度近似法的验证,Recommended for high pressure die casting filling. Faster than a full turbulence model, less memory required, provides adequate results.,过滤网,Filters in FLOW-3D are described with the porous

24、 medium model. A filter is represented by a porous geometry component. Each porous component is characterized by several parameters describing flow losses and heat transfer. Filtration of inclusions is not included in the model.,location of filters,过滤网: 定义,Activate a porous media model in Physics Ad

25、d a new component, of type Porous Define porous media properties: porosity, between 0.0 (fully blocked) and 1.0 (fully open) specific area (for heat transfer) drag coefficients (flow losses) capillary pressure,过滤网: 目测查看,Porous Media is visualized on 2D Display and 3D FAVORizer plots as arrays of sma

26、ll tetrahedrons. 3D Iso-surface plots in Display are based solely on the volume fraction values.,acceleration,inertia,pressure,viscous forces,gravity,flow losses in porous media,过滤网: 定义阻力系数,Two main flow loss models: 1. Linear, porosity-dependent drag coefficient: 2. Quadratic, Reynolds number-depen

27、dent drag coefficient:,DArcy permeability:,过滤网:线性拖曳模型,U - bulk metal velocity u - microscopic metal velocity,a,porous media,过滤网:二次拖曳模型,d the average pore size,bulk velocity U,Dp,Positive capillary pressure describes hydrophilic media. Negative capillary pressure describes hydrophobic media. Hint: su

28、rface tension model does not have to be activated to include capillary pressure in porous media.,过滤网: Capillary pressure毛细管压力,Capillary pressure accounts for the surface tension forces in porous media acting at metals free surface inside the pores; these forces can add resistance to the metal (hydro

29、phobic media) or promote flow (hydrophilic media),过滤网: Heat Transfer热传,Heat transfer with porous media is defined the same way as with the mold, except for one additional variable for each porous component: Specific Surface Area It describes the surface area of the filter per unit bulk volume and is

30、 a function of its microscopic structure. It is typically provided with the filter specifications. The default value is 0.0 which means no heat transfer inside the filter.,过滤网: Output输出,Whenever a porous media model is activated the normalized drag coefficient is stored as a spatial variable in the

31、flsgrf data file: where Dt is the time step size. DRG = 0.0 outside porous media (K=0.0) DRG 0.0 inside porous media DRG is closer to 1.0 for higher flow losses.,Air Back Pressure空气背压,Air can be trapped inside metal and mold and pressurize providing additional resistance to the metal flow. In the on

32、e-fluid approach to modeling filling, air can be approximated as bubbles with uniform properties (pressure and temperature); no air flow is computed within the bubbles. Three main modes of treating the air: instant and complete venting: constant pressure bubbles: gravity filling in sand molds no ven

33、ting at all: adiabatic bubbles: high pressure die casting partial venting: adiabatic bubbles + vents: gravity die casting,individual bubbles,To use the adiabatic bubble model: activate the model in Physics define Gamma of 1.4 in Fluids property tree in the Initial tab use the absolute value for the

34、atmospheric pressure for Void initial state pressure,g = 1.4 for air,背压: Adiabatic Bubbles绝热气泡,For a gas bubbles without heat and mass exchange with the surroundings, its pressure is a simple function of the volume: where V0 and P0 are the initial bubble volume and pressure.,1,2,3,Bubbles connected

35、to a fixed-pressure mesh boundary with fluid fraction F=0.0, assume the boundary pressure. It is recommended to always use the adiabatic bubble model for permanent mold filling problems, especially high pressure die casting. Works well with the oxide film and air entrainment models.,绝热气泡: 最终备注,Alumi

36、num structural component die casting,恒量, 非均匀模温,Thermal die cycling is modeled in die casting to obtain a realistic temperature distribution in the die. If filling is then carried out as a restart simulation, these temperatures are used as the initial conditions. Since filling time is very short comp

37、ared to the thermal diffusion time scale in the die, it makes sense to assume that the die temperature does not change during filling. This simplification provides significant gains in the speed of the filling simulation over 10 times shorter times and much less memory.,恒量, 非均匀模温, 接上页,Due to the ass

38、umption of the constant die temperature, the metal looses more thermal energy during filling, equivalent to the average temperature of about 5-10 degrees for aluminum castings. It is important to provide all the same thermal properties for the die as in the full heat transfer model, e.g., during the

39、 thermal die cycling simulation. Otherwise, the metal/die heat transfer calculations will not be accurate. Note: a non-uniform, constant mold temperature can also be defined in the initial conditions, instead of taking it from a results file.,凝固过程中孔隙度,Porosity forms as a result of the change in meta

40、l density during solidification. In many (but not all) alloys the solid phase density is larger than the density of the liquid phase. As a result, the metal shrinks as it cools and solidifies. Porosity is the #1reason for rejected castings: Adequate feeding must be provided to compensate for the shr

41、inkage and/or move it to “safe” locations.,macroporosity,microporosity,- poor mechanical properties - leakage - surface defects,Two models are available in FLOW-3D to predict porosity: - macroposity (Rapid Shrinkage) - microporosity Both models assume a linear dependency of metal density r on solid

42、fraction fs: where rs and rl are the user-defined constant solid and liquid phase densities, respectively. Both models are based on the thermal solution. Feeding is computed from the three-dimensional distribution of temperature and solid fraction. The models can be used in combination with each oth

43、er. The rapid shrinkage model assumes no fluid flow. The microporosity model can be used with any flow and solidification model,凝固过程中孔隙度 : 模型建立,Several feeding modes are considered in the model: at early stages of solidification, below solid coherency point, gravity is the dominant factor. later dur

44、ing solidification, feeding occurs by interdendritic flow, independently of gravity. no feeding occurs when the solid fraction exceeds the critical value. The distinction is made using critical and coherency solid fractions.,凝固过程中孔隙度 :快速收缩模型,g,solid,I. Feeding by gravity: min(Fs) FSCO,II. Interdendr

45、itic feeding at later stages: min(Fs) FSCO,cooling,solid,liquid,mushy,very mushy,凝固过程中孔隙度 :钢锭范例,FSCO=0.15 FSCR=0.67 (default),FSCO=0.0 FSCR=0.67,FSCO=0.7 FSCR=0.8,Macroporosity is best displayed in terms of fluid fraction: F = 1.0 sound casting F 1.0 contains pores: (1-F) times 100% by volume The si

46、ze and location of the porosity can be controlled by adjusting the values of the critical and coherency solid fractions.,凝固过程中孔隙度 : 微缩孔模型,Computes the formation of micro-pores when solid fraction exceeds the critical value, Fs FSCR. Takes into account the bulk modulus of metal: a - speed of sound in

47、 the solidifying metal, p0 intensification or initial pressure. When pressure drops below a critical value, porosity forms. Thus, pressurization of the metal can reduce the amount of porosity. Calculates and stores % porosity by volume.,new in 9.3,凝固过程中孔隙度 :设置微缩孔模型,Requires different solid and liqui

48、d phase density and a critical solid fraction less than 1.0,A380 die casting,凝固过程中孔隙度 : 微缩孔模型输出,Flow in Mushy Zone,Several options are available to model flow of semi-solid (mushy) metal. In the standard (default) model, flow is divided into three regimes: 0.0 fs FSCO - solid crystals float freely i

49、n the liquid phase (suspension flow); the mixture viscosity increases with fs. FSCO fs FSCR solid crystals form a rigid coherent structure, liquid phase seeps through it (porous media flow); a DArcy type drag coefficient is applied. FSCR fs 1.0 liquid does not form continuous flow regions; no flow on macroscopic scale: K .,Flow in Mushy Zone:凝固阻力系数,The coefficient of solidification drag, Ks, is a function of the dendrite arm spacing, DAS. It defaults to 1.0, which may be too small for many alloys. A value in the range o