Modelling for Solidification Defect

Search - Scholars Portal Journals

About | Contact | Mobile | Français

Search

Browse

Figures

Search for for::

in Anywhere Anywhere Add Rows >>

Refine Refin e By Date:

All OR

202007 07

to

PrPresent esent

Search Help

Search

Simulation-enabled casting product defect prediction in die casting process Internat Inte rnational ional J ourn ournal al of P rodu roducti ction on Research R esearch (Sept (S eptem ember ber 2009), 2009), 47 (18), pg. 520 5203-5216 3-5216

M.W. Fu; Fu; M.S. Yong

CAE CA E sim simulat ulation ion | high pressure die casting | product quality quality assurance | defect prediction

In current casting industries, product development development paradigm is shifting from traditional traditional trial-and-error trial- and-error in the workshop to CAE-enabled simulation and ‘proof-of-concept’ by computer. The product development paradigm shift is thus from heuristic know-how and experience to more scientific simulation, evaluation, analysis and calculation. CAE simulation plays an important role in the new product development paradigm as it models the entire casting process and reveals the dynamic behaviour of the casting system in working conditions. In addition, the product quality panorama and product defects are explored via simulation in such a way that the root-causes of casting defects are pinpointed and the solutions to avoid them can be determined. In this paper, the CAE technology for casting process simulation is presented. The modelling of the casting process is first articulated and the detailed simulation issues are then described. The information related to the filling and solidification process and after-ejection behaviour are revealed by CAE simulation. Through case studies, how the CAE simulation helps identify and predict the process-related defects is illustrated and its efficiency is finally validated validated..

Affiliation: 0001 a Department of Mechanical Engineering, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong 0002 b Singapore Institute of Manufacturing Technology, 71 Nanyang Drive, 638075 Singapore

Table of Content Content :

1. Introduction 2. Modelling of the casting process 3. Simulation of the casting process system 4. Behaviour and information revealed via CAE simulation 4.1 Filling F illing process process sim simulat ulation ion 4.2 Solidification process simulation 4.3 Str S tress ess analy analysis sis in castin casting g 5. Case studies 6. Conclusions


1. Introduction In today's casting product development, the product development paradigm is basically trial-and-error based on knowhow and experience. These know-how and experience, however, are normally linked to long years of apprenticeship and skilled craftsmanship. This product development paradigm appears to be more heuristic and experience-based than deep scientific simulation, evaluation, analysis, and calculation. It is thus time-consuming, error-prone, and needs a lot of experimental tryout and verification in the workshop for ‘proof-of-concept’. Currently, casting products, especially for high pressure die castings (HPDC), have been widely used in many industries due to its near-net shape or net shape characteristics, high productivity and complicated geometries and features. As the market demands for shorter design and manufacturing lead-times, good dimensional accuracy, overall product quality and rapid change of product design and process configuration are increasingly significant. How to meet these demands has been a bottleneck in casting production industries. The traditional product development paradigm is obviously handicapped in this competitive marketplace. To address these issues, an efficient product development paradigm supported by efficient enabling technologies is needed. Conventionally, CAD/CAM technologies, as the efficient enabling technologies for representation of design intent and solutions and helps realisation of design physically, provide an essential part of solutions to address the above issues. The technologies greatly enhance design quality and shorten design and manufacturing lead-times. However, it is difficult to address some critical issues in the design of the casting process, tooling structure, product properties configuration and finally the quality control and assurance by using CAD/CAM technologies alone. Computer-aided engineering (CAE) simulation technology, on the other hand, fills this gap as it helps practitioners generate, verify, validate and optimise the design solutions before they are practically implemented and physically realised. From product quality and defect prediction perspective, CAE simulation is a most technologically efficient and cost effective technology for analysis, prediction and evaluation of casting product quality and defects. In CAE simulation, the simulation is the representation of a physical system by models that imitate the dynamic behaviour of the system in working processes and conditions. The numerical simulation employs numerical methods such as finite element method (FEM) or finite difference method (FDM) to quantitatively represent the working behaviour of physical systems. The numerical results are correspondingly related to the physical content of the physical systems to be simulated. Taking the casting process as an instance, the fluid dynamics of the metal in cavity, the thermal phenomena and solid state transformation of the melt in filling and solidification processes need to be modelled by physical and mathematical models. The final simulation results will thus be related to the behaviour of the casting process and the properties of the casting products. From the production process perspective, the numerical simulation results will associate the structure, quality, property and defect issues of the products. This up-front process and casting system simulation is critical as 20% of design activities at the up-front design stage commits to about 80% of product cost and product quality issues (Fuh et al.2004 ). Furthermore, it is reported that about 90% of product defects are related to the mistakes made in the design stage and only 10% is due to manufacturing problems (Louvo 1997 ). In addition, it has also been calculated that the costs to change design in the up-front design process is ten times higher in the subsequent design and manufacturing processes. From the product quality assurance and control perspective, any technologies, which predict product quality and defects in the casting process to ensure ‘right design the first time’ and reduce trial-and-error in the workshop, will help cut product development cost and shorten time-to-market. CAE simulation technology is one of those technologies. Presently, the applications of CAE simulation technology in casting product development are monolithic. They are basically focused on casting design, process determination, flow pattern prediction, tooling design, quality control and product stress analysis. From the casting design perspective, CAE simulation helps casting design through filling simulation, solidification analysis, stress evaluation and optimisation of casting geometries and features (Sequeira et al.2001 , Sturm et al.2001 , McMillin et al.2002 ). Casting design is critical as it is the first step among the design activities and affects the subsequent design processes. From the process determination point of view, simulation helps determine process routing and process parameter configuration (Lewis and Ravindran 2000 , Midea et al.2000 , Cleary et al.2002 , Mirbagheri et al.2002 , Barriere et al.2003 , Naher et al.2003 , Hsu and Yu 2006 , Krimpenis et al.2006 ). It also helps verify the die design based on the revealed flow behaviour and solidification phenomena (Ulysse 1999 , Hu et al.2000 , Dai et al.2003 ). From the product quality control and assurance aspect, through revealing the filling and solidification behaviour related to product quality and defect forming mechanism, the simulation provides physical basis and useful information for product quality improvement and defect avoidance (Bird et al.1960 , Guo et al.2005 , Mochnacki et al.2005 , Monroe and Beckermann 2005 , Neumann et al.2005 , Peng et al.2005 , Venkatesan et al.2005 , Zhang et al.2006 ). These prior researches, however, are basically monolithic and did not systematically present a complete simulation-enabled product defect prediction paradigm and articulate how the CAE simulation helps defect prediction and avoidance. Furthermore, they also did not reveal what information can be revealed through CAE simulation; what kind of information is needed to identify


the corresponding defects; what defects may occur in filling and solidification processes; and how to come out with remedies for product defect avoidance based on the information provided by simulation. In this paper, the casting process is first modelled by numerical methods and the association between the real process, modelling, simulation and output variables are presented. The information flow in the simulation process and solution generation cycle is articulated. A process-based simulation paradigm for prediction of casting defects is proposed. The categorised information revealed by simulation and the defects categories, which can be predicted based on the output information of simulation, are presented. Through case studies, the simulation process, procedure and how the defect occurs in the casting process is analysed based on the identified simulation process are extensively described. 2. Modelling of the casting process To dynamically simulate the casting process, modelling is the first issue to be addressed. Modelling will represent the casting processes by models from physical and mathematical perspectives. From the mathematical aspect, models are formulated as governing equations and boundary conditions. For most engineering problems, the models are non-linear in terms of both the geometry and material properties of the casting systems to be modelled. Numerical methods are thus employed to convert the non-linear equations into simultaneous and algebraic equations. These equations further represent the physical relationship of the casting system in the form of action-behaviour-property relationship. In the HPDC process, the action is the high pressure generated by the fast movement of plunge in the chamber; the behaviour is the flow of the metal melt. The behaviour is then decided by the reheological properties of the melt. Figure 1 presents this relationship from physical and mathematical modelling points of view. In addition, Figure 2 further illustrates the association between the real process, simulation procedure, physical phenomena to be modelled, governing equations representing the specific physical behaviour, and the output variables. In real casting processes, materials and material properties, equipment and working parameters are the input to model the physical behaviour and phenomena of the casting process. The simulation results, on the other hand, reveal the physical information related to the casting process and the final microstructures, defects, quality and property of the casting. From the modelling perspective, there are three phenomena to be considered. They are mould filling, solidification and cooling, and stress and strain distribution of casting after ejection. Taking the modelling of filling process as an instance, there are three physical phenomena viz., melt momentum balance, mass balance and energy balance, to be represented and modelled. These phenomena are modelled by the following governing equations: Continuity equation (when T >Ts): (1) Momentum equation (Navier-Stokes equation, when T > Ts): (2) Energy Equation: (3) where:t –time, x

–space, ρ

–density, μ

–viscosity, g

–gravity, Cp

–heat capacity,


λ

–conductivity, U

–velocity, T

–temperature, Ts

–solidus temperature, and Q

–heat source. For the open surfaces, a volume of fluid function (VOF), defined as a ratio of metal melt to actual volume, is used to track the moving free surface of the metal melt. The VOF function is governed by the equation in the following: (4) The governing Equations (1–4 ) are non-linear in terms of the geometry of melt flow path and the material properties of the melt. It needs numerical methods to linearise and discretise and then a set of simultaneous and algebraic equations are obtained. Through solving the linearised equations, the velocity, pressure and temperature of the melt are obtained. For solidification modelling, the Fourier heat conduction equation is used. Phase transformation enthalpies like melt heat are considered. Through modelling of heat balance in the solidification process, the temperature distribution in casting is determined and the solidification behaviour is identified. To determine the casting stress and strain, the equilibrium equation and Hooke's law are employed for representing the relationship between displacement, stress and strain. Through solving these equations, the displacement, stress and strain are determined. 3. Simulation of the casting process system To reveal the casting defects, the simulation of the entire casting system is needed. Figure 3 presents a process-based simulation framework for prediction of casting defects. From the figure, it can be seen that the casting product design is based on the voice of customers (VoC), product functional requirements and design specifications. The product geometry is used as the basis for subsequent casting geometry design, casting process determination and process parameter configuration. The whole casting system is then determined after the die design is fixed. Therefore, the whole casting system is formed through the product and casting design, process determination, die design, casting equipment selection and working parameter configuration. In this research, the casting system is first modelled by way of establishment of physical, mathematical and numerical models of the system and then input into a CAE simulation system for simulation. In the CAE modelling process, the physical model idealises the real engineering problems and abstracts them to comply with certain physical theory with assumptions. The mathematical model specifies the mathematical equations such as the differential equations in FEM analysis the physical model should follow. It also details the boundary and initial conditions and constraints. The numerical model describes the element types, mesh density and solution parameters. The solution parameters further provide detailed calculation tolerances, error bounds, iteration specifications and convergence criteria. Currently, most CAE simulation packages have part of the built-in content of these models, but users still need to prepare and input most of the model information into the CAE simulation systems. The information includes CAD STL models, element types, mesh density and number, material and the casting process related parameters, including pouring, liquidus and solidus temperatures etc. Upon the completion of simulation, the filling-, solidification-, thermal-, and property and quality-related data and information are available for analysis and identification of any potential product quality issues and defects in the casting. The defects could be caused by the irrational design of process, tooling, and product/casting. They are thus called process-related, tooling-related, product/casting-related defects. For any product quality and defect issues, the design needs to be changed. With the new or modified design, new simulation is needed until the defect-free casting is obtained. 4. Behaviour and information revealed via CAE simulation As stated in the previous section, CAE simulation of the entire casting system reveals filling and solidification behaviour


filling, solidification and after-ejection behaviour affect the product quality and defect formation significantly, the information revealed by CAE simulation related to this behaviour is summarised in the following subsections. 4.1 Filling process simulation In the filling process, through modelling and simulation of the momentum, mass and energy balances of the metal melt, the melt velocity, flow direction, pressure and temperature are determined. These parameters reveal the physical behaviour and thermal behaviour in the filling process and thus the whole panorama of the filling behaviour is figured out: Melt-front positions, turbulence in the melt movement and filling smoothness; Air entrapment in the casting and cavity and identification of whether the venting conditions are needed; Thermal behaviour and temperature distribution in the casting; Filling sequence in the casting and determination of overflow and venting locations; Melt velocity distribution and its relationship to die erosion in the filling process; Surface finishing prediction based on velocity distribution, especially in HPDC casting process; Undesirable/irrational filling behaviour such as split melt stream, misruns and cold shut etc. in the filling process. Through filling process simulation, it is revealed whether the filling pattern is reasonable and the quality casting is ensured. In addition, the solutions for design improvement and casting quality enhancement can be developed based on the simulation results. 4.2 Solidification process simulation In the solidification process, the physical, thermal and metallurgical phenomena concurrently occur in the process. The interaction and interplay of these phenomena are simultaneous. It is thus difficult, if not impossible, to reveal the individual phenomenon by using traditional technologies. The CAE simulation technology, however, provides an approach to revealing these phenomena in process conditions and identifying the needed information to evaluate and predict product quality and product defects in the solidification process. Through modelling and simulation of the solidification process, the entire solidification behaviour and the following information are presented: The last solidification area and the location of ingate; Solidification sequence and the temperature distribution in the casting and die; Rationality of cooling layout and design; Runner system design; Hotspot in the casting; Shrinkage and porosity distribution in the casting and the solutions for uniform-shrinkage and porosity-free casting; Feeder size and its location definition. 4.3 Stress analysis in casting After the solidification of casting, the casting is ejected from die and some physical phenomena related to this process include: Temperature distribution in the casting and die after the ejection of casting; Distortion of the casting; Residual stress in the casting and its distribution; Product quality and defect existing; Optimised casting system design such as feeding system design;


Die stress analysis and thermal deformation prediction and optimal design. The stress and strain in casting and die are important as they affect the casting geometry accuracy, casting product function and service, and die fatigue life. The accurate determination of stress and strain helps generate solutions for casting and tooling quality improvement.

5. Case studies To illustrate how the filling and solidification process simulation helps identify casting defects and reveal the root-causes of the defects, two industrial cases of HPDC are presented. The two cases are both the four-cavity HPDC process. Figure 4 shows the melt flow path and the layout of the casting system of the first case. The CAD models as shown in Figure 4 are created in Unigraphics, a commercial CAD/CAM system for product design and development, and then converted into STL format through CAD model data exchange. The generated STL CAD models are directly imported into the casting CAE simulation systems for filling and solidification process simulation. In this case, the cast material is AlSi9Cu3, a widely-used die cast material. The die material is X38CrMoV5. The CAE simulation is Magmasoft, which is a popular and commercial casting simulation system in industries and academia. The element type is cuboid and the cast is meshed into 1.5 million elements. In this case, the pouring temperature of the melt is 670°C. The liquidus and solidus temperatures are 578°C and 479°C, respectively. All the die components have an initial temperature of 150°C. Five cycles of simulation are conducted to reach a stable condition of the simulation and in such a way that the simulation outcomes are reliable. It takes 11 hours to complete the five-cycle simulation. In the figure, the melt is shot from the chamber driven by the plunge movement and goes into the inlet and biscuit. The melt flows along the runner and enters into the cavity where the casting is moulded. The ingate shown in the figure is the gate between the runner and casting cavity. The melt flow velocity at the ingate (40–100 m/s) is an important parameter as it significantly affects the filling behaviour and casting quality. In this case, the velocity at the ingate is set to be 40 m/s. After the cavity is filled up, the extra melt, dirty metal and the air in the melt go into the overflow portion. The material in overflow is trimmed away after the casting is ejected. Figure 5 presents the filling process and the position of melt-front advancement (MFA) during the filling process in Case I. The MFA describes the movement status of the melt flow and the arrival sequence in the filling process. It also shows the melt position for the given percentage of the filling. The MFA reveals the flow phenomenon behaviour and the defects related to the imbalanced MFA can be identified. Through analysis of MFA, the defects can be avoided via the rational design of the MFA in the entire filling process. In addition, the MFA also indicates the last area to be filled up. The last area is usually the location of overflow which is the container of dirty melt and air. Figure 5(a) shows the filling status at 85% where most of the casing is filled up. But the two boss features behind are not filled. That would mean that they are the last portions to be filled up and air entrapment could exist there. Figure 5(b) presents 90% filling of the casting. It is found that the last area to be filled up in the casting is the right-side boss feature. Air entrapment exists in the boss feature as the air is enclosed already before the melt started to flow into the overflow. The air entrapment could be the cause of not filling the boss feature and further becomes defect. To solve the problems, there are two approaches. One is to employ an air venting mechanism on the top of the boss feature. The other is to arrange an overflow there. But the arrangement of overflow there needs to shift the parting line location to this place. For this case, it is not feasible. Figure 6 presents another case, viz., case II, of the filling and solidification process simulation. The materials of die and cast are the same as the previous case but the mesh density and the simulation time needed are different. In this case, the meshed elements are 1.6 million and it takes about 12 hours to complete the five-cycle of simulation. The simulation reveals the melt flow behaviour and the defects caused by the unreasonable flow pattern and irrational temperature distribution during the solidification process. Figure 6 first presents the flow path and the layout of the casting system, which includes inlet, biscuit, runner, ingate, casting and overflow. Similar to Case I, Case II is also a four-cavity HPDC casting. From the figure, it is found that there are two overflows in each cavity. The design of the overflow is in such a way that the two-overflow locations are located at the last filling places, which is consistent with the simulation findings, as shown in Figure 7 . In Figure 7 , there are three last filling areas. Two of them are selected as the overflow location. The third one is not selected as any overflow location or air venting mechanism location. Potential defects could exist there. Figure 8 further articulates the issue for the third last filling place shown in Figure 7 . From the figure, it is found that the melt stream is divided into two by the cores for moulding the arc-slot and central hole, as shown in Figure 8(a) . The two melt streams push the air into the centre of the flange and finally there is air entrapment in the centre of the flange. Since there is no efficient venting mechanism and the entrapment air blocks the melt flow and finally there is a concave feature in the flange, as shown in Figure 8(b) . To avoid the defect, a change of flow pattern is necessary. The melt flow needs to reverse this direction. The ‘big’ arrow as shown in the figure illustrates that if the melt flow is from this direction, there will be no two streams facing each other and no air entrapment in this area. To realise this idea, a ‘bridge’


create an air venting hole in the centre of the flange where the concave feature defect is located. Figure 9(a) presents the solidification result of case II predicted by CAE simulation. It shows the hottest area and thus it is the last solidification area in the cast. On the other hand, a fracture defect is formed in the real part. In addition, a defect called ‘orange skin’ is found in the real part. From the figure, it is seen that the hottest area has more than 500°C. As the solidus temperature of the material is only 479°C, this would mean that there is no melt feeding for the last solidification area when it is solidified. When the melt is fully solidified, the volume contraction from liquid to solid occurs in such a way that the product surfaces have many small concave features, which are the so-called ‘orange skin’. If the volume contraction is big, the fracture may happen, as shown in Figure 9(b) . To void this uneven temperature distribution in the solidification process, a better designed cooling system is needed. In addition, a local cooling mechanism such as spot cooling could absorb the ‘extra heat’ locally and make the melt solidify at the same pace with its surrounding areas. In this way, the solidification related defects can be avoided. 6. Conclusions The design and development of a casting product, process and system is a trial-and-error process based on heuristic know-how in current casting industries. The solution generated in such a way lacks scientific calculation and analysis. Since the casting process is very complicated in terms of casting geometry and the physical, metallurgical and thermal behaviour involved in the casting process, this kind of process determination needs a lot of tryout in the workshop. The product quality and defects are difficult to predict in the up-front design process. The simulation-enabled casting quality prediction and defect evaluation reduces trial-and-error in the workshop as the process is virtually realised and verified by computer and any quality issue can be pinpointed and the related solutions can be proposed. In this paper, the panorama of simulation-enabled casting quality and defect analysis and prediction were articulated. The framework, procedure and process of simulation-enabled casting defect prediction were presented. Through industrial case studies, the efficiency was verified and validated.

Figure 1. Modelling of the casting process from physical and mathematical perspectives: action-behaviour-property relationship.

Figure 2. Association among the process, modelling, simulation and output variables.

Figure 3. The process-based simulation paradigm for prediction of casting defects.


Figure 4. The layout of the casting filling system in Case I.

Figure 5. The MFA positions in the filling process. (a) Filling at 85% of the whole process; (b) filling at 90% of the whole process.


Figure 6. The layout of the casting filling system in Case II.

Figure 7. The melt-front advancement position in filling process.


Figure 8. Unreasonable filling pattern and the defect caused. (a) Simulation predicted irrational flow; (b) the real defect revealed by experiment.

Figure 9. Defect caused by the unreasonable temperature distribution in the solidification process. (a) Simulation result; (b) real part.

Acknowledgements

The authors would like to thank the Hong Kong Polytechnic University for the research grant G-YF67 to support this research.

Reference:


0001 .Determination of the optimal process parameters in metal injection molding from experiments and numerical modeling Journal of Materials Processing Technology , v.143–144 , p.636 , 2003 , Barriere T et al. 0002.Transport Phenomena, 1960 , J . Wiley & Sons , Bird RB et al. 0003 .Flow modeling in casting processes Applied Mathematical Modeling , v.26 , p.171 , 2002 , Cleary P 0004 .Effects of runner system design on the mechanical strength of Al-7Si-Mg alloy castings Materials Science and Engineering: A , v.354 (1) , p.315 , 2003 , Dai X 0005.Computer-aided Injection Mould Design and Manufacture, 2004 , Marcel Dekker , Fuh J YH 0006 .Modelling of materials properties and behavior critical to casting simulation Materials Science and Engineering: A , v.413–414 , p.465 , 2005 , Guo Z 0007 .Computer simulation of casting process of aluminum wheels – a case study Proceedings of the I MECH E Part B Journal of Engineering Manufacture , v.220 (2) , p.203 , 2006 , Hsu Y L et al. 0008 .Design and optimization of runner and gating systems for the die casting of thin-walled magnesium telecommunication parts through numerical simulation Journal of Materials Processing Technology , v.105 (1) , p.128 , 2000 , Hu BH 0009 .Simulation-based selection of optimum pressure die-casting process parameters using neural nets and genetic algorithms International Journal of Advanced Manufanuring Technology , v.27 (5–6) , p.509 , 2006 , Krimpenis A 0010 .Finite element simulation of metal casting International Journal for Numerical Methods in Engineering , v.47 (1–3) , p.29 , 2000 , Lewis RW et al. 0011.Louvo A 1997 Casting simulation as a tool in concurrent engineering In: International ADI and simulation conference 1997 Espoo. 28–30 May Finland 0012 .CAE opens new frontier in casting design Engineered Casting Solutions , v.Spring 2002 , p.29 , 2002 , McMillin T et al. 0013 .Using computer modeling to optimize casting processes Modern Casting , v.90 , p.4 , 2000 , Midea A 0014 .Simulation of melt flow in coated mould cavity in the casting process Journal of Materials Processing Technology , v.142 (2) , p.493 , 2002 , Mirbagheri SMH 0015 .Multiscale model of segregation process Journal of Materials Processing Technology , v.162–163, May 2005 , p.766 , 2005 , Mochnacki B et al. 0016 .Development of a hot tear indicator for steel castings Materials Science and Engineering: A , v.413–414 , p.30 , 2005 , Monroe C et al. 0017 .Simulation of the stir casting process Journal of Materials Processing Technology , v.143–144 , p.567 , 2003 , Naher S et al. 0018 .Prediction of texture induced anisotropy by through-process modelling Materials Science Forum , v.495–497 , p.1657 , 2005 , Neumann L 0019 .Numerical study on the low pressure die casting of AZ91D wheel hub Materials Science Forum , v.488–489 , p.393 , 2005 , Peng YH 0020.Optimization of die casting part design, process parameters and process control using newly die casting simulation tool, proceedings of die casting in the 21th century, 2001 , Sequeira W 0021.Optimized development for castings and casting processes, 19th CAD-FEM users' meeting 2001, international congress on FEM technology, 2001 , Sturm J C


0022 .Optimal extrusion die design to achieve flow balance International Journal of Machine Tools and Manufacture , v.39 (7) , p.1047 , 1999 , Ulysse P 0023 .Simulation of casting solidification and its grain structure prediction using FEM Journal of Materials Processing Technology , v.168 (1) , p.10 , 2005 , Venkatesan A et al. 0024 .Numerical methods to improve the computational efficiency of solidification simulation for the investment casting process Journal of Materials Processing Technology , v.173 (1) , p.70 , 2006 , Zhang XP et al.

©2010 Ontario Council of University Libraries. All rights reserved. Terms of Use | Privacy Stat ement | Contact Support

Modelling for Solidification Defect

Recommend Documents