Modelling and simulation of die casting process in aluminium alloys

Universidade do Minho Escola de Engenharia Nélson Moura Pereira Duro Modelling and Simulation of Die Casting Process on Aluminium Alloys October 2024 Universidade do Minho Escola de Engenharia Nelson Moura Pereira Duro Modelling and Simulation of Die Casting Process in Aluminium Alloys Master’s Dissertation Integrated Master’s Degree in Mechanical Engineering Specialization in Advanced Manufacturing Work supervised by. Hélder Puga Inês Varela Gomes October 2024 I DIREITOS DE AUTOR E CONDIÇÕES DE UTILIZAÇÃO DO TRABALHO POR TERCEIROS Este é um trabalho académico que pode ser utilizado por terceiros desde que respeitadas as regras e boas práticas internacionalmente aceites, no que concerne aos direitos de autor e direitos conexos. Assim, o presente trabalho pode ser utilizado nos termos previstos na licença abaixo indicada. Caso o utilizador necessite de permissão para poder fazer um uso do trabalho em condições não previstas no licenciamento indicado, deverá contactar o autor, através do RepositóriUM da Universidade do Minho. Licença concedida aos utilizadores deste trabalho Atribuição-NãoComercial-SemDerivações CC BY-NC-ND https://creativecommons.org/licenses/by-nc-nd/4.0/ II State of integrity I hereby declare having conducted this academic work with integrity. I confirm that I have not used plagiarism or any form of undue use of information or falsification of results along the process leading to its elaboration. I further declare that I have fully acknowledged the Code of Ethical Conduct of the University of Minho. Acknowledgment -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- III ACKNOWLEDGMENT I would like to express my deepest gratitude to my family and friends for their unwavering support, patience, and encouragement throughout the duration of my studies and during the preparation of this dissertation. I am profoundly thankful to my mentors, Prof. Hélder Puga and Prof. Inês Gomes, for their guidance, trust in my capabilities, and for challenging me to push beyond my limits. Your insightful advice and the knowledge you imparted were crucial to my success over the past year, and I am deeply grateful for your mentorship. I extend my thanks to Eng. Igor Villalobos from the ESI Group for his support, providing vital information and recommendations on the most efficient ways to utilize QuikCast. I would also like to thank Eng. Sérgio from Donelab for his help and expertise in design for manufacturing of the inserts. Your assistance was fundamental to the completion of this work. Finally, I am grateful to everyone who, in one way or another, contributed to the realization of this dissertation. Your encouragement and support made this journey not only possible but truly meaningful. This work was carried out within the framework of the “Agendas para a Inovação Empresarial” (Project nº 49, acronym “INOV.AM”, with reference PRR/49/INOV.AM/EE, operation code 02/C05i01.01/2022.PC644865234-00000004), supported by the RRP - Recovery and Resilience Plan and by the European Funds NextGeneration EU. http://www.recuperarportugal.gov.pt/ Abstract -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- IV ABSTRACT In today's modern world, the need for increased production to meet demand is ever-present. However, due to time constraints, the primary focus is on boosting productivity and minimizing timeconsuming tasks. By doing so, better solutions can be found, or the intrinsic costs of processes can be reduced. The Die Casting technology is no exception to this requirments. The time the engineers spent in the development and validation of gating design remains one of the bottlenecks of the process. Therefore, a solution must be found in order to improve the modelling and simulation of the die casting process. This dissertation explores the potential of using Python as a programming tool for the gating design in die casting manufacturing. The knowledge gained has been compiled into software named Highly Efficient Labor-saving Program Die Casting. This software can accommodate various casting alloys, it emphasizes aluminum alloys. The software was applied at two stages of gating design validation. First, it was used to generate the gating design, overflow, and venting system for a valve cover, along with the corresponding dimensional data for each section. Second, it was employed during the die casting process simulation to calculate the boundary conditions and initial conditions. The validation of the gating designs focused on three key aspects: die cycle simulation to ensure a stable thermal gradient, analysis of potential defects (such as shrinkage porosity, air entrainment, and cold shuts), and an evaluation of the process dynamics (based on the minor and major losses of the molten metal). In the process dynamics analysis, two distinct phases were identified where the losses behaved differently. Keywords: Programming aided design; High-pressure die casting process; Aluminum alloys; Finite differences method Resumo -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- V RESUMO No mundo moderno de hoje, a necessidade de aumentar a produção para satisfazer a procura é constante. No entanto, devido a restrições de tempo, o foco principal está em aumentar a produtividade e minimizar o tempo despendido tarefas demoradas. Desta forma, podem ser encontradas melhores soluções ou os custos intrínsecos dos processos podem ser reduzidos. A tecnologia de fundição em alta pressão ( Die Casting ) não é exceção a estes requisitos. O tempo que os engenheiros gastam no desenvolvimento e validação do design do sistema de enchimento continua a ser um dos estrangulamentos do processo. Portanto, é necessário encontrar uma solução para melhorar a modelação e simulação do processo de fundição injetada. Esta dissertação explora o potencial de utilizar Python como ferramenta de programação para o design do sistema de enchimento no processo de fundição injetada. O conhecimento adquirido foi compilado num software denominado Highly Efficient Labor-saving Program Die Casting . Embora este software engloba várias ligas de fundição, é dada ênfase às ligas de alumínio. O software foi aplicado em duas fases da validação do design dos canais de enchimento. Primeiro, foi utilizado para gerar o design do sistema de enchimento, dos overflows e do sistema de ventilação de uma tampa de válvula, juntamente com os dados dimensionais correspondentes a cada secção. Em segundo, foi utilizado durante a simulação do processo de fundição injetada para o cálculo das condições de fronteira e condições iniciais. A validação dos designs do sistema de enchimento concentrou-se em três aspetos principais: simulação do ciclo do molde para garantir um gradiente térmico estável, análise de possíveis defeitos (como rechupe, aprisionamento de ar e frentes frias), e uma avaliação da dinâmica do processo (com base nas perdas localizadas e distribuidas do metal fundido). Na análise da dinâmica do processo, foram identificadas duas fases distintas onde as perdas se comportaram de forma diferente. Keywords: Modelação assistida por programas; Processo de fundição injetada; Ligas de alumínio; Método das diferenças finitas VI INDEX ACKNOWLEDGMENT ............................................................................................................. III ABSTRACT .......................................................................................................................... IV RESUMO ............................................................................................................................. V INDEX................................................................................................................................ VI FIGURE INDEX ................................................................................................................... VIII TABLE INDEX ...................................................................................................................... XII LIST OF ACRONYMS ............................................................................................................ XIII LIST OF SYMBOLS .............................................................................................................. XIV CHAPTER 1 - INTRODUCTION ................................................................................................... 1 1.1. Motivation ............................................................................................................. 2 1.2. Aim ....................................................................................................................... 2 1.3. Structure of the dissertation ................................................................................... 3 1.4. Chapter references ................................................................................................ 4 CHAPTER 2 - DIE CASTING TECHNOLOGY .................................................................................... 5 2.1. Equipment and their elements ............................................................................... 6 2.2. Materials of the process ...................................................................................... 10 2.3. Variables of the process ....................................................................................... 10 2.4. Virtualization of the high-pressure die casting process .......................................... 14 2.5. Chapter references .............................................................................................. 20 CHAPTER 3 - HIGHLY EFFICIENT LABOUR SAVING PROGRAM GATING DESIGN ................................... 27 3.1. System requirements .......................................................................................... 29 3.2. Interface User - Software ..................................................................................... 32 3.3. Workflow ............................................................................................................. 35 List of Acronyms -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- XIII LIST OF ACRONYMS CAD Computer-Aided Design CAM Computer-Aided Manufacturing CAE Computer-Aided Engineering DIN German Institute of Standardization FDM Finite Difference Method FEM Finite Element Method FVM Finite Volume Method H.E.L.P. Highly Efficient Labor-Saving Program HPDC High Pressure Die Casting IGES Initial Graphics Exchange Specification ISO Internation Standard Organization NADCA North American Die Casting Association STEP Standard for the Exchange of Product model data STL Stereolithography VBA Virtual Basic for Applications VCF Volume Correction Factors List of Symbols -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- XIV LIST OF SYMBOLS Symbol Description Units 𝒗𝒑𝒍𝒖𝒏𝒈𝒆𝒓,𝟏𝒔𝒕 Plunger’s velocity in the first stage [m/s] 𝒅𝒑𝒖𝒏𝒈𝒍𝒆𝒓 Plunger’s diameter [mm] %𝒇𝒊𝒍𝒍𝒊𝒏𝒈 Shot sleeve’s filling ratio [%] 𝑳𝒑𝒍𝒖𝒏𝒈𝒆𝒓,𝟏𝒔𝒕 Plunger’s displacement in the first stage [mm] 𝑽𝒑𝒂𝒓𝒕 Part’s volume [mm3] 𝑽𝑶𝒗𝒆𝒓𝒇𝒍𝒐𝒘𝒔 Overflow’s volume [mm3] 𝑨𝒑𝒍𝒖𝒏𝒈𝒆𝒓 Plunger’s area [mm2] 𝒗𝒑𝒍𝒖𝒏𝒈𝒆𝒓,𝟐𝒏𝒅 Plunger’s velocity in the second stage [m/s] 𝑳𝒑𝒍𝒖𝒏𝒈𝒆𝒓,𝟐𝒏𝒅 Plunger’s displacement in the second stage [mm] 𝒗𝒊𝒏𝒈𝒂𝒕𝒆 Velocity at ingate [m/s] 𝑨𝒆𝒇𝒇,𝒊𝒏𝒈 Effective ingate area [mm2] 𝑪𝒅 Loss coefficient [%] 𝒕𝒇𝒊𝒍𝒍 Filling time [s] 𝑲𝟏 Constant based on die’s material [s/mm] 𝑻𝒑𝒂𝒓𝒕 Avarage thickness [mm] 𝑻𝒊 Injection temperature [ºC] 𝑻𝒇 Minimum flow temperature [ºC] 𝑻𝒅 Die’s temperature [ºC] 𝑺 Percentage of solidified metal at the ending of the filling [%] 𝒁 Solids units conversion factor [ºC/%] 𝑨𝒂𝒄,𝒊𝒏𝒈 Actual ingate area [mm2] 𝜽𝒆𝒇 Flow angle [º] 𝑻𝒂𝒄,𝒊𝒏𝒈 Actual ingate thickness [mm] 𝑱 Constant for atomized flow [-] 𝝆 Density [kg/m3] List of Symbols -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- XV 𝒓𝒊𝒏𝒈,𝒓𝒖𝒏 Ingate-runner ratio [-] 𝑨𝒗𝒆𝒏𝒕 Venting area [mm2] 𝑲 Thermal conductivity [W/mK] 𝑪𝒑 Specific heat [J/kgK] 𝒕 Time [s] x, y, z Cartesian coordinate system [-] u, v, w Component x, y and z of the velocity vector [m/s] 𝒑𝒙, 𝒑𝒚, 𝒑𝒛 Component x, y and z of the pressure vector [Pa] 𝝁 Dynamic viscosity [Pa s] 𝛁,𝛁𝟐 Nabla operator, Laplace operator [-] 𝑭𝒙, 𝑭𝒚, 𝑭𝒛 Component x, y and z of the External force vector [Pa] 𝒎 Mass [kg] 𝑸𝒊𝒏, 𝑸𝒐𝒖𝒕 Flow inward and outward [kg/s] 𝑸 Heat transfer [W] 𝑻 Temperature [ºC] 𝒉 Heat transfer convection coeffcient [W/m2K] 𝑨 Surface area [m3] 𝑻∞ Temperature of the fluid out of the boundary layer [ºC] 𝝈 Stefan-Boltzmann constant [W/m2K4] 𝜺 Emissivity [-] 𝑭𝒈 Geometric factor [-] 𝑻𝟏, 𝑻𝟐 Temperature of two different surfaces [K] 𝒈 Gravity accelaration [m2/s] 𝜷 Coefficient of volume expansion [1/ºC] 𝑳 Characteristic length [m] 𝑮𝒓 Grashof number [-] 𝑷𝒓 Prandtl number [-] 𝑹𝒂 Rayleigh number [-] 𝑵𝒖 Nusselt number [-] 𝑭𝒍𝒐𝒄𝒌𝒊𝒏𝒈 Locking force [N] 𝒑𝟑𝒓𝒅 Third stage pressure [Pa] List of Symbols -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- XVI 𝑨𝒑𝒓𝒐𝒋𝒆𝒄𝒕𝒆𝒅 Projected area [mm2] 𝒑𝒅𝒊𝒆       Average pressure on the die surface [Pa] 𝑭𝒎𝒍𝒐𝒄𝒌𝒊𝒏𝒈 Maximum locking force [N] 𝑨𝒄𝒐𝒏𝒕𝒂𝒄𝒕 Contact area [mm2] 𝑪 Numerical identification of each cell's volume [-] Chapter 1 – Introduction -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 1 Chapter 1 - INTRODUCTION The modern tools for the development of new products englobes CAD (Computer-Aided Design), CAE (Computer-Aided Engineering), and CAM (Computer-Aided Manufacturing) softwares. In the CAD software, the virtual model is conceived. CAE software plays a crucial role in validating these designs through the analysis and post-processing of simulation results. The CAM branch is directioned for the manufacturing of the product, it will not be addressed in this thesis. In the foundry industry, the usage of the CAD/CAE technological advances has been remarkable. The implementation of casting simulation has significantly reduced costs, time, labor, shopfloor trials and the number of prototypes to test. Furthermore, it is possible to optimise designs and processes without the creating the all the prototypes of each iteration step. Therefore, the development of new methodology to improve the CAD/CAE environments in HighPressure Die Casting (HPDC) and reduce gating design time is necessary. The proposed methodology is grounded in the existing guidelines outlined in the North American Die Casting Association (NADCA) manual. This will increase the productivity and the accuracy of the solution and a reduction in high qualified labour costs. Nevertheless, this new methodology will not entirely replace the expertise required for HPDC processes, particularly in gating design and casting simulation. Instead, it serves as tool to complement and enhance the knowledge and skills of industry professionals. Chapter 1 – Introduction -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 2 1.1. MOTIVATION The beginning of the casting simulation starts in the CAD software in the gating design’s and mold’s development. Then the CAE environment for the casting simulation which validates the gating design and process. This type of CAE software are already widely used in the foundry industry which simulates numerically the filling and solification processes and locate possible defects [1, 2]. These two stage reduce the number of shop-floor trials, time and costs, increasing performance and productivity. The old bottleneck of the development of molds in die casting process, which was trial and error with different molds, was turned into the usage of CAD/CAE softwares. Even though it was an improvement, these two stages still requires qualified manpower and time [3]. Thereby, one way to improve the efficiency of the process, time, cost and manpower is to improve the modelling and simulation of the die casting process. This improvement can be achived increasing the virtualization of this bottleneck. The digital virtualization of the modelling process is an underdeveloped branch with the following benefits: I. Reduced gating design time: modelling the gating design is a time consuming task for a highly qualified manpower and increases productivity; II. Reliable solution: the methodology is always the same unless the program is modified III. Storage of data: if the generated design is not accordingly with the user’s desire, the generated data is kept stored for a later modification; The virtualization of the die casting simulation is already in development, with some softwares already possessing some sort of Macros which automates through programming the user’s tasks. 1.2. AIM This dissertation thesis aims to develop and use the power of the programming in the creation of a gating design in HPDC process in aluminum alloys and use it to generate a gating design for the valve cover. This dissertations will explore essential topics for the virtualization of the gating design for HPDC using the North American Die Casting Association (NADCA) gating manual [4] and other manuals [5–7]. Part of the gating design formulation will be translated into Python code for the programmed modelling. Not only the formulation of the gating design will be explored but all the variables to its design and fluid’s mechanics principles of its dynamic. Chapter 1 – Introduction -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 3 At last, the numeric simulation of the die cycling and casting process will be done using QuikCast from ESI Group to predict defects. Properties such as density, specific heat/entalphy, thermal conductivity and solidus and liquidus temperature will be determined for the aluminum alloy which are required to the simulation.This analysis will include the evaluation of the process’s dynamics and the defects within the casting. A comparison between different gating designs will be made and compare it the final solution. 1.3. STRUCTURE OF THE DISSERTATION In the first chapter, the motivation and objective of the dissertation are presented, along with the context of the problem, which involves the modeling and simulation of the high-pressure die casting process for aluminum alloys. In chapter 2, the high-pressure die casting process is discussed, as well as the process’s variables and the NADCA methodology for developing gating systems for die casting. In chapter 3, the interface of the software developed for the modeling of the gating system and overflows is introduced. This software was designed in Python and uses the Autodesk Inventor API for automated modeling. The software also allows data to be stored in an Excel sheet format, which is later used in the simulation of the casting process. In the fourth chapter, the preprocessing for the validation of the gating system is detailed, using the QuikCast simulation software. The inserts contain the mold cavity and cooling channels, which will be manufactured via additive manufacturing. This chapter also addresses the thermophysical properties of materials relevant to the simulation of the AlSi9Cu3 alloy and 1.2709 steel. The validation of the gating systems will consist of two simulations: thermal cycles and filling. The thermal cycle simulation aims to achieve a stable thermal gradient for the filling simulation. To reduce computation time, only a single solidification scenario will be considered in the thermal cycle simulation. In the fifth chapter, three solutions will be evaluated: the final solution, solution A, and solution B. This evaluation will include the analysis of the presence of shrinkage (and the reasoning for its location), air porosity, and cold fronts (if the defect occurs). The preprocessing presented in the fourth chapter refers exclusively to the final solution. The differences between the solutions will be described in the corresponding chapter, noting that these changes do not affect the performance of the gating system. In chapter 6, the conclusions of the work are presented, along with proposals for future developments of the software. Chapter 1 – Introduction -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 4 1.4. CHAPTER REFERENCES [1] E. Flender and J. Sturm, Thirty Years of Casting Process Simulation. [Online]. Available: https:// link.springer.com/article/10.1007/BF03355463 (accessed: Sep. 23 2024). [2] H.-J. Kwon and H.-K. Kwon, "Computer aided engineering (CAE) simulation for the design optimization of gate system on high pressure die casting (HPDC) process," Robotics and Computer-Integrated Manufacturing , vol. 55, pp. 147–153, 2019, doi: 10.1016/j.rcim.2018.01.003. [3] B. Ravi, "Casting Simulation and Optimisation: Benefits, Bottlenecks and Best Practices," ReseachGate , pp. 1–2, 2008. [Online]. Available: https://www.researchgate.net/publication/ 228975218 [4] NADCA, NADCA Gating Manual . Illinois, United States of America: NADCA. [5] H. Bakemeyer, Operating the die casting: Machine. [Online]. Available: https://www.dykast.com/ user/files/operating_the_die_cast_machine.pdf (accessed: Sep. 15 2024). [6] UPMOLD, Casting Technique . Estados Unidos da América: UPMOLD. Accessed: Jan. 5 2024. [Online]. Available: http://www.upmold.com/ [7] NADCA, NADCA Product Specification Standards for Die Casting: Aluminum, Aluminum-MMC, Copper, Magnesium, Zinc and ZA Alloys, 9th ed. Illinois, United States of America: NADCA, 2015. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 5 Chapter 2 - DIE CASTING TECHNOLOGY The casting manufacturing process allows for the production of parts with near-final geometries, also known as near net shape. Currently, the two predominant casting processes in the market are sand casting and permanent mold casting [1]. In permanent mold casting, the mould material is steel, and consequently, it cannot be destroyed to remove the part from its interior compared to sand casting [2, 3]. Permanent mold casting has three branches: (i) low pressure, (ii) gravity, and (iii) high pressure [4]. In low-pressure casting, the molten metal is directed into the mold cavity by the pressure exerted on it, typically ranging from 0.1 to 1 bar [5]. In gravity casting, the molten metal is poured into the mold cavity, and gravity forces the cavity to fill. High-pressure casting is characterized by the application of higher pressures (up to 1200 bar) [6, 7] by a piston to push the molten metal into the die cavity. Consequently, the molds and the clamping mechanism used in high-pressure casting must withstand higher pressure levels. At the beginning of the high—pressure die casting process, it is sprayed die lubricant in the die surface and then the hydraulic system closes the mold. Meanwhile, the metal is heated up until it reaches the pouring temperature. Then, the metal poured into the shot sleeve (if it is a hot chamber machine the shot sleeve is inside the furnace) [8]. In the shot sleeve, the molten metal is pushed by a plunger at high pressure into the die cavity which possess the negative of the casting. At last, the casting alloy solidifies under high pressure applied by the plunger [9]. The solidification process is controlled by the cooling fluid (for example water or oil) passes through the cooling channels [10]. Lately, the hydraulic press opens the mold and the casting (Fig 2.1) is removed from the die cavity. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 6 The casting consists of the biscuit, runners, part and overflows. The biscuit dimensions are based on the shot sleeve diameter and typically have a thickness of 20 mm [11]. The runners’ and overflows’ cross sections are calculated based on the Bernoulli’s equation [6]. These parts are cut-off in the postprocessing of the casting [12]. Although the Bernoulli’s equation is used for determining the dimensions of the runners [13], these are verified in CAE software whose rulling equations are the NavierStokes’s. Fig 2.1 – Model constituted by the different elements. 1 – Biscuit; 2 – Runner; 3 – Ingate; 4 – Component; 5 – Overflow (adapted from [14]). 2.1. EQUIPMENT AND THEIR ELEMENTS The equipment mentioned in this chapter are relevant in the virtualization of the gating design and in CAE software’s boundary conditions. The equipment used in high pressure die casting which influence the gating design are [15]: • Chamber • Cooling system • Venting system The type of chamber influences the energy efficiency of the process, rulling the temperature drop of the molten metal. The type of chamber also constraint the type of casting alloy and the cycle time. The cooling system modifies the solidification dynamic of the casting and is design to reduce the shrinkage porosity of the casting during solidification. Also it partially controls the cycle time by controlling the solidification time of the casting. At last, the venting system removes the air in the die cavity, reducing the air porosity of the casting. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 13 Tab 2.2 – Intensification pressure based on the alloy (adapted from [30]). Alloy Pressure [𝒃𝒂𝒓] Aluminum 400 – 1000 Magnesium N/A Zinc 100 – 400 𝐹𝑙𝑜𝑐𝑘𝑖𝑛𝑔=𝑝𝑝𝑖𝑠𝑡𝑜𝑛,3𝑟𝑑𝐴𝑝𝑟𝑜𝑗𝑒𝑐𝑡𝑒𝑑 [2.5] In equation [2.5], 𝐹𝑙𝑜𝑐𝑘𝑖𝑛𝑔 is the required locking force, 𝑝𝑝𝑖𝑠𝑡𝑜𝑛,3𝑟𝑑 is the third stage intensification pressure and A the projected area of the component normal to the part line. Pouring and mold's temperature The pouring temperature of an alloy is determined based on its liquidus temperature, and the casting alloy is typically overheated by 50ºC. This overheating is necessary to compensate for the temperature drop that occurs during the casting process [63]. The extent of this temperature drop is influenced by several factors, including the: • Type of chamber: cold chamber has an higher temperature drop than the hot chamber [64]; • Geometry of the runners: higher length of the runner results in an higher temperature drop. However, its geometric section outweights the runner’s length during solidification; • Shot velocity: a slower shot allows the casting alloy to lose more heat; • Shot sleeve filling percentage: it can be a low filling ratio (20%-30%) [55] or high filling ratio (60%-80%) [56]. An higher filling ratio prevents an high temperature drop, due to the thermal inertia; The die’s temperature is also a parameter that influences the temperature drop, since heat transfer is proportional to the thermal gradient. Initially, to maintain the desired thermal conditions, a cooling-heating system and warm-up shots are used to heat the mold and regulate its temperature [65]. Approximately 20 thermal cycles [66] are performed to establish a stable operating thermal gradient. Throughout these thermal cycles, the coolingheating system also cools the die, playing a critical role in the solidification dynamics of the casting. It is desirable that mold’s thermal inertia can hold enough energy to prevent re-heating the mold. The parameters which influence the heat transfer is the temperature gradient, contact area, the type of flow and the thermophysical properties of the cooling fluid. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 14 The initial mold temperature and pouring temperature are key variables required to calculate the filling time (detail in 2.4). 2.4. VIRTUALIZATION OF THE HIGH-PRESSURE DIE CASTING PROCESS The gating design presented by NADCA [6] follows a methodology well defined which allows the user to adquire knowledge of the process and its effects on the gating design. The manual is divided into 9 different steps: 1Casting quality requirements 2Determine the flow pattern and location of the ingates and outgates 3Segment volumes, Cavity fill time and flow rate 4Process flow rate 5Determine the ingate parameters 6PQ2 analysis 7Design the fan and tangential runners 8Design the overflows and vents 9Simulation The algorithm which will be presented is a simplified and compacted version of the actual algorithm. In this simplified version, step 1 and 2 are a prior task which will not be detailed, since those are mainly conceptual which can not be virtualized into the algorithm. The simplified algorithm is: • Ingate area: englobes the 2nd, 3rd and 4th steps shown previously. The 6th step is performed if desired, which is only to validate the machines capacity (detailed in chapter 0); • Runners definition: englobes the 7th step; • Overflows and venting: englobes the 8th step; • Simulation: detailed in chapter 4 and 5. Ingate Area Initially, it is done the conceptual division in segments of the component based on the number of ingates required by the component. Then, based on the segment’s volume (𝑉𝑠𝑒𝑔𝑚𝑒𝑛𝑡) and the filling time (𝑡𝑓𝑖𝑙𝑙), it is possible to use the continuity equation to determine the ingate area (equation [2.3]). The equation to calculate the filling time of the part is shown in equation [2.6] [64]. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 15 𝑡𝑓𝑖𝑙𝑙=𝐾1𝑇𝑝𝑎𝑟𝑡(𝑇𝑖−𝑇𝑓+𝑆𝑍 𝑇𝑓−𝑇𝑑) [2.6] In equation [2.6], 𝑡𝑓𝑖𝑙𝑙 is the filling time of the component and overflows, 𝐾1 constant based on die’s material (related to the thermal conductivity), 𝑆 is percentage of solidified metal at the ending of the filling, 𝑍 solids units conversion factor, 𝑇𝑝𝑎𝑟𝑡 avarage wall thickness and 𝑇𝑓, 𝑇𝑖 e 𝑇𝑑 as minimum flow temperature, injection temperature at the ingate and die’s surface temperature, respectively. For different alloys, the recommended ingate velocity is shown in Tab 2.3. The ingate velocity (𝑣𝑖𝑛𝑔) definition is based on the components. To calculate the effective ingate (𝐴𝑒𝑓𝑓,𝑖𝑛𝑔) area is based on continuity equation [2.3] [58]. 𝐴𝑒𝑓𝑓,𝑖𝑛𝑔=𝑉𝑠𝑒𝑔𝑚𝑒𝑛𝑡 𝑡𝑓𝑖𝑙𝑙𝑣𝑖𝑛𝑔 [2.7] Tab 2.3 – Ingate velocity for different alloys (adapted from [59]). Alloy Ingate velocity [𝒎/𝒔] Aluminum 18 – 40 Magnesium 25 – 50 Zinc 23 – 50 Due to fluid’s inertia, the actual ingate area is bigger than the effective ingate area as it shown in Fig 2.8. The relation between the these two parameters is the flow angle which is presented in equation [2.3]. The flow angle (𝜃𝑒𝑓) is a parameter to verify in the postprocessing of the casting simulation. 𝐴𝑎𝑐,𝑖𝑛𝑔=𝐴𝑒𝑓𝑓,𝑖𝑛𝑔 cos𝜃𝑒𝑓 [2.8] Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 16 Fig 2.8 - Geometric representation of the effective and actual ingate area in a tangential runner. The last parameter to a total definition of the ingate profile is the thickness. Normally, ingates vary from 0.75 mm to 2 mm based on ingate velocity. However, the usage of equation [2.9] is more precise. 𝑣𝑖𝑛𝑔 1.707𝑇𝑎𝑐,𝑖𝑛𝑔𝜌=𝐽 [2.9] In equation [2.9], 𝐽 is a constant based on the material for atomized flow, 𝑣𝑖𝑛𝑔 is the ingate velocity (m/s), 𝑇𝑎𝑐,𝑖𝑛𝑔 is the ingate thickness (mm) and 𝜌 is the density of the fluid (kg/m3). The NADCA manual also provides some information for typcal thickness for ingate values for 𝐽 equal to 998000 [67]. In the end of the ingate definition step, the geometry of the ingate (trapezoid with 10º inward (Fig 2.9 (a))), ingate velocity and fill time are totally defined. The ingate dimensions will be required to the runner’s definition. The ingate velocity will be required to the definition of the plunger’s velocity (first and second stage). The fill time will be required to the definition of the transient boundary conditions in the CAE software. Runners definition The runners are divided into 3 categories: (i) curved sided fan, (ii) straight sided fan and (iii) tangential runner. The difference between these three is the way the the metal is distributed. IN both fans, the flow’s direction in the fan during filling is colinear with the fan’s length. In other hand, the tangential runner’s flow’s direction is perpendicular [6]. The selection of the runners is based on the part’s geometry. The fans generate a strong centre fill. Tangential runner are more compact and can distribute the metal over a longer distance [30]. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 17 The difference between the fans is the width and thickness profile across the fan’s length, but both profile section is the same (Fig 2.9 (a)). The thickness profile of a straight sided fan varies nonlinearly with the length of the tangencial runner. In other hand, the curved sided fan has a linear thickness profile. This happens in both fans to avoid the creating of voids in the fans edges and guarantee that the whole fan is complely full during the filling stage. The profile of the fans and tangential runner are presented in Fig 2.9. The equations to determine the parameters for the fans and tangential are presented in Tab 2.4 and Tab 2.5, respectively. Fig 2.9 – Runners profiles. (a) - Fan; (b) - Tangencial runner with 30ºC approach (adapted from [6]). Tab 2.4 - Equation for fan’s dimensions (adapted from [6]). Parameter Equation 𝒉 √𝐴𝑠𝑒𝑐 𝑟𝑎𝑠𝑝 𝒄 and 𝒅 ℎcos80° 𝒃 𝐴𝑠𝑒𝑐 ℎ+2ℎcos80° Tab 2.5 - Equation for tangencial runner’s dimensions (adapted from [6]). Parameter Equation 𝒉 √𝐴𝑠𝑒𝑐 𝑟𝑎𝑠𝑝 𝒅 ℎcos80° 𝒄 ℎcos30° 𝒃 𝐴𝑠𝑒𝑐 ℎ+𝑑+𝑐 The 𝑟𝑎𝑠𝑝 is the aspect ratio, which is the ratio between the avarage width and the height (typically between 3:1 to 1:1). The higher the aspect ratio, the slower it solidifies, because the module of the section Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 18 is maximized. The parameter 𝐴𝑠𝑒𝑐 is the area of the cross section. The NADCA methodology presents a 9 cross section for the fans and a 5 cross sections calculation for the tangencial runner. The distance between these section is calculated based on the flow angle. The calculation the 𝐴𝑠𝑒𝑐 parameter is different for tangential runners and fans, Fig 2.10. The tangential runner’s cross section’s area increases linearly from the shock absorved to the main runner. Meanwhile, the fan cross section increases linearly from the ingate to the main runner. The main runner’s area is based on the ingate area multiplied by the ratio ingate/runner, 𝑟𝑖𝑛𝑔,𝑟𝑢𝑛. The parameter 𝑟𝑖𝑛𝑔,𝑟𝑢𝑛 is the ingate/runner ratio, which increases the area of the runner to guarantee that the ingate is complely full before the beginning of the 2nd phase of the piston (injection), which varies with the casting alloy (Tab 2.6). Tab 2.6 - Typical values of 𝑟𝑖𝑛𝑔,𝑟𝑢𝑛 (adapted from [6, 67]). Alooy 𝒓𝒊𝒏𝒈,𝒓𝒖𝒏[𝒎𝒎] Aluminum 1.1 – 1.5 Magnesium >1.1 Zinc 1.05 – 1.15 Fig 2.10 - Cross section of the curved sided fan (a) and tangencial runner (b) (adapted from [6]). The main runner profile section is the same as the fan. In the main runner, it is important to notice that the section area have to remain constant throughout the runners. Overflows and venting The overflows are typically found either in the last regions to be filled or in areas where air can be encircled away from other overflows originating high porosity region. Therefore, an early predition of a filling simulation is typically performed to predict the filling sequency of the casting. Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 19 The metal enters in the overflow through the outgate (the sum of all outgate areas represents half of the sum of all ingate areas), whose thicknessis presented in Tab 2.7. A large portion of overflow’s volume is filled before the molten enters the vent region. The sum of all overflow’s volume varies from 75% to 25% based on segment’s volume, casting’s thickness and required quality. The visual representation of an overflow is presented in Fig 2.11. Fig 2.11 - Overflow overall dimensions (adapted from [68]). Tab 2.7 - Outgate thickness (adapted from [59]). Alloy 𝒕𝒐𝒖𝒕 [𝒎𝒎] Aluminum 1< Magnesium 0.5< Zinc 0.5< The vents are extension of overflows to delay even further the molten metal, since these are region are the last to fill. Through the vents, the air is removed from the casting [68]. Therefore, vents must have high minor and major loss to delay the molten metal. The vents area can be calculated based on the actual ingate area, [2.10]. Its thickness is typically 0.2 mm. Typically, the exit thickness is less half of the initial thickness. 𝐴𝑣𝑒𝑛𝑡=𝐴𝑎𝑐,𝑖𝑛𝑔 4 [2.10] Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 20 2.5. CHAPTER REFERENCES [1] T. V. Sai, T. Vinod, and G. Sowmya, "A Critical Review on Casting Types and Defects," IJSRSET . [2] H. S. Sodhi, "APPLICATION OF DIE CASTING: A REVIEW PAPER," ReseachGate , 2018. [Online]. Available: https://www.researchgate.net/publication/342349012_APPLICATION_OF_DIE_ CASTING_A_REVIEW_PAPER [3] M. S. Ramaprasad and M. N. Srinivas, "Permanent Molding of Cast Irons – Present Status and Scope," in Science and Technology of Casting Processes , M. Srinivasan, Ed.: InTech, 2012. [4] M. Srinivasan, Ed., Science and Technology of Casting Processes : InTech, 2012. [5] Rapiddirect, Low Pressure Die Casting: A Simple Guide to Understand the Process. [Online]. Available: https://www.rapiddirect.com/blog/low-pressure-die-casting/ (accessed: Sep. 19 2024.871Z). [6] NADCA, NADCA Gating Manual . Illinois, United States of America: NADCA. [7] Q. Hu, W. Guo, and H. Zhao, "Influence of wall thickness and intensification pressure on the microstructures and mechanical properties of AlSi7-SiC composites fabricated by the vacuumassisted HPDC process," Materials Science and Engineering: A , vol. 819, no. 3, p. 141470, 2021, doi: 10.1016/j.msea.2021.141470. [8] F. Bonollo, N. Gramegna, and G. Timelli, "High-Pressure Die-Casting: Contradictions and Challenges," JOM , vol. 67, no. 5, pp. 901–908, 2015, doi: 10.1007/s11837-015-1333-8. [9] C. Bharambe, M. D. Jaybhaye, A. Dalmiya, C. Daund, and D. Shinde, "Analyzing casting defects in high-pressure die casting industrial case study," Materials Today: Proceedings , vol. 72, no. 4, pp. 1079–1083, 2023, doi: 10.1016/j.matpr.2022.09.166. [10] G. M. Gonçalves, "Implementação de um Sistema de Controlo de Processo e Projeto de Coquilha numa Empresa de Fundição," Master's Degree Thesis, Escola de Engenharia, Universidade do Minho, Guimarães, 2014. [11] H. Bakemeyer, Operating the die casting: Machine. [Online]. Available: https://www.dykast.com/ user/files/operating_the_die_cast_machine.pdf (accessed: Sep. 15 2024). [12] H. Liang and J. Qiao, "Analysis of Current Situation, Demand and Development Trend of Casting Grinding Technology," MPDI , 22 Sep., 2022. https://doi.org/10.3390/mi13101577 [13] G. Bar-Meir, Fundamentals of Die Casting Design . Illinois, United States of America, 2012. [Online]. Available: https://zenodo.org/records/8173341 [14] EKK Inc, HPDC Gating Design. [Online]. Available: http://www.ekkinc.com/consulting/gatingdesign/ (accessed: Mar. 22 2024.248Z). Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 21 [15] M. T. Murray and M. Murray, "High pressure die casting of aluminium and its alloys," in Fundamentals of Aluminium Metallurgy : Elsevier, 2011, pp. 217–261. [16] Butler W. A., "Die Casting," Encyclopedia of Materials: Science and Technology , pp. 2147–2152. [17] Mohamed Refaat Aboel-fotouh, Ahmed Shash, and Prof. Dr. Abd El-Halim El-Habak, "SEMIAUTOMATED GATING SYSTEM DESIGN WITH OPTIMUM GATE AND OVERFLOW POSITIONS FOR HPDC," Philosophy degree, Unpublished, 2018. Accessed: Sep. 16 2024. [Online]. Available: https://www.researchgate.net/publication/330727487 [18] M. Okayasu, S. Yoshifuji, M. Mizuno, M. Hitomi, and H. Yamazaki, "Comparison of mechanical properties of die cast aluminium alloys: cold v . hot chamber die casting and high v . low speed filling die casting," International Journal of Cast Metals Research , vol. 22, no. 5, pp. 374–381, 2009, doi: 10.1179/174313309X380413. [19] Oliveira, Steven Richerd Pires de, "High Pressure Die Casting of Zamak alloys," Master's Degree Thesis, Faculdade de Engenharia da Universidade do Porto, Universidade do Porto, Porto, 2018. [20] Butler W. A., Encyclopedia of Materials: Science and Technology: Die Casting (Permanent Mold) . United States of America: Elsevier Science. [21] Aços Ramada, Camisas de Injeção. [Online]. Available: https://www.ramada.pt/pt/aplicacoes/ acos-ligas_/industria-automovel/moldes-de-fundicao/camisas-de-injecao.html (accessed: Mar. 22 2024.741Z). [22] E. spa, What is Cooling System in Die Casting Mold – What Affects the Cooling Rate of Die Casting. [Online]. Available: https://www.diecasting-mould.com/news/what-is-cooling-system-indie-casting-mold-what-affects-the-cooling-rate-of-die-casting (accessed: Sep. 16 2024.661Z). [23] L. Maldonado, C. Shi, C. Sinclair, C. Vian, and J. Ostanek, "Rapid die preheating using a highpower infrared lamp array," Elsevier , 2022. [Online]. Available: https:// pdf.sciencedirectassets.com/271641/1-s2.0-S1359431121X0018X/1-s2.0S1359431121011455/main.pdf?X-Amz-Security-Token= IQoJb3JpZ2luX2VjEN7%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaCXVzLWVhc3QtMSJIMEYCIQDL kbSXapp8qnu5hnNymbTWX6AkeRNyr40IslgylzknbwIhAIlaXhdPfxFT7gKbDuJdBMCiQ95XRODMGn ddHArT304WKrMFCBcQBRoMMDU5MDAzNTQ2ODY1Igy16ImY19ezUpK6OfQqkAUJTprKeL2u2TC OYEfd8No5LuDn8KtnS7HEjQK%2F0%2B2DN%2BJJy6cXjhM36BYdIUDdufIxO1RCi62XmPehua5tFl Xssf2%2B9J5eQ8%2Frq2paSh4bIZCt9mKS93x70mDVeLmpwE8fgRcSq9O8f%2BLfEKD69isDeOaA DbEwEgsdEYRHQ97h0CkOVRFaJ3OZiC5IooBK8xGQK3aW7fHPmXD60B4UGbfxRWd1aXnz%2BNq dJduZJf6FRa9la0gPgt91YFr38gjnsEZL8JpHOHsyfX8IsQZ%2FW%2B3Z18P4NkQY2OvD2527SX8BU Chapter 2 – Literature review -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 22 jX11E5MO71fWqdte1e434GWActjeWudLP6wxHZAcFOYvPvExc0XZrc7no%2B0nNxjT6xd4TafCnBHz xb7UaYg8FFmkzHNWxN%2B6NqKBpR5KpGiyfV1O748VCx5t%2FIqWpc9aZm%2FwYTMtmSmhxbN %2B3Q%2Fjo1D2PzQm4JIvLNCPk%2FRs%2F02GzRVjAsPXyDUc7y81iFASFcxkxGRmuICdgsU5hlnp Drq9xNM7tBAntrypI8%2F3l9BJTga6Z9aHM1IABMQD%2BT8m8v024f3RzIy%2Foo2H37hJISxlaTivO UvxeaAMu07CP1v1PkXXDE02U5Efp597nXWNuK%2Bxj6DsXDu5PhyfjjwIxxl6krQxgnyByHyiyvEx0Lc 3H9hJN3cYnvgefyHbl4E8id%2FgjR6HV8bVDn%2BCUFdxpv%2BUhsh%2FAcEcN1CQPygRdJas9Ipd Kn2N0wCecFZfR6eGl%2BqOUfRVYMtivoYnYUJhVvu7jn%2FcrlCch6GSHJfPFqumopRq4eDZ%2FSQ W9sQgduxHjl2C5Ffwy3886XD7Q6nsLglp0teQPKR0Ee1gqlzBl9NX1F0UTqGPKcS%2FVI7SIe5sTDX 8qC3BjqwAf%2B%2BI2v%2F3tV1HJpKbQo%2FmWTI0ohrXRlL8ijxB048hRH8EZONhrXlgt4Yt17Nrkd hj5isMhLG4MeZr7pU%2FlGQcQNWKC5CymCiwjktmDPfi%2FtFDCrWJizCi3DMQIZz0wMpI%2Fn1gxZ lqe8rDO4STbGuZmjRMONkm%2BRJK3lVvevGzifrGADmtdXgeYhQ5FYZ2HzUGwTUuS7ulU5jhMhlho v9PvlPVOTlAx0RRCUdwzdD%2B52s&X-Amz-Algorithm=AWS4-HMAC-SHA256&X-Amz-Date= 20240916T151557Z&X-Amz-SignedHeaders=host&X-Amz-Expires=300&X-Amz-Credential= ASIAQ3PHCVTYQC2DT7EW%2F20240916%2Fus-east-1%2Fs3%2Faws4_request&X-AmzSignature=d363799883566bdbf99146c1ac4b6e991f8b1af624d2b9845a6b2cdb3585c6a8& hash=b2ca3136ca4096a270169a82c36c8c9bd376d8f2288e5e05438a2f043fb7ee42&host= 68042c943591013ac2b2430a89b270f6af2c76d8dfd086a07176afe7c76c2c61&pii= S1359431121011455&tid=spdf-67ff3871-2bfb-425f-bf73-db8010f004b5&sid= 0d38888225c940461429a8d308fdc73cddd4gxrqb&type=client&tsoh= d3d3LnNjaWVuY2VkaXJlY3QuY29t&ua=1e105a035355055556&rr=8c41d09b19fe692a&cc=pt [24] A. Long, D. Thornhill, C. Armstrong, and D. Watson, "The Impact of Die Start-Up Procedure for High Pressure Die Casting," SAE Int. J. Mater. Manf. , vol. 6, no. 3, pp. 416–426, 2013, doi: 10.4271/2013-01-0829. [25] A. Long, D. Thornhill, C. Armstrong, and D. Watson, "Predicting die life from die temperature for high pressure dies casting aluminium alloy," Applied Thermal Engineering , vol. 44, pp. 100–107, 2012, doi: 10.1016/j.applthermaleng.2012.03.045. [26] P. Callegaro, Comparative analysis of temperature control systems for high pressure die casting dies: Numerical methods for the thermal analysis of the HPDC process are needed to shorten the cost, the design and production time and the samples to involve in finding acceptable operating conditions. [Online]. Available: https://www.enginsoft.com/expertise/comparative-analysis-oftemperature-control-systems-for-high-pressure-die-casting-dies.html (accessed: Sep. 16 2024.864Z). Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 29 A previous study of the path to be taken in order to match the proposed challange was made and the following reasons defined the programming language and the CAD software: •Python is indeed a widely used programming language, as it is a high-level language that is easy to learn and has an excellent interface with any type of database/data storage software, including Excel. Another reason for using Python is the possibility of the software also interfacing with CAD software [2]. •The program is only available for Autodesk Inventor. To make it compatible with SolidWorks, it would be necessary to "translate" the language to Virtual Basic for Applications (VBA). For Autodesk Fusion 360, only a slight change in keywords is needed since the core modeling program is similar to that of Autodesk Inventor, as both software are developed by the same company.. •Another near-solution for programmed modeling is to have the model fully parameterized linked to an Excel spreadsheet, then an overwrite of these parameters would be made. However, this is not a viable solution. According to the content studied in the NADCA's "Gating Design" manual, there are various models for gating—Curved Sided Fan, Straight Sided Fan, and Tangential Runner. Considering the different scenarios, the number of files to be parameterized could exceed 48 different scenarios (excluding the various parameter values). This would make it impractical to have software with 48 associated CAD files to calculate a single situation—the one requested by the user—with at least 41 parameters to be filled out in every CAD file. 3.1. SYSTEM REQUIREMENTS The development of H.E.L.P. Die Casting was conceived for a one-user only, which in this scenario is it’s author, the software ergonomics was not the main concern during its development. So a organization model was simplified which can be divided into 3 main folders, Fig 3.2. In the folders “Excels”, “Figures” and “Parts” are stored the databases in Excel file (“.xlsx”), figures used in the software to help the user to fill the gating parameters and 3D model of gating design and overflow, respectively. The Python files, which are responsible for the software itself, are presented in Tab 3.1. A new folder, QuikCast Macro, was being developed in the free time of this dissertation which main goal was to allow the user to save time. In this macro folder, a new macro was conceived to help the user at defining the boundary conditions in Visual-Cast. This macro stablished all the boundary conditions after defining the surfaces where they would be attributed. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 30 Fig 3.2 - Organization of H.E.L.P. Gating Design. Tab 3.1 – Function of each Python file. Python file Function Biscuit Generates the 3D model of the biscuit Brain Initializing file and is the bridge for every file Display Interface HELP Gating Design – user Gating System Generates the 3D model of the gating design Overflow_vent Generates the 3D model of the overflow Ratios Provides ratios for a proportional of the gating system STEP_X Calculates every step’s variables based on input As a consequence of unecessary user-friendliness, there are some pre-requirements are needed in order to run the program which can be allocated in two main groups: • Files location: The software is designed to write information to an external database (Excel, in this case) stored in the same folder as the executable. In addition to the database, other folders, such as 'Figures' and 'Parts,' used by the software require their file locations to be specified; Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 31 •Python libraries: the Python libraries is a requirement only to the user who runs the Python files instead of an executable or an installable. In such cases, Python libraries are annexed to the program increasing the file size. Files location The file directory could be a variable which the user could define. However due to its timeconsuming activity of searching for its location and one user-only, another approach was choosen. It was defined a fixed folder was defined to facilitate the file’s search (Fig 3.3). If the program was converted into an installable, this problem would be supressed, since during the installation it is defined a permanent file directory. In Fig 3.3, the variable “rootDir” could be changed based on the user’s file location. In this case, the file to find is in the excel folder (Fig 3.2). If this variable is empty (notation is: “”) the program would have to search for the desired file in the entire computer storage, increasing the time consumption (aout 30 seconds everytime it was required to open a new file). Fig 3.3 – Block code which searches for each file in a fixed folder. Python libraries As mentioned previously, Python libraries required a previous installation if the user intends to run the program from a Python file. In Tab 3.2 is presented all libraries required to create the software. It is worth mentioning that libraries such as “numpy” and “math” were not mentioned, since they were pre-installed when downloading Python. The "tkinter" library [3] allows programmers to design a functional GUI that enhances the HELP Gating Design user interface, making it more user-friendly. Without this interface, users would need to input all parameters through a command terminal. The "openpyxl" library [4] is responsible for reading and writing Excel files. It handles data input provided by the user and is utilized during the modeling phase. While entering data into the filling system, Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 32 the Excel files must remain closed, as the program needs to save the data and close Excel during execution. However, during the modeling phase, users can access the Excel files. The "OS" library [5], short for "operating system," is included with the Python installation. Its presence should be confirmed, as it facilitates communication between Python and the computer. This library is used to locate files that the software needs to read or write, such as Excel sheets or images for the GUI. Similarly, the "win32com.client" library [6] is derived from the computer's operating system and allows Python to access all applications installed on the computer. To run the software without having to open any Python file, those files can be compiled and then executed just like a generic application with “Auto Py-to-Exe” library [7]. Tab 3.2 – Required libraries. Library Function tkinter Interface Python – User openpyxl Interface Python – Excel os Interface Python – System 32 win32com.client Interface Python – Inventor Auto Py-to-Exe Create an executable file of the program 3.2. INTERFACE USER - SOFTWARE In the design of the Guided User Interface (GUI), the programmer ensured that the information required from the user was organized in a way that minimized the number of windows the user would need to fill out. The arrangement of the information was carefully considered to reduce the number of steps needed for user input. For instance, the software needs to be informed about the two points that define the machine's curve (required for step 6) (Fig 3.4, parameter (g)) before it receives information about the volume of the part (required for step 3) (Fig. 4.5, parameter (a)). In the initial GUI (Fig 3.4), the user first specifies the type of chamber, based on the material to be used (parameters (a) and (b)), the mold material (parameter (c)), the average thickness of the thinnest walls (parameter (d)), the number of parts in each mold (parameter (e)), and the piston diameter (parameter (f)). The machine’s parameters (parameter (g)) units for pressure and flow rate are MPa and l/s, respectively. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 33 Fig 3.4 - First GUI layout which the user will fill. The last three parameters, (h), (i), and (j), are related to the type/number of gates that will be applied to the part. If the parts are identical and have identical gates, parameter (h) should be set to "Yes." This will reduce the amount of data that needs to be introduced, as the software will automatically apply the same parameters to all the gates. Depending on the type and number of gates, it may be necessary to input variables that are not common to all approaches, such as the orientation of the gates (parameters (d), (e) and (f) in Fig 3.5) when there is more than one attack. If there is more than one attack, the H.E.L.P. Gating Design provides functionality to calculate the gate layout, ensuring there are no geometric incompatibilities between the parts. To perform this calculation, additional information about the mold plate alignment and the part is required (Fig 3.7). However, certain variables must be defined regardless of the approach, such as volume, entry velocity, and the runner-to-ingate ratio (parameters (a), (b), and (g) in Fig 3.6). Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 34 Fig 3.5 - Second GUI layout which the user will fill for a scenario of 4 attacks to 4 equal parts using a curved sided fan. Fig 3.6 - Second GUI layout which the user will fill for a scenario of 1 attack to 4 equal parts using a tangential runner. In the third GUI (Fig 3.7), the alignment of the mold and the part is defined (parameters (a) and (b)), along with the information required to set up the air removal system and the final casting quality (parameter (c)). The combine overflow volume is based on the parts volume, following NADCA methodology. Defining its length and number of overflows, the implicit variable is the width of the overflow. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 35 Fig 3.7 - Third GUI which the user might fill. 3.3. WORKFLOW The recommended workflow using this software passes through three main phases, as it is presented in Fig 3.8. The first step, initial decisions, is based in understanding the components geometry and prepare it to the casting process, such as dividing the component in segments and the attacks’ location, applying drafts and fillets, definition of parting line, etc. As auxiliay tool, a solidification simulation may be performed to determine the exact location of the hotspots to the definition of the cooling system. This simulations is a mere approximation of the real casting conditions. The second phase, the gating design phase, is the where the H.E.L.P. Die Casting software is applied to reduce time consumption of the user. The process’s data is stored in databases which will be used lately in the die casting simulation. The gating design phase is the most complex phase, since it evolves the manipulation of CAD files with excel which both work as data storage and database (Fig 3.2). To distinguish the worksheets that serve as database from those which are data storage, the designation format is “Table X, Page X”. This format was based on what page and which table the data was in the NADCA Manual. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 36 Fig 3.8 – Workflow of gating design and the positioning in the iterative process H.E.L.P. Gating design software. In a detail view, in this phase is generated two CAD files, the Gating Design and the overflows, and stored information of the process introduced by the user and postprocessed by the H.E.L.P. Gating Design. The file “STEP_3"’s function is to provide information of constant values to the calculation of the injection time, as well as the flow rate for different segments. This information is presented in different worksheets if the part presents more than one segment. The file “STEP_4"’s function is to determine an approximation of ingate dynamic pressure and plunge speed. This information can be lately compared and corrected with the PQ2 analyis in “STEP_6” file, to determine the actual shot velocity (this step can be replaced based on CAE postprocessing). The file “STEP_7” and “STEP_8” are related to the attacks section dimensions and overflows, respectively. These last 2 files are exclusively to data storage. As a result of a programmed CAD modelling, some details as well as the assembly require to be done manually. These details include adding fillets, removing emboss, correcting the ingate section as it is presented in Fig 3.9 as well as the assembly. This is the easiest part of the gating design where the user takes a few minutes to make small changes and the assembly (assembly of the following components: gating design and overflow and part)depending on the user experience to the software. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 37 Fig 3.9 - Attack correction in a round surface. The simulation process can be divided in three simulations to save computation time. The first simulation is to verify the filling pattern and sequency to guarantee that the overflows are located in key regions where the air can be encircled and trapped within the casting. This simulation only deals with the filling simulation module, considerating that the fluid’s temperature remain constant throughout the filling simulation (ending when the die cavity is full). If the first simulation’s results are satisfactory, the second simulation shall be done before the casting simulation, the die cycling simulation. This simulation only deals with the solification of the casting, which provides the casting simulation a more accurate die’s thermal field. The third simulation consists in using the die’s therm field from the die cycling simulation as initial conditions for the mould for the die casting simulation. Chapter 3 – H.E.L.P. Gating Design -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 38 3.4. CHAPTER REFERENCES [1] Autodesk, Inventor : Autodesk, 2023. [Online]. Available: https://www.autodesk.com/products/ inventor/overview?term=1-YEAR&tab=subscription [2] B. Ekins, Inventor API Object Model . California, United States of America: Autodesk. Accessed: Feb. 5 2024. [Online]. Available: https://www.autodesk.pt/ [3] A. D. Moore, Python GUI Programming with Tkinter: Develop responsive and powerful GUI applications with Tkinter . Birmingham: Packt Publishing, 2018. [Online]. Available: https:// search.ebscohost.com/login.aspx?direct=true&scope=site&db=nlebk&db=nlabk&AN=1813737 [4] C. Clark and E. Gazoni, openpyxl - A Python library to read/write Excel 2010 xlsx/xlsm files — openpyxl 3.1.2 documentation. [Online]. Available: https://openpyxl.readthedocs.io/en/stable/ (accessed: Feb. 5 2024.421Z). [5] Python documentation, OS: Miscellaneous operating system interfaces. [Online]. Available: https://docs.python.org/3/library/os.html (accessed: Aug. 19 2024.878Z). [6] Python Package Index, pywin32. [Online]. Available: https://pypi.org/project/pywin32/ (accessed: Aug. 19 2024.402Z). [7] B. Vollenbregt, auto-py-to-exe 2.42.0. [Online]. Available: https://pypi.org/project/auto-py-to-exe/ (accessed: Feb. 5 2024.428Z). Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 45 The CAD model was edited (exit angles, fillets, etc) in accordance with NADCA products specifications standards [15]. Then, the part’s properties relevant to the its gating design is shown in Tab 4.5. Mass and volume properties are required to determine flow rate and dimensioning gating design, component’s size is required to determine the size of mold plates and project area required to verify the locking force. Tab 4.5 – Relevant properties to the gating design. Property Value Mass 134 g Volume 49.5×103 mm3 Dimensions (X, Y, Z) 39×125×95 mm Projected area 4753 mm2 Before using H.E.L.P. Die Casting to the gating design, an initial idealization of segments based on part’s hotspots. In case of a small and uniformal component the unique present segment is the entire component. However, as it is presented in Fig 4.6, it is possible to identified that there are critical regions which the solidification time gradient would be higher than others. Fig 4.7 - Identification of the hotspots. Filling time and Ingate velocity The algorithm shown in chapter Chapter 3 - 3.3 is applied to this scenario in to calculate the filling time using equation [4.1] of the entire die cavity which results in the component. In this equation, 𝐾, 𝑇𝑝𝑎𝑟𝑡, 𝑆, 𝑍, 𝑇𝑖, 𝑇𝑑 and 𝑇𝑓 values were 0.0346 s/mm, 2 mm, 0.2ºC/%, 4.6%, 710ºC, Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 46 577ºC and 300ºC respectively. Therefore, filling time (𝑡𝑓𝑖𝑙𝑙) was 0.0361 seconds. It is to note that die temperature value 𝑇𝑑 was based on initial temperature of the mold before the first cycle. 𝑡𝑓𝑖𝑙𝑙=𝐾1𝑇𝑝𝑎𝑟𝑡(𝑇𝑖−𝑇𝑓+𝑆𝑍 𝑇𝑓−𝑇𝑑) [4.1] NADCA gating design manual did not possess any information about the alloy, therefore there had to be determined. The 𝑇𝑖, 𝑇𝑑 and 𝑇𝑓 values are the pouring, die and minimum flow temperatures. 𝑇𝑓 and 𝑇𝑖 values attributed were 577ºC (liquidus temperature) and 690ºC. However prior simulations results indicated that the pouring temperature had to be increase to 710ºC to avoid an early solidification. The die temperature was based on NADCA information with small adjustments based on simulation results. 𝐾1 was a constant based on the mold’s material (steel H13), 𝑇𝑝𝑎𝑟𝑡 avarage thickness of the narrowest regions, the variable 𝑆 was based on surface finishing and variable 𝑍 value was based on the alloy volume change property which was 4.6%. As mentioned in chapter 2.4, for aluminium alloys the ingate velocity is within 18 to 40 m/s. The gating design was dimensioned to a ingate velocity (𝑣𝑖𝑛𝑔) of 20 m/s. This value was choosen in order to reduce mold’s wearing and fatigue. Furthermore, it would reduce heat loss during the injection phase since higher velocities results in an higher convective heat transfer coefficient. Ingates and Runners section Each ingate would directionate the flow to both segments previously defined as it is shown in (Fig 4.8). It was defined an non-uniformal attack’s distribution through part’s length in Y direction, since it was oriented to the most critical regions which are more susceptible to shrinkage. The attacks section are a crucial part of the design. Not only it is responsible for the fluid’s orientation but it is also responsible for the compaction phase during solidification. If the attacks solidify before the most massive region of the part, shrinkage porosity is expected. This occurs because the piston cannot push the material in the runner into the part. However, an attack section too robust may not distribute the molten metal through the part but it will concentrate it in one place and may increase the difficulty of cutting it off in the post process operation. Therefore, the used attack thickness (𝑇𝑎𝑡𝑡𝑎𝑐𝑘) was 1.8 mm. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 47 Fig 4.8 Attacks to each segment and its non-uniformal division through the parts length in Y axis and its critical regions. Analyzing the attacks position, it was idealized that the flow orientation angle was 35º, this value will be later verified and analyzed. Furthermore, it is possible to determine attacks width, equation [4.2]. 𝑤𝑒𝑓𝑓,𝑖𝑛𝑔=𝑉𝑝𝑎𝑟𝑡 𝑇𝑎𝑐,𝑖𝑛𝑔𝑣𝑎𝑐,𝑖𝑛𝑔𝑡𝑓𝑖𝑙𝑙cos𝜃𝑒𝑓 [4.2] The equation [4.2] is the continuity equation written in order to the attack width (𝑤𝑒𝑓𝑓,𝑖𝑛𝑔), 23 mm, with the correction factor for flow orientation, cos𝜃𝑒𝑓. The ingate area actual is normally 1.1 times bigger than the theoric value and the tangential runner section’s area is typically 1.2 to 1.5 times bigger than the ingate area actual. This happens in order to make sure the entire section is completely full and avoid backflow [16]. The tangential runner/ingate choosen was 1.5 in order to delay the solification. This value was found the best during the iterative process in trade off for some process efficiency. To delay the solidification even further, the selected aspect ratio for tangential runners was 1:1. Meanwhile, the main runner aspect ratio was 2:1, since it would take too long for the main runner to solidify, resulting in an time increase in cycle time. The runners for both parts and main runners profile sections are shown in Tab 4.6 and Tab 4.7, which variables visual representation are shown in Fig 4.9. As both segments were idealized as having the same volume, both attack’s section would be the same and the tangential Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 48 runner’s would be equal, therefore the software only generated the result for only one attack. The modelling process only has to read the data twice, instead of rewritting it. Tab 4.6 - Generated data using H.E.L.P Die Casting for tangential runners. Section Distance [mm] Area [mm2] Thickness [mm] b [mm] t [mm] c [mm] d [mm] A_1_Attack_1 0.00 4.00 2.00 2.54 1.46 0.54 0.54 B_1_Attack_1 16.82 28.42 5.33 6.27 4.39 0.94 0.94 C_1_Attack_1 33.63 52.85 7.27 8.55 5.99 1.28 1.28 D_1_Attack_1 50.45 77.27 8.79 10.34 7.24 1.55 1.55 E_1_Attack_1 67.27 101.70 10.08 11.86 8.31 1.78 1.78 F_1_Attack_1 84.08 126.12 11.23 13.21 9.25 1.98 1.98 A_2_Attack_1 0.00 4.00 2.00 2.54 1.46 0.54 0.54 B_2_Attack_1 16.82 28.42 5.33 6.27 4.39 0.94 0.94 C_2_Attack_1 33.63 52.85 7.27 8.55 5.99 1.28 1.28 D_2_Attack_1 50.45 77.27 8.79 10.34 7.24 1.55 1.55 E_2_Attack_1 67.27 101.70 10.08 11.86 8.31 1.78 1.78 F_2_Attack_1 84.08 126.12 11.23 13.21 9.25 1.98 1.98 Tab 4.7 - Generated data using H.E.L.P Die Casting for main runners. Feeding Area [mm2] Thickness [mm] b [mm] t [mm] c [mm] d [mm] 1 126.12 11.23 13.21 9.25 1.98 1.98 2 126.12 11.23 13.21 9.25 1.98 1.98 12 252.24 11.23 24.44 20.48 1.98 1.98 Fig 4.9 - Design parameters of the tangential runner. The codename “Section” shown in Tab 4.6 for the last line (which name is “F_2_Attack_1”), means sections F (Fig 4.9) of the second part which attack number is 1. While Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 49 the the codename “Feeding” in Tab 4.7 represents the part number of each tangential runner, i.e., “Feeding” 1 means the data in the following lines is referent to the tangential runner which is responsible for feeding the first part. While “Feeding” 12 is reffered to the runner (main runner in this case) which is responsible to feed the first and the second tangential runners. Overflows The determination of the right number of overflows is an iterative process. The initial step would be based on the flow pattern previous prediction followed by an only flow simulation is required to find out what are the last filling regions. The main reason for a only flow simulation is to have an higher accuracy in the casting process, since the negative pressure caused by the venting system influence the flow pattern. Overflow’s overall dimensions was based on part’s size, casting material and avarage thickness. The NADCA manual recommended an total overflow volume equal to 25% of the part’s volume and an outgate and vent thicknes of 0.6 and 0.5 mm, respectively. In order to reduce insert’s size insert’s volume without compromising the number of overflows and its volume, it was considerated that aspect ratio of overflow would be approximated to 1:3 and it would be a square, nominal dimensions in Fig 4.10. Lately, draft would be added as well as extractor pins to facilitate overflows extraction from the die cavity. Fig 4.10 - Overflows nominal dimensions (dimensions in milimeters). It was predictable, the air entrappement in critical region 1 would be unavoidable.While regions behind the holes (shadow areas) in each corner of the part was predictable as well as the overflow in position shown in Fig 4.11. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 50 Fig 4.11 - Overflow location in the part (dimensions in milimeters). Gating design assembly As a result of the previous iterative process, a gating design was made (Fig 4.12). Due to the fact that the component is not symmetric, the first attack of one part is directed to the critical region 1 (Fig 4.7 (a)) while the other part was attacked in critical region 2. Therefore is expected different result in both parts and defects have to be analized and checked for both parts. Fig 4.12 - Gating design made using H.E.L.P. Die Casting. Nowadays, the conformed channels (CC) technology is widely used to injection molding using polymers. This technology still in development to die casting applications [17], which part of this dissertation’s project was included in. The CC technology consists in using additive manufacturing (AM) to the production of mold’s inserts (Fig 4.13) [18]. In these inserts, the negative of the casting geometry will be allocated. To reduce the construction time using AM, only the die cavity that originates the part is produced using this technology (Fig 4.14 (b)), die cavity which originates the runners are Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 51 manufactured using the traditional way (subtractive manufacturing) (Fig 4.14 (a)). During the production stage, an additional finishing machining phase is made to guarantee geometric tolerance. Fig 4.13 - Mold's inserts assembled in the mold plate. The main advantages of using CC is to increase the local cooling rate by permiting the designer to adapt (as it is shown in Fig 4.16), keeping casting’s quality without trading it off for accuracy and allowing the replacement of high wearing regions (die cavity which originates the component), without the need of making a new mold [19]. Fig 4.14 - Mold's filling cavity. (a) - filling cavity to be produce by machining; (b) - filling cavity to be produce by AM and machining. The mould’s and inserts’ dimensions are shown in Fig 4.15. Initially, the mold dimensions were based on giving a minimum offset of 100 mm from the inserts to in all directions to guarantee the mould’s stiffness. The decision taken to define insert’s dimensions was different from the one taken previously. The insert would have at least an offset of 15 mm from the die cavity which would originate the component. This value was chosen in order to reduce AM-producing time and spent material without compromising mechanical resistance [20]. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 52 Fig 4.15 - General mold's and insert's dimensions (dimensions in milimeters). This new method of producing cooling channels also affect the die’s thermal field, due to the fact that there is not a domain continuity in the mold (die plates and inserts). This happens since it is impossible to have a perfect contact between the die and the inserts as result of surfaces’ roughness which originates a contact resistance. This contact resistance is based on surfaces’s roughness, interstitial fluid and contacts pair materials. Based on this knowledge, a contact resistance of 6971 W/m2K [21] was defined. Cooling channels configuration is shown Fig 4.16. It was possible to notice that in the critical region 1, it would require an intensive cooling to cool down the hotspot. During the iterative process, it was necessary to add the cooling system on the opposite side of the casting. Fig 4.16 - Cooling channels positioning relatively to the part. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 53 Based on the avarage thickness of the casting (defined in chapter 1) it was possible to set channel’s diameter as 8 mm. The distance between the cavity’s surface and the distance from the cooling channel was 2 time the diameter and the cooling channels pitch (distance between parallel cooling channels) as well as the distance to extractor pins (positioning in Fig 4.17) was 1.5 times the diameter [18, 20]. Fig 4.17 - Positioning of extractor pins. 4.3. NUMERICAL MODEL After the conception of the CAD model, it is needed to verify the gating design and overflows. This verification is made in computer aided engeneering software for the casting process. The choosen software is Quikcast [22] of ESI Group. This software uses a different formulation (finite differences method), when compared to other softwares like Procast [23] (finite elements) or Flow-3D (finite volumes) [24]. The main difference between these types of discretization is the computational time and the accuracy of the results. In these softwares, it is possible to simulate virtually the high pressure die casting process to predict possible to verify part’s sanity and the process without having a cost to produce a mold. The casting process can be described as a transient computacional fluid dynamics problem. Therefore the most important equation which will describe the fluid’s behaviour are the Navier-Stokes equations (equation [4.3] to [4.5]) [25, 26] and the continuity equation (equation [4.7] and [4.8]) considering time dependent variables. It is important to note that these equations are only appliable to an incompressible and newtonian fluid in which the domain is isothermal. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 54 𝜌(𝜕𝑢 𝜕𝑡+𝑑𝑢2 𝑑𝑥+𝑑𝑢𝑣 𝑑𝑦+𝑑𝑢𝑤 𝑑𝑧)= −∇𝑝𝑥+𝜇∇2𝑢+𝐹𝑥 [4.3] 𝜌(𝜕𝑣 𝜕𝑡+𝑑𝑢𝑣 𝑑𝑥+𝑑𝑣2 𝑑𝑦+𝑑𝑣𝑤 𝑑𝑧)= −∇𝑝𝑦+𝜇∇2𝑣+𝐹𝑦 [4.4] 𝜌(𝜕𝑤 𝜕𝑡+𝑑𝑢𝑤 𝑑𝑥 +𝑑𝑣𝑤 𝑑𝑦 +𝑑𝑤2 𝑑𝑧)= −∇𝑝𝑧+𝜇∇2𝑤+𝐹𝑧 [4.5] Navier-Stokes equations can be described as Newton’s second law [27] applied to fluids dynamics field to a discretized system. In the first member, it is represented the inertial forces originated by the a spatial and temporal velocity gradient. In case of a stationary study, which is not the case, the temporal inertial force is 0. The second inertial force results of a velocity gradient due to a spatial gradient of velocities [28]. The second member describes all forces applied on the fluid. The pressure gradient component considerates different pressures might be applying on fluid, such as vaccum pressure originated by the venting system. The tangential forces are originated by viscosity considerating the fluid as newtonian. Furthermore, it is considerated external forces which can be applied on the fluid such as hydrostatic and static. Other equation which was used for the calculating the fluid pressure was the Poisson’s pressure equation, equation [4.6] [29]. ∇2𝑝=𝜌∇(𝑢 󰇍 ∙∇)𝑢 󰇍 [4.6] The continuity equation shown in equation [4.8] [25, 30] is the differential form of the transient equation [4.7] which can be applied to the mathematical system in order to guarantee the mass conservation. 𝑑𝑚 𝑑𝑡+𝜌𝑄𝑖𝑛=𝜌𝑄𝑜𝑢𝑡 [4.7] 𝑑𝜌 𝑑𝑡+𝜌∙∇𝑢=0 [4.8] In equation [4.7], 𝑚 is represents the mass, 𝑄𝑖𝑛 the flow rate inward and 𝑄𝑜𝑢𝑡 the flow rate outward. Whenever there is any sort of thermal gradient, heat transfer is envolved. As well as the flow model, casting process is a transient heat transfer problem. However this numeric model Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 61 indicate the fraction of casting metal which cell will be fulfilled [40]. As it is possible to verify in Fig 4.25 (a), mold’s surface intercept some cells, which will have its factor respective factor to calculate each cell’s properties (such as area, volume, density), Fig 4.25 (b). Fig 4.25 – Influence of volume correction factors. (a) – real geometry without VCF; (b) – repesctive volume correction factors as well as its discretized geometry (adapted from [22]). Consequenly, using a coarse 2D mesh will not impact volume’s properties as much as a coarse 3D mesh. A rectangular flat surface can be described perfectly by two 2D elements, however regarding surface’s tilt, two 3D elements to discretize such geometry would result in enormous error when calculating its properties. For the generation of a nodal network which could fit the presented problem, it was defined the mesh would be non uniformal with properties shown in Tab 4.9 which generated 3180375 cells. The presented mesh definition is used to all simulations later. Tab 4.9 – 3D mesh properties used for the simulations its definition. Property Value Minimum size 1 mm Avarage size 1.8 mm Maximum size 36 mm Maximum growth rate 1.4 The generated mesh is in accordance with the discretization method of finite differences method (FDM). This discretization method approximates derivatives with finite differences for both spatial and temporal domains [41]. This descretization method convert non-linear partial different equations such as Navier-Stokes and energy equation as shown in chapter 4.3, respectively into a linear system of equations which can be solved iteratively [42, 43]. In order to have accurate Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 62 results, it requires a small element size either to delineate the component’s commplexity and to compensate for the low accuracy discretization method when compared to Finite Volume Method (FVM) [44]. The system’s discretization schematics is shown in Fig 4.26, where the determination of the central node property is based on neighbouring nodes. To determine the first and second derivative in the respective orientation, equations [4.15] and [4.16] are used, respectively [45]. Fig 4.26 – 2D staggered grid for the simulation problem with the respective connectivity of each node (adapted from [31]). 𝑑𝑃 𝑑𝑦|+=𝑃𝑖,𝑗+1−𝑃𝑖,𝑗 ∆𝑦 [4.15] 𝑑2𝑃 𝑑𝑦2=𝑑𝑃 𝑑𝑦|+−𝑑𝑃 𝑑𝑦|− ∆𝑦 =𝑃𝑖,𝑗+1−2𝑃𝑖,𝑗+𝑃𝑖,𝑗−1 ∆𝑦2 [4.16] Thermal model The casting process is a transient heat transfer problem, therefore in heat conduction equation it will be needed to add a transient term. Discretizing heat transfer equation into a finite differences problem within the same volume, as it is presented in Fig 4.26, it is possible to get equation [4.18] [46, 47]. 𝜙=∆𝑦∆𝑥𝜌𝑐𝑝(𝑇𝑖,𝑗 𝑛−𝑇𝑖,𝑗 𝑛−1) ∆𝑡 [4.17] 𝑘 ∆𝑥(𝑇𝑖−1,𝑗 𝑛−2𝑇𝑖,𝑗 𝑛+𝑇𝑖+1,𝑗 𝑛)+𝑘 ∆𝑦(𝑇𝑖,𝑗−1 𝑛−2𝑇𝑖,𝑗 𝑛+𝑇𝑖,𝑗+1 𝑛)=𝜙 [4.18] In equation [4.17] and [4.18], 𝑖 and 𝑗 is the node position in the staggered grid and 𝑛 is the number of the time-step. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 63 If it’s the case of node is in a surface, as it is presented in Fig 4.27, the convective term has to be introduced in order to describe heat transfer through the surface, equation [4.19]. In case of an interface mold/casting metal during the injection phase, the convective term, ℎ, is calculated automatically. During solification, user has to define it. ∆𝑦 ℎ(𝑇𝑖,𝑗−𝑇∞)+2𝑘 ∆𝑥(𝑇𝑖+1,𝑗 𝑛−𝑇𝑖,𝑗 𝑛)+𝑘 ∆𝑦(𝑇𝑖,𝑗−1 𝑛−2𝑇𝑖,𝑗 𝑛+𝑇𝑖,𝑗+1 𝑛)=𝜙 [4.19] Fig 4.27 - Case of a node in a surface (adapted from [42]). The equations [4.17], [4.18] and [4.19], are valid to solids, this means that are only appliable for the heat transfer within the mold domain and casting domain during solidification. During the injection phase, energy equation within the fluid domain is incorporated with NavierStokes equations, since most of the heat transfer occurs due to convection and not conduction. Flow model The flow model is only useful for during the injection phase where the main computational challenge is the solving Navier-Stokes equations. During this phase heat transfer is also present, however only the Navier-Stokes equations will be presented, since energy conservation equation is similar to the one shown in thermal model. The need of the usage of staggered grid is appear when it is required to discretize NavierStokes equation. The pressure term is calculated in the node, while the local velocities are calculated in the intermidiate points (𝑢 󰇍 𝑛). Lately, it is calculated in the nodes (𝑈 󰇍 󰇍 𝑛) the avarage local velocities [28]. 𝑈 󰇍 󰇍 𝑛+1−𝑈 󰇍 󰇍 𝑛 ∆𝑡 +(𝑈 󰇍 󰇍 𝑛∙∇)𝑈 󰇍 󰇍 𝑛=𝜇∇2𝑈 󰇍 󰇍 𝑛+∇𝑝+𝜌𝑔 [4.20] Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 64 Explicit and Implicit methods Both models presented before have different methods to solve the transient term, which can be explicit and implicit [48]. The difference between these two methods is that the explicit method uses a variable value in (𝑛)𝑡ℎ time-step to calculate it’s value in timestep (𝑛+1)𝑡ℎ. Meanwhile the implicit method solves the differencial in order to find the value when the transient term of the previous timestep, (𝑛)𝑡ℎ, equals the weight of system’s variables of the same instant, (𝑛+1)𝑡ℎ. For a dynamic problems, the method used is the explicit one, where the velocity in the timestep (𝑛+1)𝑡ℎ is determined adding the value of the velocity in instant (𝑛)𝑡ℎ to its time variation, equation [4.21]. 𝑈 󰇍 󰇍 𝑛+1=𝑈 󰇍 󰇍 𝑛+(−(𝑈 󰇍 󰇍 𝑛∙∇)𝑈 󰇍 󰇍 𝑛+𝜇∇2𝑈 󰇍 󰇍 𝑛)∆𝑡+𝜌∇𝑈 󰇍 󰇍 𝑛 [4.21] In equation [4.20], the term 𝜌∇𝑈 󰇍 󰇍 𝑛 is the pressure gradient in Navier-Stokes equation. As it is possible to verify, 𝑈 󰇍 󰇍 𝑛+1 is calculated entirely based on the previous step (𝑈 󰇍 󰇍 𝑛). For the Heat transfer problem during filling, QuikCast also uses the explicit method. In other hand, for the solificiation process, QuikCast uses implicit method. In this method, equations [4.17] and [4.18] are used to determine the temperature in timestep (𝑛)𝑡ℎ. This method is used, since it has a smaller error and computation time is the same, since the matrix for the linear problem remains constant. Both methods can be attributed to a factor in which represents the (𝑛+1)𝑡ℎ timestep’s weight in solution, equation [4.22] to [4.24]. To an explicit method, 𝑓 is 0. For the implicit method, the factor 𝑓 is different from 0, with two exceptions. These exceptions are for 𝑓 equals to 0.5 which represents Crank-Nicolson’s method and 𝑓 equals to 1 which represents a total implicit method [49]. 𝑘 ∆𝑥(𝑇𝑖−1,𝑗 𝑛−2𝑇𝑖,𝑗 𝑛+𝑇𝑖+1,𝑗 𝑛)+𝑘 ∆𝑦(𝑇𝑖,𝑗−1 𝑛−2𝑇𝑖,𝑗 𝑛+𝑇𝑖,𝑗+1 𝑛)=𝜙𝑛 [4.22] 𝑘 ∆𝑥(𝑇𝑖−1,𝑗 𝑛−1 −2𝑇𝑖,𝑗 𝑛−1+𝑇𝑖+1,𝑗 𝑛−1)+𝑘 ∆𝑦(𝑇𝑖,𝑗−1 𝑛−1 −2𝑇𝑖,𝑗 𝑛−1+𝑇𝑖,𝑗+1 𝑛−1)=𝜙𝑛−1 [4.23] (1−𝑓)𝜙𝑛−1+𝑓𝜙𝑛=∆𝑦∆𝑥𝜌𝑐𝑝(𝑇𝑖,𝑗 𝑛−𝑇𝑖,𝑗 𝑛−1) ∆𝑡 [4.24] Crank-Nicolson’s method has an even lower error to small timesteps than implicit method [42], however its computation is higher than the implicit method, since the calculation of each Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 65 temperature requires the temperature of two timesteps of neighbour nodes, (𝑛)𝑡ℎ and (𝑛+1)𝑡ℎ, while implicit only requires one. 4.5. HIGH PRESSURE DIE CASTING SIMULATION CONDITIONS The simulation of the high pressure die casting process needs 2 types of simulation. The first one is the cycling simulation and the second one the actual high pressure die casting simulation. A die cycling simulation is required to simulate the impact of previous cycles in the numeric model. The number of cycles it required validate the process was 20 cycles [50]. After 20 Cycles it was considerated that the thermal variation across the mold would remain the same spatially and temporally when compared to the same instant of the previous cycle. In this cycles will be only considerated the thermal module to reduce computational time without trading it off for accuracy. Using the flow module to simulate the effect of the casting alloy thermal gradient would increase massively the computing time. The cycling simulation would increase from 90 minutes to 16 hours and 47 minutes, an 2675% increase, without considering possible interruptions. The mesh used for both simulations was presented in Tab 4.9. After the cycling simulation, it was possible to simulate a real injection with the flow model and solidification to verify the solution. 4.5.1. DIE CYCLING SIMULATION The non consideration of previous cycles not only does not describe the casting process after each shot accuratly, but it also influence negatively the prediction of defects. Since it does not consider that the mould’s thermal field is affected by the thermal cycles. Consequently, increasing the apperence of shrinkage porosity defect in the casting postprocessing. The casting simulation requires the definition of the initial conditions and boundary conditions which QuikCast’s designation is volume manager and process conditions, respectively. The definition of the thermal cycle of the process requires the definition of the transient thermal conditions. To define such conditions, it’s needed to define it’s value, starting and ending time (Fig 4.28). These process conditions are adjust according to the available technology such as die spray machines. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 66 Fig 4.28 - Thermal scheme of the process. The solidification mentioned in Fig 4.28 is referent to the injection and solification of the casting alloy. As flow module is disable for the cycling, only the solidification stage is mentioned in the die cycling simulation. The open mold condition defines the moul’s condition when it is open after the casting is removed. Therefore, the external temperatures for both forms of heat transfer (radiation and convection) were defined as 25ºC. The emissivity was considerated 0.9 [31] for steel-steel surfaces (between mold plates) and 0.35 to aluminium-steel surfaces (between mold plates and casting). The convection heat transfer coefficient (ℎ) was calculated based on natural convection [51]. The approximation of its value was based on Grashof, Prandlt, Rayleigh and Nusselt number adimensional numbers [31], equations [4.25] to [4.28], respectively. 𝐺𝑟=𝐿3𝜌2𝑔𝛽∆𝑇 𝜇2 [4.25] 𝑃𝑟=𝜇𝐶𝑝 𝑘 [4.26] 𝑅𝑎=𝐺𝑟𝑃𝑟 [4.27] 𝑁𝑢=ℎ𝐿 𝑘= ( 0.825+0.387𝑅𝑎1 6 [1+(0.492 Pr )9 16]8 27 ) 2 [4.28] To calculate the ℎ,Casting simulation, initially it is calculated the Grashof number and Prandlt, then the Rayleigh’s. Lately, it can be calculated Nusselt number to calculate the convection coefficient, which value is 5.3 𝑊 𝑚2𝐾. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 67 The close mold condition defines the condition when the mold is closed before the injection. In this condition, it was only considerated a heat transfer between the air inside the mold and the mold surface. Relatively to the open mold condition, it was only considerated an equal convection heat transfer coefficient and it was not considerated any ratiation, since both surfaces have the same temperature. The cooling system starts at the beginnig of the injection and would last one second after the end of the solidification period. It was considerated that the cooling oil would have an avarage temperature of 160ºC and a convection heat transfer of 8595 𝑊 𝑚2𝐾[22]. This convection heat transfer coefficient is based on Nusselt number using Colburn Formula, equation [4.29]. 𝑁𝑢=0.023𝑅𝑒0.8𝑃𝑟0.33 [4.29] The cooling system is based on a new technology (conformed cooling channels) which main goal is to optimize heat transfer from the casting to the cooling fluid. Therefore, a better numerical model would be required in order to predict accurately the local heat transfer coefficient and the effect of the fuild’s temperature increase along the channel instead of an analitical approximation of constant temperature. The application of a die coating keeps the die cooled increasing mold’s lifespan [52, 53], lubricated to facilitate the casting’s removal [54], enhance surface quality and dimensional accuracy [55]. These types of die coating are a moisture with the a stripping agent and a cooling agent. The cooling agent can be water, oil or air. The stripping agent responsible for the lubrication of the die cavity can be graphite, TiO2, Al2O3 and Fe2O3, etc [56]. The die coating used is a waterbased moisture which stripping agent is graphite. The film size of a the die coating is typically between 0.1 to 0.3 mm [40] with thermal porperties shown in Tab 4.10. Tab 4.10 - Thermal properties required by QuikCast to define die coating interface (adapted from [22]). Property Value [unit] Thermal conductivity 1.05 [W/mK] Aderence resistance 10-5 [W/m2K] Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 68 The systems initial conditions are shown in Fig 4.29. It is only relevant to define for the first cycle, initial conditions for the other cycles are based on thermal fields of the previous cycle. The inserts were defined as a mold with a filling volume of 100% with tool steel 1.2709 at 300ºC. The material of the shot sleeve and mold plates was H13. The casting alloy volume which would originate the runners, parts and overflows had a filling volume of 100% at 710 ºC, while the volume casting alloy in the shot sleeve volume had a 0% filling. Even though the casting alloy volume in the shot sleeve was 0%, it was required to define it for casting simulation. If the third stage effect is used during solidification only simulation, QuikCast considers that the shot sleeve’s volume is full, even though it was defined as 0% full. This happens, because applying pressure only makes sense if there is any metal. To overcome this difficulty, an extra boundary condition is applied to impose that there is not any liquid. This boundary condition was to impose that casting alloy in shot sleeve to be at 300ºC (at same temperature than the mold’s), to minimize error. It was choosen the usage of the third stage effect since it is the best way do describe solidification. Fig 4.29 - Initial conditions to the die cycling simulation. After the cycling simulation, the resulting thermal field will be used as a start temperature of the mold. It is also possible to analise temperature during different times of the cycle to guarantee the mold’s local temperature does not reach critical temperatures. From the Fig 4.30 to Fig 4.33, are presented the thermal’s field of the last cycle in different instants. The chosen instants were at 0, 12, 15 and 25 seconds. The instant 0 second’s thermal field is the same as the last instant from the previous cycle. The instant 7 second’s thermal field Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 69 is referred to the moment after applying the die spray, which main source of thermal energy loss was the cooling agent. The instant 25 represents the mold’s temperature before the next injection. In initial instant of the last cycle, mold’s thermal field is presented in Fig 4.30. It is possible to notice very low temperatures which are surface of the cooling channels. In instant 12 seconds, Fig 4.31, die cavity’s surface is lowest during the process and its temperature gradient is the highest. This happens due to the cooling agent of the die spray during die coating, since a relatively high heat transfer is happening in the die cavity, its surface temperature is abruptally reduced. Once the the die coating process is finished and it is given enough time, die cavity’s surface temperature increases due to thermal conductivity of the thermal energy in present in the mold. (a) (b) Fig 4.30 – Mold's Thermal field in instant 0 seconds. (a) - Fix mold plate; (b) - Mobile mold plate. (a) (b) Fig 4.31 – Mold's Thermal field in instant 12 seconds. (a) - Fix mold plate; (b) - Mobile mold plate. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 70 (a) (b) Fig 4.32 – Mold's Thermal field in instant 15 seconds. (a) - Fix mold plate; (b) - Mobile mold plate. (a) (b) Fig 4.33 – Mold's Thermal field in instant 25 seconds. (a) - Fix mold plate; (b) - Mobile mold plate. 4.5.2. CASTING SIMULATION The injection simulation requires the definition of the physical system with flow conditions in order to evaluate the process’s dynamic. Therefore, the initial conditions of volumes and boundary conditions will vary of those defined in chapter 4.5. However last cycle’s mold’s thermal field will be used in the injection simulation, thermal field shown in Fig 4.33. At the initial instant of the casting process, it was stipulated that both the shot sleeve and the filling cavity pertained exclusively to the “Alloy” volume, meaning these volumes would only contain the molten material, alluminum alloy AlSi9Cu3. The pouring temperature of the “Alloy” was 710ºC. The shot sleeve was filled to a capacity of 29% full, whereas the filling cavity remained Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 77 [11] ENAC-AlSi9Cu3(Fe) / ENAC-46000 - SteelNumber - Aluminium equivalent, chemical composition, properties. [Online]. Available: https://www.steelnumber.com/en/steel_alloy_ composition_eu.php?name_id=1244 (accessed: Jul. 5 2024.533Z). [12] P. C. Bollada, P. K. Jimack, and A. M. Mullis, "Faceted and dendritic morphology change in alloy solidification," Computational Materials Science , vol. 144, no. 3, pp. 76–84, 2018, doi: 10.1016/j.commatsci.2017.12.007. [13] ASM TECHNICAL BOOKS, Symbols, Abbreviations, and Intermetallic Phase Designations | Aluminum-Silicon Casting AlloysAtlas of Microstructures | Technical Books | ASM Digital Library (accessed: Jul. 17 2024.508Z). [14] Q. Miao et al., "Comparative study of microstructure evaluation and mechanical properties of 4043 aluminum alloy fabricated by wire-based additive manufacturing," Materials & Design , vol. 186, p. 108205, 2020, doi: 10.1016/j.matdes.2019.108205. [15] NADCA, NADCA Product Specification Standards for Die Casting: Aluminum, AluminumMMC, Copper, Magnesium, Zinc and ZA Alloys, 9th ed. Illinois, United States of America: NADCA, 2015. [16] NADCA, NADCA Gating Manual . Illinois, United States of America: NADCA. [17] J. K. Pancholi, B. Padhiyar, and V. Sutariya, "Design and Analysis of Die Casting Die With Conformal Cooling Channel," International Journal of Engineering and Computer Science , vol. 9, no. 3. [Online]. Available: https://www.researchgate.net/publication/344235476_ Design_and_Analysis_of_Die_Casting_Die_With_Conformal_Cooling_Channel [18] Z.-C. Lin and M. H. Chou, "Design of the Cooling Channels in Nonrectangular Plastic Flat Injection Mold," vol. 2, 2002. [19] E. Vojnová, "The Benefits of a Conforming Cooling Systems the Molds in Injection Moulding Process," Procedia Engineering , vol. 149, pp. 535–543, 2016, doi: 10.1016/j.proeng.2016.06.702. [20] S. Feng, A. M. Kamat, and Y. Pei, "Design and fabrication of conformal cooling channels in molds: Review and progress updates," International Journal of Heat and Mass Transfer , vol. 171, p. 121082, 2021, doi: 10.1016/j.ijheatmasstransfer.2021.121082. [21] A. Souza et al., "Thermal Contact Resistance between Mold Steel and Additively Manufactured Insert for Designing Conformal Channels: An Experimental Study," JMMP , vol. 6, no. 5, p. 99, 2022, doi: 10.3390/jmmp6050099. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 78 [22] ESI Group, QuikCast | Casting Simulation Software . France: ESI Group. [Online]. Available: https://myesi.esi-group.com/ [23] ESI Group, ProCAST | Casting Simulation Software : ESI Group. Accessed: Jun. 13 2024.038Z. [Online]. Available: https://www.esi-group.com/products/procast [24] FLOW-3D, Free Gridding Saves Time. [Online]. Available: https://www.flow3d.com/ resources/cfd-101/free-gridding/ (accessed: Jun. 13 2024.523Z). [25] NATIONAL AERONAUTICS AND SPACE ADMINISTRATION, Navier-Stokes Equations. [Online]. Available: https://www.grc.nasa.gov/www/k-12/airplane/nseqs.html (accessed: Mar. 23 2024.132Z). [26] Wikipedia, Navier–Stokes equations. [Online]. Available: https://en.wikipedia.org/w/index.php?title=Navier–Stokes_equations&oldid=1217632972 (accessed: May 6 2024). [27] Glenn Research Center | NASA, Newton’s Laws of Motion | Glenn Research Center | NASA. [Online]. Available: https://www1.grc.nasa.gov/beginners-guide-to-aeronautics/ newtons-laws-of-motion/ (accessed: May 6 2024.359Z). [28] Department of Mechanical Engineering, Columbia University, A Numerical Analysis for 2-D Navier-Stokes Equations by FDM and FVM. [Online]. Available: https://www.columbia.edu /~ll2998/documents/A%20Numerical%20Analysis%20for%202-D%20NavierStokes%20Equations%20by%20FDM%20and%20FVM_Lei%20LEI.pdf (accessed: Jun. 11 2024). [29] Wikipedia, Poisson's equation. [Online]. Available: https://en.wikipedia.org/w/index.php? title=Poisson's_equation&oldid=1214748818 (accessed: Jun. 17 2024). [30] Wikipedia, Continuity equation. [Online]. Available: https://en.wikipedia.org/w/index.php? title=Continuity_equation&oldid=1221543206 (accessed: May 6 2024). [31] J. H. Lienhard, IV and J. H. V. Lienhard, A Heat Transfer Textbook , 2001. Accessed: Jun. 13 2024. [Online]. Available: http://web.mit.edu/lienhard/www/ahtt.html [32] Cadence CFD Solutions, Energy Equations: The Navier-Stokes Method of Analysis | System Analysis Blog | Cadence. [Online]. Available: https://resources.systemanalysis.cadence.com/blog/msa2022-energy-equations-the-navier-stokes-method-of-analysis (accessed: Mar. 23 2024.285Z). Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 79 [33] COMSOL Multiphysics, Heat Transfer: Conservation of Energy. [Online]. Available: https:// www.comsol.com/multiphysics/heat-transfer-conservation-of-energy (accessed: Jun. 13 2024.559Z). [34] S. K. Mohammed, M. H. Arbo, and L. Tingelstad, "Leveraging model based definition and STEP AP242 in task specification for robotic assembly," Procedia CIRP , vol. 97, no. 4, pp. 92–97, 2021, doi: 10.1016/j.procir.2020.05.209. [35] Wikipedia, STL (file format). [Online]. Available: https://en.wikipedia.org/w/index.php?title= STL_(file_format)&oldid=1210489843 (accessed: Apr. 5 2024). [36] Adobe Acrobat, What are IGES files and how do you open them? | Adobe. [Online]. Available: https://www.esi-group.com/sites/default/files/resource/publication/1284/ finegi-ans.pdf (accessed: Apr. 5 2024.094Z). [37] Scratcha Pixel, Geometry. [Online]. Available: https://www.scratchapixel.com/lessons/ mathematics-physics-for-computer-graphics/geometry/transforming-normals.html (accessed: Apr. 5 2024.481Z). [38] Two-Dimensional Conduction: Finite-Difference Equations and Solutions: Chapter 4 - Sections 4.4 and 4.5. [Online]. Available: https://www.visualslope.com/Library/FDM-forheat-transfering.pdf (accessed: Jun. 4 2024). [39] I. P. Villalobos and I. U. Martinez, Guía Práctica de QuikCASTVisual Environment: Para el proceso de HPDC , 2019. [40] ESI Group, Quikcast reference manual . France: ESI Group, 2014. Accessed: 2014. [41] Wikipedia, Finite difference method. [Online]. Available: https://en.wikipedia.org/wiki/ Finite_difference_method (accessed: May 6 2024.526Z). [42] P. Lobarinhas, Mecânica Computacional (accessed: Jun. 4 2024). [43] J. C. F. Teixeira, Métodos Numéricos em Transferência de Calor (accessed: Jun. 5 2024). [44] Bengt Fornberg, "Finite difference method," Scholarpedia , vol. 6, no. 10, p. 9685, 2011, doi: 10.4249/scholarpedia.9685. [45] M. T. T. Monteiro, Métodos Numéricos: exercícios resolvidos aplicados à Engenharia e outras Ciências (accessed: May 6 2024.526Z). [46] J. E. Marsden, L. Sirovich, S. S. Antman, J. S. Hesthaven, and T. Warburton, Nodal Discontinuous Galerkin Methods . New York, NY: Springer New York, 2008. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 80 [47] S. Patankar and D. Spalding, "A calculation procedure for heat, mass and momentum transfer in three-dimensional parabolic flows," International Journal of Heat and Mass Transfer , vol. 15, no. 10, pp. 1787–1806, 1972, doi: 10.1016/0017-9310(72)90054-3. [48] Wikipedia, Explicit and implicit methods. [Online]. Available: https://en.wikipedia.org/w/ index.php?title=Explicit_and_implicit_methods&oldid=1076468513 (accessed: Jun. 17 2024). [49] Wikipedia, Crank–Nicolson method. [Online]. Available: https://en.wikipedia.org/w/index.php?title=Crank–Nicolson_method&oldid=1188507267 (accessed: Jun. 17 2024). [50] E. Anglada, A. Meléndez, I. Vicario, E. Arratibel, and G. Cangas, "Simplified Models for High Pressure Die Casting Simulation," Procedia Engineering , vol. 132, pp. 974–981, 2015, doi: 10.1016/j.proeng.2015.12.585. [51] QuickField.com, Natural convection coefficient calculator. Accessed: Jun. 20 2024.162Z. [Online]. Available: https://quickfield.com/natural_convection.htm [52] N. D. Campos Neto, A. L. Korenyi-Both, C. Vian, S. P. Midson, and M. J. Kaufman, "The development of coating selection criteria to minimize die failure by soldering and erosion during aluminum high pressure die casting," Journal of Materials Processing Technology , vol. 316, p. 117954, 2023, doi: 10.1016/j.jmatprotec.2023.117954. [53] V. Nunes, F. Silva, M. F. Andrade, R. Alexandre, and A. Baptista, "Increasing the lifespan of high-pressure die cast molds subjected to severe wear," Surface and Coatings Technology , vol. 332, pp. 319–331, 2017, doi: 10.1016/j.surfcoat.2017.05.098. [54] FOSECO and VESIVIUS, DYCOTE: Manual. [Online]. Available: www.vesuvius.com\foundry (accessed: Jun. 21 2024). [55] Sunrise, "Die Spray & Lubricate," Sunrise Metal - Aluminium Die Casting Expert , 28 Apr., 2024. https://www.sunrise-metal.com/die-spray-and-lubricate-technology (accessed: Jun. 21 2024.298Z). [56] X. Dong, X. Zhu, and S. Ji, "Effect of super vacuum assisted high pressure die casting on the repeatability of mechanical properties of Al-Si-Mg-Mn die-cast alloys," Journal of Materials Processing Technology , vol. 266, pp. 105–113, 2019, doi: 10.1016/j.jmatprotec.2018.10.030. Chapter 4 – Casting Simulation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 81 [57] Become a master of CAD Design, What is injection phase of HPDC Die Casting | Die Casting. [Online]. Available: https://www.youtube.com/watch?app=desktop&v= SG3xR9d7lPM (accessed: Sep. 5 2024.122Z). [58] J. G. Fairman, Experiment on Fluids. [Online]. Available: https://www.grc.nasa.gov/www/k12/WindTunnel/Activities/exper_on_fluids.html (accessed: Apr. 15 2024.220Z). [59] N. Hall, Bernoulli's Equation. [Online]. Available: https://www.grc.nasa.gov/www/k-12/ airplane/bern.html (accessed: Apr. 15 2024.395Z). [60] Wikipedia, Pascal's law. [Online]. Available: https://en.wikipedia.org/w/index.php?title= Pascal's_law&oldid=1211459295 (accessed: Mar. 23 2024). [61] P. Szalva and I. N. Orbulov, "The effect of vacuum degree on the porosity and mechanical properties of die cast AlSi 9 Cu 3 (Fe) alloy," Resolution and Discovery , vol. 3, no. 2, pp. 33–38, 2018, doi: 10.1556/2051.2018.00061. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 82 Chapter 5 - GATING DESIGN VALIDATION The gating design validation consist in evaluating the CAE results which predict the existance of defects within the component’s volume and the postprocessing of process variables such as injection pressure for process’s validation. Gating design parameters such as flow angles, filling time, segments’ division defined in the gating design will be checked to determine if these were appropriate and provide H.E.L.P. Die Casting with new knowledge for future gating designs. Lately, time spent doing by-hand calculations and modelling were performed to determine the real positive impact of the developed software. In chapter 4.3 was shown both numeric models to describe the casting simulation. In chapters 4.5.1 and 4.5.2, it was described the process conditions for die cycling simulation and injection simulation. This previous knowledge to describe the casting simulation resulted in various CAE results which shall be evaluated and compared with product specifications and available technology. Without forgeting the CAE importance, the results given by QuikCast have to be compared with the result of a real casting process. 5.1. CYCLING SIMULATION The cycling simulation described in chapter 4.5.1 originated a thermal field which was shown previously. However, in order to validate such simulation, it was necessary to monitor some point in the die, shown in Fig 5.1. It was choosen symmetrical points in each component, in order Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 83 to guarantee that a different attack position defined would not result in a completely different shrinkage defect location. After monitoring these points, it was possible to track its temperature throughout each cycle to find its convergence temperature. Fig 5.1 - Mould’s position of the tracking points. The opposite points pairs are P01/P14, P02/P13, P03/P12 and P04/P11 and the "Point_Ce_1” which is located at midpoint of both casting components. In Fig 5.2 is shown the temperature variation of the each point during each cycle of the die cycling simulation. Fig 5.2 - Temperature oscillation within the die cycling simulation. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 84 It was possible to evaluate that the die temperature variation through the die cycling simulation converges. In a more detail analysis it was exclusively evaluated the maximum temperature of each point. In Fig 5.3 (a), (b) are shown the maximum temperature of each cycle to the point P01/P14 and P04/P11, P03/P12 and P02/P13 and Point_Ce_1, respectively. In Fig 5.3 (a), the temperature difference in pair P01/P14 (13ºC) is lower than the pair P04/P11’s (31ºC), this is caused by the caused by the proximity to the cooling channel in critical region 1 (defined in chapter 4.2). The cooling channel influence on die cavity’s temperature is higher in critical region 1 than the one in critical region 2, due to the cooling channel area. Fig 5.3 - Maximum temperature in each cycle. (a) - maximum temperature in each cycle in points P01, P04, P11 and P14; (b) - maximum temperature reach in each cycle in points P03, P12, P02, P13 and Point_Ce_1. Even though the temperature variation throughout the die cycling simulations converges, a relative analysis was done (Fig 5.4) to evaluate its relative variation with a convergence threshold of 1%. It is possible to verify that the points in die cavity (every point exepct Point_Ce_1) converges after the 4th cycle, however the central point (Point_Ce_1) converges after the 9th cycle. This point is exposed to a lower temperature gradient during each cycle, since it is not in contact with the casting alloys. Therefore, this point would reflect the mold’s thermal inertia and it’s equilibrium point between the process’s heat outflux (to the environment and cooling channels) and the heat influx (casting alloy). Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 85 Fig 5.4 - Relative variation of maximum temperature in each cycle and its defined threshold of 1% for acceptance. In a detailed view of temperature variation of the last cycle, Fig 5.5, it is noticeable the real effect of each process’s phase on die’s temperature. The die coating phase causes the highest temperature reduction. This happens due to the high value of heat transfer convection coefficient. However, this process’s phase is yet divided into two different step, the spraying of die coating and the air blowing which removes excess of mixture. Consequently, die cavity’s surface temperature variation is different in both stages. Lately, between the last instant of spraying the coating and close mold the die cavity’s surface temperature increases due to the its thermal inertia (as it was concluded in chapter 4.5.1). Fig 5.5 - Association between each step of the process and the die cycling defined previously. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 86 5.2. CASTING SIMULATION In the casting simulation, it is possible to predict possible shrinkage defects, air entrainment, a qualitative analysis of the existance of cold shuts, check if the flow pattern and segment division was as predictable during gating design phase. Furthermore, it is possible to determine if the overflow position was correct. Two extra gating designs were made in order to study the effect of variables such as ingate thickness and flow angle and the ingate continuity. The final gating design of the component is shown in 4.2, which will be used as final solution. The gating design – final solution will be taken in consideration as the final solution solution. The results plotted in the postprocessing of the casting simulation in order to validate the gating design are flow pattern, shrinkage porosity, air entrainment and cold shuts location (this result was evaluated based on a user defined result). Fluid’s vorticity could be evaluated to determine the location’s which are more susceptible to air entrainment, however it would require a finer mesh than the one used. Gating designs A and B analysis will based on shrinkage porosity and air entrainment. All solutions present a casting yield of 58%. he small changes described in the following chapter amount to less than a 1% variation. Gating design – Final solution The flow pattern evaluation is hard to evaluate due to geometric complexity therefore, its evaluation is done based on the filling sequency of each region in important frames to verify overflows position. In Fig 5.6, it is shown the filling sequency at 0.6004, 0.6029, 0.6116 and 0.6152 seconds. The validation of overflows position is based in avoiding the encircled air region which will prevent the air from leaving the die cavity.In frames shown in Fig 5.6 (a) and (b), it is verifies the position of the overflows 12, 13 and 14. The position of overflow 1, 2, 3, 5 and 6 are validated in frame (c). Overflows 4, 7, 8 and 9 are validated in frame (d). Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 93 When analysing the filling sequency of this gating design, it is possible to notice that the ingate does not possess a uniform filling profile, Fig 5.16. This is resulted of an high flow angle which can not be fulfilled due to fluid’s momentum. Even though the ingate design A goal was to improve the flow distribution to reduce the swirl, this design proved to have a flow angle too large. Being therefore an overall result worst than the final solution. Fig 5.16 – Ingate’s last region to fill. Gating design – B Gating design B ingate tunning is presented in Fig 5.17. When compared with reference ingate, the ingate flow angle on the side closest to the biscuit is 30º inward. This orientation was choosen to reduce the difficulty in filling the ingate (when compared to gating design A), which causes backflow and vorticity. Meanwhile, the opposite side of the ingate has a 90º angle. The cross section’s dimensions are 1.6×26 mm2. Fig 5.17 - Ingate modification of gating design B. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 94 Shrinkage porosity results are presented in Fig 5.18. It is possible to verify that the shrinkage volume is higher than the reference’s. This happens as a consequence of a less thick ingate which will solidify earlier than the reference gating design (cross section is 1.8×23 mm2). Similarly to the gating design A, in this gating design the appearance of new regions (when compared with reference) with shrinkage porosity is resulted of descontinuity of the liquid state domains within the component (Fig 5.18), which will prevent the compensation of volume change. Fig 5.18 - Shrinkage porosity at the end of the solidification process. Fig 5.19 - Descontinuity of liquid-state domains between the component and runner. In Fig 5.20, it is possible to conclude that this gating design reduce the encirclement of air volume present in the die cavity, since the air entrainment in locate almost exclusively in the critical region 1 and in the pin region. The avarage air entrainment quantity of air in the casting (including overflow) is 0.0004 g/cm3 and 0.0005 g/cm3 for component 1 and 2, respectively. This values Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 95 represent a reduction in air entrainment of 43% and 38% (for component 1 and 2, respectively) compared to the reference. This solution possesses a lower air entrainment in both components, but had more shrinkage porosity when compared to the final solution. Fig 5.20 - Air entrainment in the casting. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 96 5.3. DYNAMICS OF THE PROCESS Even though previous calculations were made in order to predict the required dynamic pressure more accurately, lately it would be required to verify experimently. The theoric value of injection of the machine was determined in chapter 4.5.2 without considering the inertial forces and losses of energy. These variables will generally increase the required injection force to grant the casting metal the required ingate velocity (20 m/s). In Fig 5.21, it is presented the injection pressure required to the process. Local peaks in in the simulated injection pressure are resulted of geometric complexity of the gating design and error cause by difference time-step sizes taken by the solver during processing. Fig 5.21 - Injection pressure evolution with piston's displacement. An optical evaluation of the ingate speed would be less accurate when compared to a macroscopic variable (such as injection pressure). This happens cause ingate velocity can only be evaluated when the runner is completely full (and its instant), geometric configuration of the gating design would increase kinetic energy losses, mesh refinement, velocity gradient in the ingate. It is very clear that the the local peaks in the graphics are resulted of a splash of the casting metal in the piston’s surface. However, these splashes are irrelevant to the overall injection cycle and do not cause air entrainment (problem solver in chapter 4.5.2). A relation between the filling and pressure is presented in Fig 5.22. It is noticeable three significant relative increases in injection pressure can be identified during the process.. The first occurs within the 1st phase of the piston’s advance when the main runner starts to be filled up, Fig 5.22 (a). This increase happens due to the decreasing of area between the shot sleeve and main Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 97 runner (minor loss). The second occurs during the 2nd phase , Fig 5.22 (b), where the flow rate increases drastically and the reduction of section area (ratio runner/ingate to guarantee a completely full ingate). The third , Fig 5.22 (c), is related to the transition to the increase of the viscous forces and the transition to the 3rd stage. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 98 Fig 5.22 - Piston's position relation to the injection pressure. (a) – Shot sleeve at full capacity; (b) – Piston’s 2nd advancing stage; (c) – Piston’s 3rd advancing stage. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 99 As verified in Fig 5.22, the kinetic energy losses happen in two distinguished locations, at the runners and at ingate. Using the minimum mean square error method [2], it is possible to determine that the energy loss for this gating design using the casting alloys AlSi9Cu3 during the 1st and 2nd stage is 0.16 and 0.35, respectively. Therefore, it is possible to divide piston’s second advancing phase in two stages as it is shown in Fig 5.23 (scenario (b) and (c)), described by two step functions. Lately, this information can be stored in H.E.L.P. Die Casting database. Fig 5.23 - Two step discretization of the machine based on piston's displacement. The difference between the pressure at distance 295 mm of the piston’s displacement in Fig 5.22 and Fig 5.23 (beginning of the second phase) is a consequence of the difference in geometric complexity. This happens due to the runner is geometrically less complex than the component which causes minor losses. It is noticable that the injection pressure mentioned previously corresponds to the sum of the dynamic pressure and static pressure during the second advancing stage of the piston and not to the static pressure during the third stage. The pressure vs displacement is a macroscopic variable that allows to evaluate a processes dynamics globally avoiding the velocity gradient at the ingate [3]. This also plays a crucial role when determining the mesh convergence, since it is possible to analise a tendency of the mesh refinement. Chapter 5 – Gating Design Validation -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 100 5.4. CHAPTER REFERENCES [1] Adam Augustyn, Pascal’s principle | Definition, Example, & Facts. [Online]. Available: https://www.britannica.com/science/Pascals-principle (accessed: Jul. 31 2024.160Z). [2] Wikipedia, Minimum mean square error. [Online]. Available: https://en.wikipedia.org/w/ index.php?title=Minimum_mean_square_error&oldid=1212895420 (accessed: Jul. 22 2024). [3] W. Heywang, K. Lubitz, and W. Wersing, Piezoelectricity: Evolution and Future of a Technology . Germany: Springer, 2008. Accessed: Jul. 24 2024.013Z. [Online]. Available: https://books.google.pt/books?id=8KyawZ92eIEC&pg=PA534&lpg=PA534&dq= macroscopic+variable+in+cae+software+engineering&source=bl&ots=JxmAz2tZa1&sig= ACfU3U2U1JXh9N2wbD3swCJFFCm3ewgyNw&hl=pt-PT&sa=X&ved= 2ahUKEwj2ksnrsMCHAxVQcKQEHaluBYoQ6AF6BAgnEAM #v=onepage&q=macroscopic%20values%20like%20displacement%20and%20force%20and%2 0internal%20loading&f=false Chapter 6 – Conclusion -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 101 Chapter 6 - CONCLUSION The conclusion of this dissertation summarizes the project giving it an overview. Therefore, it will be divided into two main topics “Summary” and “Future work”. The dissertation focused on the virtualization of the process of gating design for high pressure following NADCA methodology. The main challenges faced was during the development and study of the interface between Python and CAD software (Autodesk Inventor). The language used by most CAD software’s is VBA and almost no information for the Python language increasing the number of attempts to create any sort of feature. 6.1. SUMMARY The main goal of the dissertation was achived. It was to model and simulate the highpressure die casting process in aluminum alloys which brought the programming interface with the CAD environment. Progresses were made to make pyhton programming language a useful tool to the gating design in the manufacturing technology high-pressure die casting. The industrialization of this software would reduce the time spent modelling the gating system, however it remains a branch to develop. However, small adjustments had to be made before exporting to CAE environment. Furthermore, H.E.L.P. Die casting was designed in such way the user could to add more data about casting and mold materials which the NADCA manual would not possess. Lately, the second part of the goal was done in the CAE environment QuikCast software to validate possible solutions. Chapter 6 – Conclusion -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- 102 To an early validation of the process before doing a more intensive postprocessing of the results, it was made a cycling simulation to determine if the die temperature would reach a stable thermal gradient at the beginning of each cycle. This stable thermal gradient was found after the 9th cycle. Three gating designs were made to determine the best solution using the H.E.L.P. Die casting. In all of the three gating designs were evaluated variables such as filling pattern, flow colored path and air entrainment in the using QuikCast’s flow module and solidification time and shrinkage porosity using the thermal module. Regarding the casting process, the final solution presented a process yield of 58% and a cycle time of 42 seconds. Analysing the shot’s dynamic, it was possible to conclude that the injection pressure of the second stage was slightly inferior to the calculated. This happened due to the reference value of kinetic energy loss mentioned in the NADCA manual was higher. Furthermore, it was possible to notice the losses in two distinguished periods. The first was in the first stage of the piston where the losses were equal to 16% of the initial kinetic energy. The second was during the injection phase of the piston where the losses reach 35%. In conclusion, the H.E.L.P. Gating design contributed by allowing the user to generate as many gating designs as it wished following the structure of the NADCA manual saving time during the modelling and simulating process. 6.2. FUTURE WORK The primary advantage of the HELP Gating Design software is its ability to significantly reduce the time required for casting project execution. It can generate the dimensions of a gating system in just 15 minutes, with the data stored in Excel along with the corresponding geometric modeling. The software is also versatile, reliable, and adaptable to both the user’s preferences and the geometry of the component. However, there is always space for improvement to make it easier and clearer for users. Some future features to add would be: •Read previous data: This software remains exclusively academic and to a oneuser only (its author). Therefore it was possible to increase its user-friendliness, for example, allowing the user to use previous sessions and project parameters instead of filling all the parameters again;