Numerical simulation of metal flow and solidification in the multi-cavity casting moulds of automotive components

https://doi.org/10.1016/j.jmatprotec.2007.12.004Get rights and content

Abstract

The metal flow and solidification behaviours in a multi-cavity casting mould of two automotive cast parts were simulated in three dimensions. The commercial code, FLOW-3D® was used because it can track the front of the molten metal by a volume-of-fluid (VOF) method and allows complicated parts to be modelled by the fractional area/volume obstacle representation (FAVOR) method. The grey iron automotive components including a brake disc and a flywheel were cast using an automatic sand casting production line. Solid models of the casting, the gating system and the ceramic filter were spatially discretised in a multi-block pattern. The surface roughness and the contact angle of the mould were taken into account in the model, based on the properties of the sand mould used. The turbulent flow was simulated using the two-equation kɛ turbulence model. The D’Arcy model was used to analyse the fluid flow throughout the ceramic filter designed in the gating system. The simulation model was validated against the experimental observations. The model was used to investigate the appropriateness of the multi-cavity mould design and its running system for each automotive component.

Introduction

Numerical simulation provides a powerful means of analysing various physical phenomena occurring during casting processes. It gives an insight into the details of fluid flow, heat transfer and solidification (Flemings, 1974, Campbell, 1991). Numerical solutions allow researchers to observe and quantify what is not usually visible or measurable during real casting processes. The goal of such simulations is to help shorten the design process and optimize casting parameters to reduce scrap, use less energy and, of course, make better castings. Simulation produces a tremendous amount of data that characterize the transient flow behaviour (e.g., velocity, temperature), as well as the final quality of the casting (e.g., porosity, grain structure). It takes good understanding of the actual casting process, and experience in numerical simulation, for a designer to be able to relate one to the other and derive useful conclusions from the results.

Most of the casting modelling codes can be divided into two categories: those using the finite difference (FD) approach for solving fluid flow equations, and those that employ the finite element (FE) method (Barkhudarov, 1998). The FE method uses body-fitted computational grids leading to more accurate representation of metal/mould interfaces than generally achievable by FD methods. However, generating good quality FE grids is still a challenging task and often takes significantly more time than the simulation itself. Solution accuracy degenerates in highly distorted grids and changes in geometry, even small ones, often require a completely new grid. The FD method offers ease of mesh generation due to the structured nature of the mesh, uses less storage to describe geometry and simplifies the implementation of the numerical algorithms. However, the conventional FD methods often require fine grids to describe complicated geometry to reduce errors associated with the ‘stair-step’ representation of curved boundaries. The latter introduces inaccuracies when computing liquid metal flow along the walls and heat fluxes normal to the walls.

In this work, the commercial, general purpose, computational fluid dynamics (CFD) code FLOW-3D®, was used to simulate the filling and solidification sequences of two automotive components, cast into the multi-cavity sand moulds (FLOW-3D, 2005). The process model developed was used to investigate the appropriateness of the running and feeding systems.

Section snippets

Model theory

The CFD code FLOW-3D® is based on a finite volume/finite difference approach. Two methodologies, fractional area/volume obstacle representation (FAVOR) and volume-of-fluid (VOF), constitute the core of the software. These methods differ from methods in most other codes but offer many advantages, and are summarised below (Barkhudarov and Hirt, 1993).

Numerical simulation

Two automotive components including a brake disc and a flywheel were simulated in this work. The complete solid models of the parts were created in steriolithography (STL) format and imported to the software (Fig. 1). Due to the symmetry plane of the system, only half of each model was modelled. The multi-block meshes of the models are shown in Fig. 2. Thermo-physical properties of the cast iron parts, silica mould and ceramic filter, were derived from both literature and manufacturer's

Experimental

In order to validate the simulation model, the filling time of each component was measured carefully by a precise stopwatch. The solidification time of the castings were determined by knocking out the moulds in different times after the pouring. Table 2 shows the filling and solidification times measured for the two castings. A Minolta/Land Cyclops 152 infrared pyrometer was used to measure the melt temperature just before pouring. All castings were cut transversely after cooling down to

Brake disc part

The filling pattern of the three-cavity brake disc mould is shown in Fig. 3. The cast iron melt stream with a cross-sectional area less than that of the sprue is entered into the mould and fills up the primary runner followed by the secondary runner after about 1.0 s (Fig. 3a). The melt is then entered to the mould cavity through the second gate of the side-castings followed by the gates of the middle-casting, when the inclusion trap in the primary runner is completely filled. During the filling

Discussions

The verified model interestingly represented the correct location of the hot spots in the castings. Fig. 7 compares the simulated final location of the hot spots for the brake disc part with the micro-shrinkage that is experimentally observed, showing a reasonable agreement. It should be noted that due to the automatic moulding system being used, it was not possible to propose a suitable chilling system to avoid such micro-shrinkage. However, the simulation results showed that decreasing the

Conclusions

A 3D simulation model was developed to simulate the filling and the solidification behaviours of the automotive components, cast in an automatic sand casting production line. The verified model based on the experimental observations, showed that the four-cavity mould is more suitable than the three-cavity one, in getting a more uniform casting quality for all cast parts. The model also represented a different performance between the gates for each cast part, suggesting a smaller cross-sectional

Acknowledgments

The authors appreciate the collaboration of their colleagues at Isfahan University of Technology and Azarin Casting Industries of Isfahan, especially Mr. H. Morady.

References (10)

  • Barkhudarov, M.R., 1998. Advanced simulation of the flow and heat transfer in simultaneous engineering, Technical...
  • Barkhudarov, M.R., Hirt, C.W. Casting simulation: mold filling and solidification—benchmark calculations using...
  • J. Campbell

    Castings

    (1991)
  • M.C. Flemings

    Solidification Processing

    (1974)
  • Flow Science, Inc., 2005. FLOW-3D® User's Manual, Version...
There are more references available in the full text version of this article.

Cited by (74)

  • Optimization of multi gating system for the experimental investigation inflow parameters with effects of design in small scale industries

    2022, Materials Today: Proceedings
    Citation Excerpt :

    Reynolds‘ number was the same as in the models and actual system [14,15]. Metal flow and solidification activity in a multi cavity of two automobile cast parts were mathematically tested against experimental observations. [16]. It's related to mass manufacturing.

  • Experimental and simulation analysis on multi-gate variants in sand casting process

    2021, Journal of Manufacturing Processes
    Citation Excerpt :

    The shrinkage porosity was measured in both experimental and numerical methods for different plate thickness and observed a minimum variation at maximum thickness. Additionally, several recent studies have suggested improvement of quality products to different casting industries through validation of experimental/numerical models, using various software with defined constraints based on the type of processes [21–23]. From the above literature reviews, it was evident that the demand for the experimentation and numerical analysis is still required towards transformation of automation in sand casting industries.

  • Influence of parameters on the smart productivity of modern metal casting process: An overview

    2021, Materials Today: Proceedings
    Citation Excerpt :

    The shrinkage porosity was calculated in experimental and computational methods for the different thickness of the plate and the minimal difference was found at the highest thickness. In addition, several recent studies have suggested improving the quality of products in different casting industries by validating experimental/numeric models using various software with specified constraints based on the form of process [22–24]. This casting process is a cost effective process extensively used for mass production applications [25].

View all citing articles on Scopus
View full text