Internal void closure during the forging of large cast ingots using a simulation approach

Abstract

Large cast ingots often contain defects or undesirable microstructural features, such as voids and zones related to casting. Some of these features can remain after hot open die forging, which is an important process for converting large cast ingots into wrought components. During the initial cogging and deformation steps prior to the detailed open-die-forging operations, any internal voids should be eliminated. The present work focuses on the closure of internal voids during open die forging so as to produce a sound component. Hot compression tests were conducted to obtain the flow strength of the cast microstructure at different temperatures and strain rates. The measured flow strength data together with other appropriate material properties were used to simulate the forging steps for a large cast ingot. The numerical simulations for the forging deformation and for the internal void behavior were performed using DEFORM-3D™. Actual defects were measured in commercial ingots with an X-ray scanner. The simulation results for the void deformation behavior are compared with voids measured before and after forging. Through the comparison of experimental results and numerical simulation, a criterion for void closure is proposed. The criterion is that a local effective strain value of 0.6 or greater must be achieved for void closure during forging. Such a criterion can be used in conjunction with simulations to insure that a sound component is produced during the hot open die forging of large cast ingots.

Introduction

Recently, there is an increasing need to produce large forged components for aerospace, naval, energy, and other applications. Open die forging of large cast ingots is the primary process used to produce high quality large wrought components. Cogging or side-pressing processes are used in the primary stages during most open-die forgings. In this study we will use the industrial term “upsetting” for what is sometimes called side pressing in the literature. Upsetting in the present study is not the compression of a cylinder along its axis, but rather the compression of a cylinder perpendicular to the axis of symmetry. During the preliminary deformation processes, any internal voids from the initial cast ingot need to be eliminated. If these voids are not closed during the initial deformation stages, they may nucleate a crack or be a source for a defect during the subsequent open die hot forging steps. Fig. 1 shows an example of an internal void and the surface crack that can result if the void was not closed during the initial cogging or upsetting operations.

A number of studies have been conducted on using finite element analysis to help with the operation and control of open die forging processes. Kiefer and Shah (1990) used three-dimensional FEM to examine the effects of die width and reduction on the internal stresses and strains in an open die forging of rectangular shaped workpieces. Lee et al., 2008a, Lee et al., 2008b examined the effects of temperature and strain rate on high temperature deformation behaviors to draw deformation processing map. Dudra and Im (1990a) used two-dimensional FEM to study the axial compression of a cylinder and the plane-strain side pressing (i.e. upsetting) of a circular cross-sectioned workpiece by open die forging. The side-pressing study used dies of different configurations to investigate press loads and internal strains. Cho et al. (1998) also used three-dimensional FEM to study the effect of die configuration, die width and reduction on the open die forging of rectangular billets. They evaluated their results by comparing the calculations to small sized laboratory experiments using plasticine model materials at a scale of 1–40. Tamura and Tajima (2001) used three-dimensional FEM to study the formation of a surface deformation pattern that they called “concave defect”. They verified their results by laboratory experiments using lead as a model material. From their analysis they also provided a recommendation for a small die modification to avoid the surface defect. Tamura and Tajima (2003) and Tamura and Tajima (2004) extended their work to develop a pass schedule for open die forgings, which provides a more homogeneous strain distribution within the workpiece. Tamura et al. (2005) further extended their studies of open die forging to determine a die shape that would minimize overlap defects during multiple pass side pressing of an octagonal workpiece into a circular cross-sectioned shape. Dyja et al. (2004) used three-dimensional FEM to investigate the effect of some complex die shapes on the deformation patterns in open die forging. They validated their calculations with laboratory scale experiments using steel as the workpiece. Choi et al. (2006) used three-dimensional FEM to optimize the open die forging process for conversion of a rectangular/square cross-sectioned workpiece into one with a circular cross section. They focused on feed rate and rotation angle in the study. It is evident from these studies that FEM simulations can be very helpful in understanding and controlling open die forging processes. Furthermore, the deformation behavior has also been analyzed by FEM in conjunction with constitutive model. Yoon et al. (2010) predicted plastic flow behavior by considering an operating deformation mechanism and strain hardening model.

One of the main benefits of open die forging is the ability to close internal voids that come for cast ingots. There have been a number of previous studies examining how to operate the open die forging process to best close these internal voids. Dudra and Im (1990b) extended their two-dimensional FEM analysis to study the closure of centerline pores during side pressing of circular shaped billets using dies with various shapes. From their analysis they indicated that effective strain appears to be a better indicated of void closure than hydrostatic stress. Park and Yang (1996) proposed a bonding mechanism based on hot pressing that could help in studying the closure of voids during open die forging. Park and Yang (1997a) used three-dimensional FEM together with a Taguchi set of experiments to study void closure during open die forging of large ingots. It appears that they assumed centerline voids in the analysis. They found that the die width ratio and proper die shape are beneficial in closing voids whereas cooling of the ingot surface does not appear to be helpful for void closure. They also provide some insight on the bonding efficiency as the void surfaces came together. Park and Yang (1997b) extended their bonding efficiency study with FEM modeling and experiments. They also investigated the bonding as a function of void position in the ingot. They found that die shape in conjunction with reduction in height controls the bonding process.

Kim et al. (2002) developed a neural network algorithm for the closure of voids in open die forging of rectangular cross-sectioned workpieces. They studied a large number of process parameters and used the algorithm to recommend a forging pass schedule that would close voids effectively and efficiently. Overstam and Jarl (2004) used three-dimensional FEM to examine the closure of centerline voids in rectangular cross-sectioned workpieces. They found that the bite ratio was a critical factor in getting the pores to close. Banaszek et al. (2005) and Banaszek and Stefanik (2006) used two-dimensional FEM to study the effect of die shape and other process parameters on void closure in open die forgings. They examined voids at various locations in the ingot and provided verification with experimental tests using a highly alloyed steel. They found that die shape had a major effect on void closure and advocated the use of shaped dies during the initial stages and flat dies during the final forging stages to achieve best results. Chun et al. (2006) examined the closure of centerline voids in rectangular cross-sectioned workpieces using three-dimensional FEM. They studied the effect of die width ratio, die feed rate, die shape, and number of passes on the closure of voids. Skubisz et al. (2008) studied the closure of centerline voids during open die forging both numerically and experimentally. They found that a critical amount of effective strain was needed to close these voids. Lee et al., 2008a, Lee et al., 2008b used three-dimensional FEM to examine centerline void closure during the axial compression of cylindrical workpieces. They found that a critical amount of effective strain was needed to close the voids. Zhang and Cui (2009a) developed an analytical model for the closing of a spherical void during axial compression of a cylinder. They verified their model using two-dimensional FEM. Zhang et al. (2009b) indicated that void closure in forgings is a multi-scaled problem and they developed a mesomechanics approach to the issue. They indicated that during the early stages void closure increases with the level of stress triaxiality increases. Their criterion for void closure depends on the hydrostatic stress, the effective stress, the effective strain, the Norton exponent (essentially the inverse of the strain rate sensitivity exponent) and four numerical values that were determined as a function of the Norton exponent via FEM analysis and regression. Kakimotoa et al. (2010) used both two-dimensional and three-dimensional FEM to study the closure of centerline voids during axial compression and side pressing of circular and rectangular crossed-sectioned forgings. They calculated a void closing index, Q-value, which must achieve a value of 0.21 to insure void closure. They indicate that the Q-value uses Oyane's equation and was initially proposed by Ono et al. (1993).

Void closure has also been studied in other large shapes produced by deformation. Hamzah and Ståhlberg (1998) studied pore closure in manufacturing of heavy rings using two-dimensional FEM. They found that piercing of the initial hole in a closed container produced a pore free ring whereas open die piercing of the hole did could produce a final ring with pores. They attributed the successful manufacturing route to the large strains created during the piercing in a closed die. Hamzah and Ståhlberg (2001) extended their ideas in proposing a new route for producing pore free rings, which primarily depended on high strains during piercing and subsequent forming. They also found that the hydrostatic pressure was a less critical parameter for pore closure.

It is evident form these previous investigations that effective strain is an important parameter in determining whether or not a void will close. Although stating that effective strain is important, many studies do not give the specific amount needed for effective void closure. It is also interesting to note that several of the studies have found that hydrostatic stress is a less important parameter in void closure as compared to effective strain. Although Kakimotoa et al. (2010) have proposed a very specific criterion for void closure during open die forging, their criterion is fairly complex and requires knowledge not only of the effective strain, but also the effective stress, the hydrostatic stress as well as some flow properties of the material. Although a good theoretical study, it would be difficult to implement their criterion in a production environment.

The present study is concerned with the elimination of the internal voids in large ingots so as to obtain a sound final product. Most of the previous studies have only investigated voids in the center of the workpiece rather than throughout. In the present study voids at various locations in the workpiece are investigated and a specific criterion for the amount of effective strain needed for void closure is proposed base on the analysis of the results generated. The criterion for eliminating internal voids provides important information for the design and operation of an open die forging process that results in a good quality, sound product.

In this study, hot compression tests were conducted to obtain the flow strength of cast ingot material at different temperatures and strain rates. FEM simulations were performed to investigate the deformation behavior of cast ingots during the various forging stages. The measured flow strength data as well as appropriate thermal property data were used to simulate the upsetting process of cast ingot. DEFORM-3D™ was used to perform the numerical analysis of void closure. The calculated results for void deformation behavior are compared to experimentally measured results before and after upsetting of actual ingots/forgings, which were determined by X-ray scans. From the comparison of the numerical simulations and experimental results, the criterion for the amount of deformation needed for void closure was developed. Fig. 2 shows the flow chart for this investigation into void closure of cast ingot during the initial upsetting stages of open die forging.