Finite element simulation of machining Inconel 718 alloy including microstructure changes

Highlights

• The most suitable Johnson–Cook material model was identified for machining simulation of Inconel 718.

• FE numerical model is customized to predict grain refinement and affected layer during dry and cryogenic machining of Inconel 718 alloy.

• Fe model can be used to predict parameters related to the surface integrity in turning processes of Inconel 718 alloy.

Abstract

Inducing thermo-mechanical loads during the machining of hard materials lead to the severe grain refinement and hardness variation into the machined surface. This variation significantly affects the performance and the service quality of the products. Inconel 718 superalloy is one of the difficult-to-machine materials employed widely in aerospace industries and its surface characteristics after final machining process is really important. The main objective of this study is to implement a reliable finite element (FE) model for orthogonal machining of Inconel 718 alloy and prediction of the microstructure changes during the process. At first, experimental results of cutting forces, chip geometry and maximum temperature were taken into account to identify the most suitable material model out of the seven models found in the literature. Then, the FE numerical model was properly calibrated using an iterative procedure based on the comparison between simulated and experimental results. Moreover, a user subroutine was implemented in FE code to simulate the dynamic recrystallization and, consequently, to predict grain refinement and hardness variation during the orthogonal cutting of Inconel 718 alloy. Zener–Hollomon and Hall–Petch equations were employed to respectively predict the grain size and microhardness. In addition, the depth of the affected layer was controlled using the critical strain equation. As overall, a very good agreement has been found between the experimental and simulated results in term of grain size, microhardness and depth of the affected layer.

Introduction

Nickel-based superalloys were created in the 1940s to be used in gas turbine applications because of their excellent performance and service quality. Inconel 718 superalloy is one of the hard materials among other nickel based alloys which are also used extensively in many industries. Inconel 718 has superior properties such as wear resistance, high melting temperature, high corrosion and creep resistance and maintaining strength and hardness at high temperatures. Consequently, this alloy finds a wide range of applications in the aerospace industry and in the rotary parts of gas turbine engines such as blades, shafts and disks. The mentioned properties are responsible for poor machinability of the Inconel 718 from various points of view such as surface integrity [1], [2], [3].

Surface integrity of a machined workpiece is one of the most specified customer׳s requirements. It becomes more essential in machining of difficult-to-cut materials (such as Inconel 718) because of the severe thermo-mechanical loads induced during the process. Surface and subsurface alterations including microstructural and microhardness changes are significant factors of the surface integrity [4], [5]. Machining is known to be a fast chip formation operation where a workpiece material is exposed to high temperature, strain and strain rate during the process. When the amount of these dislocations exceeds a critical value, dynamic recrystallization (DRX) occurs and leads to the grain refinement in the machined surface. Microstructural changes and consequently microhardness variation during the finish turning process are much more important than most of the manufacturing processes, since these phenomena occur near the machined surface and directly influence the performance and service quality of the final products [5], [6]. However, experimental investigation of the surface integrity in machining processes is very expensive and time-consuming.

Recently, the finite element method (FEM) has been developed as a beneficial and efficient tool in order to simulate the machining processes [6], [7] and to investigate the mechanical and thermal variables as well as the some parameters related to the surface integrity avoiding, at the same time, unnecessary experiments [6]. Unfortunately, it should be highlighted that various indicators of the surface integrity, such as residual stresses and microstructural changes, are difficult to model and required accurate inputs [6], [8]. Therefore, reliability of the machining simulation is a very important task, especially when valuable outputs such as the surface integrity parameters are modeled.

In this regard, the material constitutive model is the most important prerequisite for simulation of the cutting process and plays a significant role on the FE numerical results. In other words, the identification of the most reliable material model is the first step for an accurate simulation of the machining process. In general, the Johnson–Cook׳s (J–C) constitutive equation is used to model the behavior of the workpiece material during the cutting simulation [6], [9]. Furthermore, the reliability of the numerical models drastically depends on the fracture criterion, frictional and thermal parameters between the chip, tool and workpiece interfaces [6]. Therefore, a suitable selection of these parameters is also an essential step to obtain a reasonable accuracy for the prediction of the cutting forces, temperature fields, and chip geometry [6], [9], [10]. As overall, it is evident that, the correct selection of the mentioned parameters is truly beneficial for predicting the surface integrity on the machined workpiece by FE analysis.

Currently, significant improvements have been reported for machining simulation of difficult-to-cut materials [11], [12], especially for the machining of nickel and titanium based alloys [3], [13], [14], [15]. Recently, Ahmed et al. [16] modeled the distribution of thermal strain in the machined surface during conventional and ultrasonically assisted turning of Inconel 718. Courbon et al. [17] investigated the effect of high-pressure coolant on the thermal and mechanical loads induced during the machining of Inconel 718. Ozel et al. [18] conducted an experimental and numerical investigation in order to evaluate the effect of the tool nose radius on the cutting forces and residual stresses in the machining of Inconel 100 alloy.

So far, several J–C material models have been introduced for aged and annealed Inconel 718 superalloy, although the degree of success of these material constitutive models has not been compared with various machining experiment tests to identify the most reliable material model for the simulation of the cutting process (e.g., a sort of benchmark). In addition, only few works are found in literature to simulate the microstructural changes during the machining of other materials [3], [19], but also it has been never conducted for machining of Inconel 718 alloy.

In this study, orthogonal cutting process of Inconel 718 was modeled using the FE method in order to improve the knowledge on the J–C material constitutive model for Inconel 718, to highlight advantages and drawbacks of different set of material constants and to enhance the FE analysis when surface integrity parameters need to be predicted. To do that, the process was firstly simulated using different J–C material models and the results were compared with experimental data to identify the most reliable material model. The effective parameters of simulation including thermal and frictional conditions and fracture criterion were calibrated and then validated with several experimental data. Once the most suitable J–C material constitutive equation was found, a user subroutine was implemented in the FE software to simulate microstuctural and microhardness changes. Zener–Hollomon parameter and Hall–Petch equation were taken into account to simulate the grain refinement and hardness variation, respectively. In addition, the depth of the affected layer was controlled using the critical strain considered as a function of the temperature and strain rate. At the end, a very good agreement was reported between predicted results and corresponding experimental data.