**1. Introduction**

Glass materials have been widely used in our daily life due to their superior properties, but machining these materials has always been a challenge due to their hard and brittle nature. There are various machining techniques such as turning, milling, drilling, grinding, and laser machining. However, it is very difficult to study the complex process of crack initiation and propagation through experimental observation and theoretical analysis. In order to understand the dynamics of random crack initiation and propagation in the glass cutting process, discrete element method is adopted to model and simulate the cutting process [1].

The discrete element method (DEM) is a numerical technique which models solid structures as bonded particles. These particles can deform and displace from one another and interact through contacts or interfaces between them. Unlike the finite element method (FEM), the discrete element method has the advantage of modeling brittle fracture. It can describe nonlinear behavior of brittle materials and handle the complex particle contact physical process with coupled shear and bulk deformation effects. Since the DEM was first introduced by Cundall [2], it has been widely applied in various areas such as simulating crushable soil [3], granular flow [4], and even behavior of the earthquakes [5]. In recent years, the DEM also has been used in simulating the cutting process of various materials such as rock, ceramics, and carbon fiber reinforced polymer [6–8].

Nowadays, hybrid machining has become more and more popular compared to traditional machining methods. It combines different machining actions on the material that need to be removed and makes use of the combined advantages to avoid or reduce some adverse effects [9]. For example, vibration-assisted machining combines machining with small-amplitude tool vibrations. During this process, the cutting tool loses contact with the chips on a specified amplitude, resulting in decreased machining forces and improved tool life and surface finish [10]. Chemical-assisted micromachining

combines micromachining and chemical reaction of the workpiece. For silicon, the bonding forces between Si particles on the surfaces can be weakened by the hydrofluoric acid. A low concentration of hydrofluoric acid added to the abrasive slurry in ultrasonic machining can increase the material removal rate and surface quality [11]. Laser-induced crack-assisted machining combines laser micromachining and traditional orthogonal cutting. Before orthogonal cutting is carried out, the workpiece is treated by a femtosecond laser to induce seed cracks on the workpiece. This process is able to reduce cutting force, subsurface damage, and tool wear [12]. Due to the hard and brittle nature, it is very difficult to study the complex process of crack initiation and propagation through experimental observations and theoretical analysis. It also brings big challenges to researchers when developing new machining processes. Therefore, the DEM simulation is used in this study to help us understand the dynamics of random crack initiation and propagation during the cutting process.

In this paper, the discrete element method is used to simulate the orthogonal cutting of soda-lime glass with different seed cracks. The purpose is to predict the effects of different seed cracks on the cutting process. The first step is to create a synthetic material that behaves like soda-lime glass. Then, the macro-properties are calibrated by adjusting the micro-parameters of the DEM model to match the mechanical properties of the real soda-lime glass. Orthogonal cutting experiments are conducted in order to validate the cutting forces, subsurface cracks, and chips. Finally, the cutting simulations with four different types of seed cracks are conducted in order to optimize the cutting force, random cracks, and surface roughness. Through this study, the effects of different types of seed cracks during the orthogonal cutting is predicted.
