Model description

4.2.1.1

DEM simulation model at elevated temperature (650

_C)

As mentioned earlier, the MC carbide is the most prevalent carbide type in HSS work roll [19, 20] and it has various complicated shapes such as rod-like, chicken-feet, branch, chrysanthemum or coral-like carbides [29, 132]. However, the rod-like and chicken-

feet-shaped MC carbides are two popular MC shapes as shown in Figure 4.7. Thus due to

the complexity of the 3D structure in creating other shapes and the expensive computational time in the modelling, the current work in this Section focuses only on two shapes of MC carbides. The rod- and chicken feet MC carbides and their orientation will be distributed into the oxide scale with the desired distribution, position and orientation.

113 Moreover, the hard iron oxide debris particles trapped on strip surface have been modelled to investigate their size influence on the HSS wear rate. In addition to that, the penetration depth of these iron oxide debris particles has also been considered to investigate their influence on the roll wear. The current work has taken advantage of DEM that can naturally produce material removal to predict HSS roll wear at 6500_C.

Figure 4.7 (a) “Chicken feet”- shaped and (b) “Rod” MC Carbides by SEM micrographs of MC carbides [29, 132]

Due to their significant hardness compared to that of the oxide scale, the two MC carbide types, rod- and chicken-feet carbides were simulated as rigid bodies (as cluster of spheres) to simplify the model and save the simulation time (there is no bond between these spheres). In the EDEM Solutions commercial software, there is a feature of merging spheres to form a rigid body, each “rod”- shaped carbide is formed by merging 8 different small spheres whereas the “chicken feet” shapes are represented by joining 11 small spheres with the dimension as shown in Figure 4.8.

114 Figure 4.8 Carbides (rigid bodies) formed by merging small spheres (a) chicken-feet

carbides type and (b) rod-like carbides type

4.2.1.2

DEM model of oxide layer on HSS work roll at 650

_C

The oxide scale (Fe2O3) in DEM model consists of 22404 spheres (radius 3𝜇𝑚) were

randomly generated in the pre-defined box. This box was created in EDEM with the size of 480µm x 170µm x 96µm in x, y, z direction, respectively as shown in Figure 4.9. All the oxide particles were dropped and stabilized into this box thanks to the gravity. A total of 21 MC carbides (including 2 different shapes with total 9.65x10-4_{mg in mass) were created}

in three “virtual planes” (each plan contains 7 random MC carbides of two types); their position and orientation can be pre-defined through a feature of EDEM Solutions 2.7.3 [100]. After that, a total of 7 MC carbides in each group were randomly dropped on the top of the oxide material. The whole box was compressed by a rigid plane until reaching the desired packing density of around 64% [121-123] and bonded at the pre-defined bond time. After this time, the contact-bonds among the particles (and carbides) are formed and all bonds are activated as shown in Figure 4.9.

115 Figure 4.9 Distribution of “rod” (green color)- and “chicken feet” (red color)- shaped

carbide in the oxide layer (yellow color)

The MC carbides morphology and distribution formed on HSS have been investigated with scanning electron microscopy and optical microscopy by Luan et al. [29]. They were positioned on the surface of the oxide to ensure the sliding tip will be affected by these carbides in this study. Moreover, experimental results in [132] have shown that the carbides are distributed randomly by group on HSS material. The carbides with “rod”- and “chicken feet” carbide shapes are randomly generated in 3 groups near the surface of HSS roll, each group contains 7 carbides as can be seen in Figure 4.10.

Figure 4.10 Group distribution of carbides (a) in experiment [29] and (b) in simulation (Case 2, 90-degree carbides)

116 Three simulations will be carried out to investigate the effect of MC carbides orientation on wear resistance and a case without carbides was also implemented to compare with three cases with the presence of carbides. The 180-degree oriented carbide (case 1); 90-degree oriented carbide (case 2) and 0-degree oriented carbide (case 3) have been created and positioned in (XYZ) as (-1, 0, 1), (0, 1, 0) and (1, 0, 1) respectively as shown in Figure 4.11. All the MC carbides were generated on the top surface of the oxide sample (at the same Z level) in order to investigate their effect on the HSS wear rate during the tip scratching on the oxide layer. The MC carbides in case 1 (180-degree) and case 3 (0-degree) have been positioned along with the scratching direction while the carbides in case 2 (90-degree) are normal to the scratching direction.

117 The abrasive wear of the HSS roll surface is simulated by a single scratching tip with the rounded tip radius 𝑟𝑡𝑖𝑝 = 15𝜇𝑚, which is modelled as a rigid body. The tip initially

penetrates on the HSS surface with a depth of 30𝜇𝑚 and then slides at this constant depth through 3 groups of carbides as shown in Figure 4.12. The sample was constrained by the rigid wall at the bottom; two sides and the end face of the box; this will keep the sample stable during simulation. Because the sample was constrained by the wall at the end surface, the resistance force on the tip will increase significantly when it approaches the end plate. This will be discussed further in the next Section.

Figure 4.12 Configuration of scratching simulation

The aim of this simulation is to determine how much wear for each case during the scratching simulation and investigate the influence of carbides orientation, different scratching tip size and scratching tip depth on HSS roll wear at 6500_C.