**2. Results**

**Figure 1.** Schematic of erosion testing device: 1-rotating shaft; 2-disk for controlling erosion angle; 3-furance; 4-sample

of materials in seawater and aqueous environments with or without solid particles to reveal the synergistic effects [15, 16, 30–32], the relationships between microstructural parameters (e.g., Fe2B lamellar spacing and orientation in flowing zinc) and flow hydrodynamics under erosion-corrosion condition of materials in flowing liquid metal are poorly understood. Dybkov et al. studied the dissolution of a solid in a liquid metal and pointed out that the flowing liquid metal had a strong effect on the thickness of boundary layer and dissolution constant [33–35]. Meanwhile, corrosion interface morphology can directly affect the corrosion process, which may be governed by inhibitors or corrosion barrier [36, 37]. Therefore, erosion corrosion caused by flowing corrosive mediums and material microstructures are of significant importance to probably generate an interface/fluid interaction (e.g., multi-scale material microstructure/environment interactions) on surface films, thus affecting the diffusion, mass transfer, and penetration as well as flow-induced vibration and film rupture during the

In present work, the interface film morphology and erosion-corrosion behavior of directionally solidified (DS) Fe-B alloy in flowing zinc have been investigated to reveal the effect of Fe2B lamellar spacing and hydraulic flow on erosion corrosion and interface structures. Accordingly, a flow-induced localized corrosion and cracking is also discussed in flowing liquid zinc based on the Fe2B lamellar spacing regulation, which may understand the combined

dipped into zinc to a depth of 40 mm. Erosion surface of the alloy with Fe2B [001] vertical to the interface was chosen for test owing to the oriented doweling effect. Vertical to interface

A rotating-disk technique (erosion set-up in **Figure 1a**) was employed to conduct flow zinc erosion test. Erosion thickness loss was obtained by single-side measurement. Then the ero-

was prepared from oriented alloy,

effects of hydrodynamics and interface film morphology in flowing liquid metal.

holder; 5-crucible; 6-test sample; and 7-liquid zinc.

erosion-corrosion process [27–34, 38, 39].

124 Cavitation - Selected Issues

**1.3. Method and characterization**

sion rate was evaluated using Eq. (1):

Erosion sample with dimensions of 140 × 15 × 5 mm3

was chosen for test owing to the oriented doweling effect.

### **2.1. As-cast microstructure of DS Fe-B alloy**

**Figure 2** shows the morphologies and XRD pattern of as-cast DS Fe-B alloy. It can be seen that the longitudinal morphology displays rod-like quadrangular prism along Fe2B [002] orientation (**Figure 2a**). It is a typical dual-phase microstructure with α-Fe and Fe2B. From the transverse section, most of the eutectic Fe2B shows an irregular shape except for some rectangular borides with hollow structures extracted from the DS alloy (**Figure 2b**). **Figure 2c** shows the XRD patterns of the DS Fe-B alloy. Clearly, a strong (002) peak of the Fe2B phase appears in the transverse section, and only a small (004) peak of Fe2B is detected except the (002) plane. However, many crystal planes are detected in the prism. It indicates that [002] orientation of Fe2B is the sole preferred growth direction. In addition, it can be seen that the strongest peak of Fe is the (110) plane, which indicates that α-Fe may grow along [110] orientation under DS condition.

#### **2.2. Effects of erosion time and Fe2B lamellar spacing on erosion-corrosion rate**

**Figure 3** shows the erosion-corrosion rate of DS Fe-B alloy in flowing zinc as a function of erosion time and Fe2B lamellar spacing (e.g., the spacing between two columnar Fe2B edges showing in **Figure 3a**). It is clear that the erosion-corrosion rate of DS Fe-B alloy with Fe2B [002]

**Figure 2.** Microstructure and XRD of DS Fe-B alloy: (a) longitudinal; (b) transverse; and (c) XRD.

**Figure 3.** Erosion rates of DS Fe-B alloy with different Fe2B lamellar spacing in flowing zinc: (a) erosion-corrosion rate vs. erosion time; (b) erosion-corrosion rate vs. Fe2B spacing.

of Fe2B existing at the erosion-corrosion interface can be seriously destroyed and swept by flowing zinc (**Figure 4b**), while no adhesive film with an interfacial pinning effect develops except for some granular and incompact ζ products when larger size of Fe2B appears (e.g.,

**Figure 4.** Erosion-corrosion interface morphologies of DS Fe-B alloy with different Fe2B lamellar spacings: (a) vertical

Accelerated Cavitation Damage of Steels in Liquid Metal Environments

http://dx.doi.org/10.5772/intechopen.80769

127

**Figure 4c** displays the erosion-corrosion interface morphologies of parallel sample erosion for 30 h. Obviously, a layer-by-layer spallation of Fe2B occurs at the erosion-corrosion layer, and there are a plenty of erosion products with different sizes randomly distributed in the layers. The difference of erosion-corrosion interface structure in various Fe2B lamellar spacings is the spalling amounts of Fe2B and products caused by flow, and unsuited Fe2B lamellar spacing can hardly generate adhesive interface and strong synergistic effect of interfacial oriented Fe2B and epitaxial grown ζ-FeZn13 (i.e., multiphase protective film). Therefore, it uncovers that the erosion corrosion strongly refers to not only the interface structure but also the fluid

**Figure 5** shows the EBSD analysis of DS Fe-B alloy before and after erosion in flowing zinc. Clearly, the DS Fe-B alloy is mainly composed of α-Fe and Fe2B to form dual-phase textured microstructure (**Figure 5a**). The {002} poles of oriented Fe2B grains are located at Y-axis in forms of lightest spot area, while {100}, {110}, and {111} poles of Fe2B grains distribute within quadrants, which indicates that Fe2B [002] orientation dominates its preferred growth direction, as recorded as the strong peak (002) plane in XRD of the transverse section (**Figure 2c**). Furthermore, the {110} poles of α-Fe grains display concentrated spot area in the vertical axis, which may reveal that α-Fe grains generate an orientation growth in the [110] direction.

**Figure 5c** and **d** show the interfacial orientation map and grain boundary distribution collected by EBSD in a well-distributed directional area of Fe-B alloy in flowing zinc. The

λFe2B = 5.94 μm) due to their distinctive interface structures and morphologies.

sample for erosion 5 h; (b) vertical sample for erosion 30 h; and (c) parallel sample for erosion 30 h.

**3.1. EBSD analysis on microstructure and erosion-corrosion interface**

hydrodynamic effects in flowing zinc [14–18, 42].

**3. Discussion**

orientation perpendicular to the erosion interface decreases sharply at first and then gradually declines to a stable level. However, the erosion-corrosion rate of the parallel sample almost linearly decreases, and it maintains the higher erosion rate after the erosion steps into the steady stage (**Figure 3a**). Compared to two erosion patterns (e.g., vertical or parallel sample), it is revealed that there exists an erosion initiation effect or incubation period of the interface ζ formation for the DS Fe-B alloy in flowing zinc (**Figure 3b**). Obviously, the adhesive film with interfacial pinning effect in the vertical sample at initial erosion-corrosion stage does not form. However, the erosion-corrosion rates decrease with the increase of erosion time, for example, erosion from 5 to10 h (**Figure 3b**).

It indicates that the interface structure may undergo continuous changes owing to the combined effect of the oriented Fe2B and epitaxial grown ζ-FeZn13 products under flow erosion condition. Clearly, epitaxial grown ζ-FeZn13 products at the interface demonstrate an accumulation/pile-up effect and synergistically generate a buffer layer with oriented Fe2B to resist the flowing zinc erosion with the prolonged erosion time. It therefore means that the erosion-corrosion interface structure is dominant by flow-accelerated diffusion of liquid zinc and ζ product accumulation at the interface during the prolonged erosion-corrosion process.

#### **2.3. Interface morphological evolution during flowing zinc erosion corrosion**

**Figure 4** shows the erosion-corrosion interface of DS Fe-B alloy in flowing zinc with different erosion times as a function of oriented Fe2B lamellar spacing. It is clear that small and large Fe2B spacing in vertical sample in flowing zinc can be damaged in the form of both Fe2B dissolution and numerous spallation at the front of the erosion-corrosion interface for 5 h (**Figure 4a**). Evidently, only suitable lamellar size of oriented Fe2B can resist the erosion corrosion (e.g., λFe2B = 3.67 μm) at the initial erosion stage. However, after erosion for 30 h at a steady erosion stage (**Figure 4b**), there exists an adhesive film with the interfacial pinning effect in DS Fe-B alloy with Fe2B lamellar spacing λFe2B = 3.67 μm. An adhesive product film comprising the oriented Fe2B and epitaxial grown columnar ζ-FeZn13 builds up at the erosion-corrosion interface as a buffer layer to resist flowing liquid zinc damage. Smaller size

**Figure 4.** Erosion-corrosion interface morphologies of DS Fe-B alloy with different Fe2B lamellar spacings: (a) vertical sample for erosion 5 h; (b) vertical sample for erosion 30 h; and (c) parallel sample for erosion 30 h.

of Fe2B existing at the erosion-corrosion interface can be seriously destroyed and swept by flowing zinc (**Figure 4b**), while no adhesive film with an interfacial pinning effect develops except for some granular and incompact ζ products when larger size of Fe2B appears (e.g., λFe2B = 5.94 μm) due to their distinctive interface structures and morphologies.

**Figure 4c** displays the erosion-corrosion interface morphologies of parallel sample erosion for 30 h. Obviously, a layer-by-layer spallation of Fe2B occurs at the erosion-corrosion layer, and there are a plenty of erosion products with different sizes randomly distributed in the layers. The difference of erosion-corrosion interface structure in various Fe2B lamellar spacings is the spalling amounts of Fe2B and products caused by flow, and unsuited Fe2B lamellar spacing can hardly generate adhesive interface and strong synergistic effect of interfacial oriented Fe2B and epitaxial grown ζ-FeZn13 (i.e., multiphase protective film). Therefore, it uncovers that the erosion corrosion strongly refers to not only the interface structure but also the fluid hydrodynamic effects in flowing zinc [14–18, 42].
