The following paper is condensed from chapter 2 of "Solid Particle Erosion and Erosion-Corrosion of materials" A. V. Levy, published by ASM International, 1995

Introduction

This chapter describes the mechanism of erosion of metals based on physical observation of erosion surfaces at high magnifications with the great depth of field that is possible in scanning electron microscopy (SEM) correlated with measured metal losses. This mechanism of erosion can account for those aspects of erosion behavior that the microcutting mechanism could not, as well as those that it did explain. The mechanism consists of sequential plastic deformation processes that account for each of the separate occurrences that result in the overall surface degradation.

The Platelet Mechanism of Erosion

In work conducted by the author to determine how specific steel microstructures affected erosion, an erosion weight loss curve depiction was used that resulted in basic doubt concerning the validity of the microcutting mechanism. In order to learn more about the initiation of the erosion that occurred, an incremental weight loss rate curve was used rather than the cumulative curve generally reported in the literature. The erosion was conducted on steel, incrementally, 60 g of particles at a time, and the weight loss caused by each 60 g increment was measured.

The initial erosion rate caused by the first 60 g of silicon carbide particles was much lower than that of subsequent 60 g batches of erodent. Also, extrapolating the curves down to 0 erosion shows that a number of grams of particles had impacted the surface before erosion losses began. With subsequent increments of impacting particles, the metal loss rate increased until it reached steady-state conditions where each increment of particles caused the same metal-loss rate to occur. Steady-state conditions occurred after a relatively short exposure period.

If microcutting was the primary mechanism of erosion, the erosion rate of the initial, uneroded surface should be higher than subsequent incremental rates because work hardening of the surface due to the machining action would have reduced the machinability of the surface and the resulting amount of material loss. Weight loss by erosion also should have started with the first impinging particles that would have machined off metal. This relationship also raised doubts concerning the effect of the hardness and strength of the metal on erosion. The lowest hardness, lowest strength, spheroidized condition of the 1075 eutectoid steel tested had the lowest erosion rate.

The effect of work hardening of the spheroidized steel was also investigated. It was expected that as the material was work hardened its erosion resistance would increase with the resulting hardness increase. The erosion rate for the initial 60 g of erodent was monitored for 1075 steel specimens that were cold rolled to various percentage reductions. The hardness doubled between the annealed steel and the 80% cold-reduced steel, but the initial erosion rate, rather than decreasing with increasing hardness, increased significantly. It did not achieve the steady-state incremental erosion rate, but approached it.

The data from these experiments, coupled with micrographs showing extensive piling up of severely plastically deformed material that was extruded up and out of the craters produced by single particle impacts, established doubts regarding the microcutting mechanism of erosion.

Microscopic Sequence of Erosion

The development of an eroded surface on 1100-0 aluminum was metallographically observed a few impacts at a time. This entailed developing a technique that would locate the same microscopic area using SEM after sequential, very short erosion exposures. To do this, microhardness indentation markers and much patience was required.

The appearance of the surface at relatively low magnifications at the beginning of the erosion exposure was examined.. The changes in the pyramid-shaped area as it was struck by individual particles shows the development of flattened platelets as each subsequent 0.1 g of erodent struck the total eroding surface. It can be seen that after 0.4 g of erodent had been used, essentially no material loss had yet occurred in the observed area. Just the tip of the pyramid was knocked off. This sequence accounts for the fact that there is a threshold period in erosion when severe plastic deformation is occurring with no material loss. It is followed by an increasing material loss as more particles strike the surface, until steady-state conditions occur. Carrying this sequential erosion experiment to steady state erosion produced the highly textured surface. Hundreds of platelets in various states of deformation were seen. There were marked changes that occurred after the second gram of particles struck the whole surface. The platelets were knocked off and new platelets have been extruded out from the main area of platelets.

In summary, the loss of metal from an eroding surface was observed to occur primarily by a combined extrusion-forging mechanism at all particle-impact angles. The evidence indicates that the platelets are initially extruded from shallow craters made by particle impacts. Once formed, they are forged into the distressed condition in which condition they or parts of them are vulnerable to being knocked off the surface by subsequent particle impacts.

Surface Observations

Evidence of extrusion being the initiating mechanism of platelet erosion was obtained in an experiment where a thin layer of copper was plated on a 1020 steel substrate that was subsequently eroded with a few silicon carbide particles. On a cross section of the eroded surface area copper was seen beneath the surface, indicating that the steel had been extruded over it from the nearby shallow craters. The sequence of extrusion followed by forging of the extruded material can readily account for the surface and subsurface locations of the copper plating. The presence of the thin layer of copper over the entire surface of the craters indicates that it is extrusion that forms the lips or platelets rather than microcutting, as machining would have removed the thin copper layer.

Surface Heating

While there is no direct evidence of heating of the eroding surface or any direct temperature measurements, considerable indirect evidence was gathered during the course of the investigation that temperatures near the recrystallization temperature of the target metal probably occurred at the immediate eroding surface. Evidence has been observed of aluminum that has been melted and resolidified on the surface of a silicon carbide erodent particle that was used to erode 1100-0 aluminum. Almost all of the eroding particles that were captured and observed after the erosion process had some once-melted aluminum on them. Considerable recent work and some older work in the literature supports the fact that adiabatic shear heating and, possibly, some frictional heating occurs on the surface during the erosion process.

Work Hardening

The extensive plastic deformation that occurs on the surface of an eroding metal should cause some amount of work hardening to occur somewhere in the eroding surface region. Because the immediate surface was determined to be at an elevated temperature, it was thought that the work hardening might occur just beneath the heated surface. To determine where work hardening occurred, microhardness measurements were made on cross sections of eroded aluminum and steel specimens. The first measurements were made 5 m beneath the actual surface. The lower hardness, immediate surface region can be seen, particularly for the 30º impact angle, where erosion rates are highest. The hardness increased to a subsurface work-hardened zone followed by a hardness decrease to that of the base metal. The evenness of the geometry of the microhardness indentations indicates that valid near-surface readings were obtained.

Platelet Mechanism Description

The erosion-heated surface, 5 to 15 m thick, consists of platelets at various stages of generation as the result of large plastic strain deformation. Beneath the platelet zone is a work-hardened zone that developed during the early stages of the erosion exposure. This zone lies beneath the heated surface region and strain hardens as a function of the strain-hardening coefficient of the target metal. Beneath the cold-worked zone is base metal at its initial condition. It is proposed that the following sequence occurs in the erosion process. Initially, platelets are formed without loss of material. Adiabatic shear heating of the immediate surface region begins to occur. Beneath the immediate surface region, the mass of target material forms a work-hardened zone because the kinetic energy of the impacting particles is enough to result in considerably greater force being imparted to the metal than that required to generate platelets at the surface.

When the surface has been completely converted to platelets and craters and the work-hardened zone has reached its stable hardness and thickness, steady state erosion begins. The higher steady-state erosion rate compared to the initial rates is due to the subsurface cold-worked zone acting as an anvil to increase the efficiency of the hammerlike impacting particles, to extrude/forge platelets in the now fully heated and most deformable surface region. When the anvil is fully in place and the platelets are fully formed and heated, maximum, steady-state, material removal rates will occur.

Erosion Mechanism of Brittle Scales on Metals

In most elevated-temperature erosion environments, the eroding surface is undergoing corrosion as well as erosion. The eroding surface region, therefore, is some combination of deposited erodent particles, surface scale, and base metal. To better understand the behavior of in situ formed, brittle surfaces on structural metal substrates undergoing combined erosion-corrosion, the erosion of scales formed under oxidation conditions on metal substrates has been studied.

From the investigation of the erosion behavior of relatively thick, duplex nickel oxide scale, it was determined that scales formed in situ on metals erode sequentially down through their thicknesses by a cracking and chipping mechanism rather than being knocked off the metal in pieces at the scale/metal interface. The presence of the ductile metal substrate does not appear to have a major effect on the way that the brittle scale erodes. However, the ductile substrate does have a small effect. The thinner scales can transfer more of the kinetic energy of the impacting particles to the ductile nickel and, hence, they crack and chip and then erode at lower rates than the thicker scales. As the thicker scales are removed, their rates of erosion decrease to a rate that becomes nearly the same for all scale thicknesses. In the case of the duplex nickel oxide scale, the harder, more dense, columnar grain outer scale protected the inner, softer, more porous, equiaxed scale from erosion as long as it was present.

Return to Technical Session Home Page

Return to InterCorr® Home Page