Localised laser preheating for residual stress mitigation during laser repair of wear-resistant cast iron components

Casting defects in the form of porosity can render wear resistant cast iron (WRCI) components for use in the mining industry useless. Such components are often scrapped or recycled, adding cost to the manufacturer and ultimately the end user. Repair of such defects is most often not an option as these materials are classified as un-weldable due to their susceptibility to cracking of any repairs by welding process. High hardness, high carbide volume fraction, high thermal contraction and low ductility makes these materials unsuitable for the rapid heating and cooling cycles of welding. The purpose of this research is to solve this industrial manufacturing problem, thus reducing cost and wastage due to casting porosity defects in 27Cr3C WRCI alloy components. The aim of this research is to develop a method of repair for crack-free laser metal deposition (LMD) WRCI deposits by mitigating residual stress but maintaining wear resistance qualities. The methods used were predominantly experimental using LMD with powdered material feedstock that is the same composition as the 27Cr3C WRCI alloy substrate being repaired. A novel localised laser preheating (LLP) process was developed using a de-focused laser to direct energy to a localised area of the material for a specific duration prior to LMD. The preheating process is used to raise the temperature and reduce the thermal gradient across the substrate, allowing for plastic flow to occur at elevated temperatures upon cooling of the LMD deposit, thus reducing final residual stress state. Numerical simulation was used for LLP parameter definition with a three-dimensional thermal model and experimentally validated material properties. Analysis of microstructure of deposits was carried out via optical microscopy, SEM, XRD, EBSD, neutron high intensity powder diffractometry and Vickers hardness testing. The effect of post-LMD heat treatments on microstructure was also investigated. Cast heat treated benchmark samples were found to be consistent with the literature, containing a coarse network of M7C3 carbides and fine, discrete M23C6 carbides in an α/α’ matrix. As-deposited LMD microstructures have a dendritic morphology with a predominantly γ-Fe matrix and discrete, blade-like nano-scale M7C3 carbides. A 960°C hardening heat treatment resulted in almost complete martensite transformation and slight coarsening of M7C3 carbide phases but dendritic structure was maintained. A 1200°C solution treatment homogenised the microstructure, removing the dendrites and caused carbide coarsening through coalescence and morphology change from blade-like to more globular geometry, however, carbide phases were still discrete within martensitic matrix. Despite coarsening after solution treatment, carbide phases are still significantly smaller than those of cast substrates. Thus, heat treatment is effective in producing LMD deposits with identical phase types and distributions to cast materials. Coarsening occurs through heat treatments but morphology of deposits does not match those of cast microstructures. Analysis of stresses was carried out via visual inspection (dye-penetrant), one-dimensional transverse (layer removal method) and 3-dimensional (neutron diffraction strain scanning). While the presence of cracks was easily identified, quantification of stresses proved difficult due to lack of ductility, high carbide volume fractions and phase variations between deposits and substrates. It was found however that without preheating, cracking occurs close to the deposit/substrate interface. LLP is a functional method for preventing cracking occurring in WRCI LMD deposits. LLP reduces residual stresses, however, stresses still occur in the deposit/substrate interface zone as well as at the deposit surface. Wear resistance testing using Coriolis method was carried out on a) full-surface LMD deposits and b) WRCI repair deposits. Analysis of wear magnitude, relative wear resistance (compared to cast control samples) and wear mechanisms was performed. It was found that as-deposited WRCI has inferior wear resistance compared to heat treated LMD and cast WRCI due to predominantly austenitic matrix and extremely fine, discrete carbides which offer little matrix protection. Solution heat treating full-surface WRCI LMD deposits results in improved wear performance due to martensite transformation and carbide coarsening but is not equivalent to that of cast samples. Heat treated repair surfaces, which include LMD deposits and cast sections of the wear path, are close to equivalent wear resistance of cast surfaces for both fine-sand and garnet abrasives tested. A case study involving LLP-LMD on full-scale WRCI throat nozzle component from slurry pump is described. It was found that the LLP-LMD process was successful in producing crack-free repair deposits at the edges of a large WRCI component. However, cracking occurred around repairs closer to the centre of the part due to the surrounding material acting as a heat sink with conduction drawing thermal energy away from the deposit area and causing a higher cooling rate than that seen at the edge deposits. Surrounding material of the centre deposits also acted as a restraint, preventing stress-reducing strain that was able to occur in the 'open' edge deposits. Research outcomes include the fact that LMD technology can be used to produce crack-free WRCI deposits when used in conjunction with preheating to at least 600°C in order to reduce cooling rate and increase hot ductility to allow for strain to occur during processing. LLP is an effective method of preheating WRCI components prior to LMD repairs, however, further work is required for effective commercialisation on full scale components. Microstructures of LMD 27Cr3C WRCI deposits are described in detail and compared to cast alloy of the same composition. Cracking in LMD deposits of WRCI occurs due to excessive residual stresses caused by thermal contraction of the deposit and the lack of ductility of the substrate and deposit. Erosive wear resistance in Coriolis scenario (2-body, low angle, low impact) shows that the carbide mean free path correlates directly to wear performance more than the carbide volume fraction or bulk material hardness. I.e. in an otherwise identical material, larger carbides offered greater resistance to erosive wear. It must be noted however, that wear resistance is a relationship between material and attacking bodies, not an intrinsic material property, meaning that results in different wear environments would differ. While LLP-LMD method is effective in some circumstances at producing crack-free repairs in WRCI components for service in the mining industry, further work is required in order to make it more commercially viable.