Article Preview
TopIntroduction
Nanofluids are colloidal suspensions of nano-size-powders (1 nm to 100 nm) in a base fluid that are studied for their potential as enhanced alternatives to ordinary cooling fluids. The first reference to nanofluid enhanced heat-transport properties were reported by Masuda et al (1993) and Choi et al (1995). Since then nanofluids have evolved (Singh, 2008) as typically mixtures of solid metal nano-particles (as alumina, silica, titanium dioxide and copper oxide of up to 5% of nanoparticles), of carbides and nitrides, and of carbon nanotubes or nanofibers in cooling fluids (as water, and its solutions with alcohols). Nanofluids are predicted to have higher thermal conductivity and heat transfer coefficient than those of the base fluids, because of the solids much larger thermal conductivity, and of nanoparticles much larger surface-to-volume ratio and higher mobility (Yu, 2008). Therefore, nanofluids are a promising alternative as coolants for critical-cooling systems, as advanced nuclear systems (Buongiorno, 2006), large engine radiators, and microchip cooling (Wong, 2010). For instance, a computational simulation of ideal nanofluid cooling in a Cummins 500hp diesel engine showed that radiator size could be reduced by 5% (Saripella, 2007). Recently M'hamed et al (2016) reported experimental increases of up to 196% for maximum heat transfer coefficient by a 0.5% nanoparticle volume concentration of multi-walled carbon nanotubes in water/ethylene glycol coolant in an actual radiator system.
However, many concerns remain about the practical applications of nanofluids, mainly because of their potential corrosion-erosion effects on cooling-system material surfaces. Addition of solid particles in flowing fluids is known to lead to higher erosion rates on conduit materials, but the effect of adding smaller size ones, as nanoparticles (1nm-100nm) are largely unknown. Use of nanofluids instead of conventional cooling fluids requires a better understanding of their likely wear/erosion and corrosion effects. Solid-contact models (derived from contact-mechanics theory) are not appropriate to assess the erosion rates from multiple particles which occur in actual application purposes. Early work on the study of multiple macroscopic-particle impact and erosion was carried out by Brainard and Salik (1980) for annealed copper and aluminum erosion produced by both single- and multiple-particle impacts of 3.2-millimeter-diameter steel balls in air at up to 140 m/s; their results agreed with findings of Cousens and Hutchings (1983). Rao and Buckley (1983) investigated erosion and morphology of 6061-T6511 aluminum alloy when treated by normal impingement-jets of a mix of spherical glass beads and angular crushed-glass particles. They showed that the time evolution for gas-jet erosion involved an “incubation period”, an acceleration-deceleration period, and a final steady state period, and they correlated the length of incubation and acceleration periods with decreasing particle impact-velocity.
The work of Gee and Hutchings (2002) describes the four common “dry-type” erosion test systems (e.g., gas jet; centrifugal accelerator; wind tunnel; and whirling arm tests) and includes a discussion on the important variables (as particle impact velocity, particle impact angle, particle size, shape and material, and ambient temperature). The ASTM G76 - 07: “Standard Test Method for Conducting Erosion Tests by Solid Particle Impingement Using Gas Jets” (ASTM, 2007) covers testing of material loss by solid particle impingement with gas-carrier jet-type erosion equipment. There is abundant data from these particle-in-gas-jet erosion tests of metals (including of copper and aluminum alloys) and of polymers (Ruff, 1979). The mechanisms of observed gas-jet erosion strongly depend on the size of impacting particles, a fact that is relevant to the research work in this paper, and the logarithm of the volumetric erosion rate relates linearly to the logarithm of the impact velocity. For large particles impacting ductile materials, the erosion rates peak for a 20° angle of the incident jet with respect to surface, an effect that is attributed to particle ploughing and cutting by fragmented particles (for brittle materials, however, erosion would peak for the 90° angle, because of crack formation on the impacted crater). But for small particles, of less than one micrometer, flow patterns produce large disturbances from the nominal attack angle, and in particular for ductile materials the same erosion rates seem to occur for a wide range of attack angles around the 90° one.