Modern electronic and communication technologies are witnessing explosive growth in miniaturization and system speeds. As a result, the need for efficient cooling systems is critical.
Most conventional coolant fluids have limited heat conductivity, except for metals which are generally unusable at typical processing temperatures. Thus, despite the use of turbulence and increased heat exchange area, the intrinsic weakness of traditional coolants remains their low conductivity.
Cooling fins and microchannels have already been studied intensively, but these are inadequate for the needs of newer devices in electronic and optical devices. Moreover, microelectromechanical systems (MEMS) and nanotechnology are novel fields which demand enhanced cooling power, not to mention larger devices that call for compact but high-powered cooling.
Researchers are therefore working frantically to study heat transfer at nanoscale, in order to find better ways of cooling.
The idea of a solid suspended in a liquid to form a more effective coolant was pioneered many decades ago, initially using particles at microscale or larger. The utility of these fluids was restricted by the following factors:
The particles tended to rapidly form a surface layer which inhibited further heat transfer
More vigorous fluid circulation tends to erode the system tubes and channels, pushing up the costs
The particles tend to block microchannels in the cooling system
The presence of particles causes a higher pressure drop
A higher concentration of particles is required to produce an increase in conductivity; but this promotes the above issues as well
Thus this field was not very promising until nanofluids came into the arena.
What are the characteristics of heat transfer in nanofluids?
Nanofluids are fluids containing suspended nanoparticles. The concept of nanofluids has helped to revitalize the use of suspended solids in liquids to form effective coolants. This is because the properties exhibited by nanoparticles are quite different from those shown by the same material in bulk form.
Some of the properties useful in the enhancement of heat transfer include:
High aspect ratio
Nanoparticles offer a very large surface area for heat transfer, since one in five of their atoms are found on the surface. This makes plenty of electrons available for heat transfer. Different types of nanoparticles offer different levels of thermal transfer enhancement. Carbon nanotubes (CNTs) have a thermal conductivity of about 3000 W/mK, and a high aspect ratio of about 2000.
Nanoparticles have low particle momentum and very high mobility.
The small size of the molecules allows for free movement and hence microconvection, which promotes heat transfer.
The fluid may show much faster heat dispersion owing to these two factors.
Particle size
Small particles weigh less, and are therefore less likely to undergo sedimentation as a result. If stabilizing agents are used, the fluid may remain stable for months at a time.
Nanoparticles are ideal for use in microchannels which handle large heat inputs, because of their high conductivity and aspect ratio. Their use avoids the danger of clogging that is associated with larger particles. In addition, the small particle size reduces wear and tear, extending system life.
Cost-effectiveness
Typically, in order to double the heat transfer of a coolant fluid, the pumping power must be increased ten times, or the conductivity must be increased three-fold. This does not hold good if the viscosity of the fluid increases.
Nanofluids display disproportionate increases in conductivity with a very slight or insignificant increase in particle volume fraction. This helps to bring down the required pumping power, and thus reduces costs significantly. For instance, one study at ANL showed a 150% increase in heat conductivity when the volume fraction of nanoparticles was increased by just 1%. In this case, multi-walled carbon nanotubes (MWCNTs) suspended in engine oil were used. MWCNTs offer an increase in conductivity of 20 000 times relative to engine oil.
Relation to temperature and concentration
The nanofluid maintains perfectly Newtonian behavior because of the small particle concentration and insignificant rise in viscosity, with hardly any change in the pressure.
Nanofluids show a three-fold increase in the rate of rise of conductivity with increasing temperature, which indicates the strong possibility that the movement of particles within the nanofluid undergoes drastic alterations as the temperature rises.
The increase in conductivity was found to be not only proportional to the concentration of the nanoparticles but also inversely proportional to particle size.
How do nanofluids achieve heat transfer?
Classical theories about heat transfer occurring in nanofluids have not succeeded in explaining or predicting the behavior of nanofluids. Attention has therefore shifted to the motion, surface action and electrokinetics of nanoparticles. The microconvective movement of these particles also may give rise to hydrodynamic force.
The layering of fluid around the particles may help with rapid heat transfer by providing a pathway. Ballistic heat transport was a possibility because of the nanoparticle dimension of the phonon free path. However, with nonlocal nonequilibrium conduction it has been demonstrated that effective conductivity is actually lowered rather than increased, which debunks this explanation.
Fractal geometrical models also fail to explain nanofluid behavior in the absence of adsorption.
Another new approach is using the field factor in combination with depolarization and a dielectric constant. This uses liquid layering with a couple of adjustable parameters, relating to the thickness of the liquid layer around the nanoparticle and its conductivity, and yields accurate values, matching measured results.
The disadvantages of this model and its modifications are the assumptions that the liquid layer size is very large, and that it shares the same thermal conductivity as the solid layer, both of which are unproven experimentally. In fact, the only experiment that measured the liquid layer showed it to be only three atomic diameters thick, and this is confirmed by theoretical calculations.
Another way to explain nanofluid conductivity enhancement uses a drift velocity model for particle motion, assuming the presence of nanoconvection in the inter-particle spaces. This obviates the need for adjustable parameters.
Brownian motion has been used to produce a model for nanofluid behavior based on four methods of energy transfer, namely: the thermal conductance of the fluid based on the collisions between the liquid molecules; heat diffusion in nanoparticles; Brownian-motion-induced collisions between nanoparticles; and interactions between the base liquid and the moving nanoparticles. With very small particles, the order of magnitude of random motion becomes exaggerated, with increasing importance being given to convection and related effects. This model is capable of accurately predicting the conductivity in terms of particle size and temperature.
A similar model deals with both stationary and moving particles. Stationary particles show geometrically increasing surface areas for unit volume as the size of the particles decreases. Heat flows both through the liquid and the solid particles, in parallel paths. This explains why thermal conductivity goes up with volume fraction, and in inverse proportion to the particle diameter. With moving particles, the kinetic theory of gases is invoked to explain how the thermal conductivity goes up with temperature. The values and the order of magnitudes agree with those predicted by the kinetic theory, unlike some other models.
How does the type of nanoparticle affect heat transfer?
With oxide nanoparticles, heat transfer by convection actually increases moderately, but viscosity also increases. With the use of metallic nanoparticles, a very small increase in particle concentration leads to almost unchanged viscosity but a high enhancement in heat transfer. With CNT nanofluid use, effective thermal conductivity goes up with temperature and with particle concentration, but more with temperature.
With increasing concentrations, however, heat transfer by convection goes up significantly. Thus the increase of convective heat transfer is much more pronounced that can be explained by the enhancement of effective heat conductivity. This may be explained by the rearrangement of particles, increase in heat conduction as a result of shear, lowering of the thickness of the thermal boundary layer because of the presence of nanoparticles, and high aspect ratio of CNTs.
With a suspension of oxide nanoparticles, boiling effects deteriorate, probably because of the plugging of microcavities on the fluid surface by the nanoparticles. This causes nucleation site density to fall, hindering boiling. This may be used to produce fluids which can prevent boiling or cause boiling to occur only when a preset surface temperature is reached, as may be necessary in some heat treatments or material processing.