UCLA

A unique multiferroic type of thermomagnetic generator is being investigated in order to establish its thermodynamic efficiency at different size scales. This device generates electricity when a magnetic material interacts with a thermal gradient by means of a spring-magnet mechanism. This unique technology is compared to other thermal-electric energy harvesting technologies to show that these devices have a similar goal of achieving a maximum theoretical efficiency of around 50% relative to Carnot. The first approach towards achieving improved performance relies on the analytical modeling, and experimental verification, of several subsystems stemming from the original design, which include the optimization of the magnetic force component, the optimization of the heat transfer process and the efficiency of the energy conversion process. The method to improve the magnetic force component is not recommended and neither is the method to improve the heat transfer process. Nevertheless, the energy conversion subsystem is successfully modeled and verified; thereby suggesting that an electromagnetic induction coil may be better suited for the energy conversion process over a ferroelectric transduction mechanism at bulk scale. A cascade design is also investigated as a method to improve device efficiency; though analysis reveals a design flaw, which leads to other methods for improving efficiency.

Two models of thermomagnetic generator thermodynamic efficiency are developed, which are based on distinct approaches taken by Solomon and Brillouin in order to compare this unique system to a Carnot engine. The model based on a modified form of Solomon's approach results in a relative efficiency of 0.5%, which compares well with an estimate of efficiency based on provided data from the original design. This representative model of efficiency was then applied to a survey of pure elements for comparison, which confirms gadolinium as the best material for use as a working body with a relative efficiency of around 20%. The model based on Brillouin's approach represents a more rigorous thermodynamic analysis, which qualitatively agrees with the results based on Solomon's approach, though it predicts larger values of efficiency for most of the materials in the survey. The conservative model based on Solomon's approach is then applied to a hypothetical system that uses a single-domain magnetic material as a working body. This effort is pursued since single-domain nanostructures exhibit a remanent magnetization, which is shown to increase magnetic energy density. The resulting analysis predicts efficiencies on the order of 30% relative to Carnot for this nanoscale system, though the effects of size are not considered in this model.

A model is developed to correlate the effects of size on thermodynamic efficiency for this device. Considering a nanoscaled nickel structure as a working body, this model combines three existing models to predict relative efficiency values that are comparable to the bulk scale, although this system may operate closer to room-temperature. This result is unexpected since the absolute efficiency is shown to increase as a function of decreasing size, though this discrepancy is explained as a consequence of Curie point suppression. The combined model is also applied to a hypothetical composite made of separated layers of nickel with distinct thicknesses. This composite material is predicted to spread the ferromagnetic transition across a much larger temperature range as compared to bulk nickel, such that this material may be better suited for different applications; for example, as a sensor or as a thermal switch. Moreover, this combined model is also shown to be a lower-bound estimate of thermodynamic efficiency, since the actual performance depends on material characterizations that have yet to be determined.

The magnetization of ferromagnetic nanostructures as a function of both applied field, and temperature represents a current engineering challenge. This may be due to current manufacturing techniques that produce defects, and may also be due to resolution limitations of commercially available magnetometers. These challenges also affect a proposed hypothesis regarding the existence of a distribution of transition temperatures within a single nanostructure. This hypothesis aims to resolve whether the diffuse transition behavior of ferromagnetic nanostructures is due to volume-averaged magnetometer measurement techniques, and/or due to surface effects from defects resulting from imperfect manufacturing techniques. These issues would need to be resolved before considering a nanoscale design, which may potentially be useful as a secondary energy-recycling device in thermally assisted magnetic recording applications. Nevertheless, a proof-of-concept experimental setup is offered that may be useful for future designs at the smallest relevant scales.