Gregory W. Swift

Thermoacoustics: A Unifying Perspective for Some Engines and Refrigerators

The Atomic Processes in Plasmas Conference is a bi-annual international conference on topics covering high-energy-density plasmas, magnetically confined fusion plasmas, astrophysical plasmas, fundamental atomic data and advanced modeling and plasma diagnostics. The conference lets international researchers share cutting-edge results in plasma creation, plasma experiments and plasma modeling.

A thermoacoustic Stirling heat engine

Electrical and mechanical power, together with other forms of useful work, are generated worldwide at a rate of about 1012 watts, mostly using heat engines. The efficiency of such engines is limited by the laws of thermodynamics and by practical considerations such as the cost of building and operating them. Engines with high efficiency help to conserve fossil fuels and other natural resources, reducing global-warming emissions and pollutants. In practice, the highest efficiencies are obtained only in the most expensive, sophisticated engines, such as the turbines in central utility electrical plants. Here we demonstrate an inexpensive thermoacoustic engine that employs the inherently efficient Stirling cycle. The design is based on a simple acoustic apparatus with no moving parts. Our first small laboratory prototype, constructed using inexpensive hardware (steel pipes), achieves an efficiency of 0.30, which exceeds the values of 0.10–0.25 attained in other heat engines, with no moving parts. Moreover, the efficiency of our prototype is comparable to that of the common internal combustion engine (0.25–0.40) and piston-driven Stirling engines, (0.20–0.38).

A thermoacoustic-Stirling heat engine: Detailed study

A new type of thermoacoustic engine based on traveling waves and ideally reversible heat transfer is described. Measurements and analysis of its performance are presented. This new engine outperforms previous thermoacoustic engines, which are based on standing waves and intrinsically irreversible heat transfer, by more than 50%. At its most efficient operating point, it delivers 710 W of acoustic power to its resonator with a thermal efficiency of 0.30, corresponding to 41% of the Carnot efficiency. At its most powerful operating point, it delivers 890 W to its resonator with a thermal efficiency of 0.22. The efficiency of this engine can be degraded by two types of acoustic streaming. These are suppressed by appropriate tapering of crucial surfaces in the engine and by using additional nonlinearity to induce an opposing time-averaged pressure difference. Data are presented which show the nearly complete elimination of the streaming convective heat loads. Analysis of these and other irreversibilities show which components of the engine require further research to achieve higher efficiency. Additionally, these data show that the dynamics and acoustic power flows are well understood, but the details of the streaming suppression and associated heat convection are only qualitatively understood.

Analysis and performance of a large thermoacoustic engine

Measurements and analysis of a 13‐cm‐diam thermoacoustic engine are presented. At its most powerful operating point, using 13.8‐bar helium, the engine delivered 630 W to an external acoustic load, converting heat to delivered acoustic power with an efficiency of 9%. At low acoustic amplitudes, where (linear) thermoacoustic theory is expected to apply, measurements of temperature difference and frequency agree with the predictions of theory to within 4%, over conditions spanning factors of 4 in mean pressure, 10 in pressure amplitude, 6 in frequency, and 3 in gas sound speeds. But measurements of the square of pressure amplitude versus heater power differ from the predictions of theory by 20%, twice the estimated uncertainty in the results. At higher pressure amplitudes (up to 16% of the mean pressure), even more significant deviation from existing thermoacoustic theory is observed. Several causes of this amplitude‐dependent deviation are identified, including resonance‐enhanced harmonic content in the acoustic wave, and a new, first‐order temperature defect in thermoacoustic heat exchangers. These causes explain some, but not all, of the amplitude‐dependent deviation of the high‐amplitude measurements from existing (linear) theory.

Microchannel crossflow fluid heat exchanger and method for its fabrication

A microchannel crossflow fluid heat exchanger and a method for its fabrication are disclosed. The heat exchanger is formed from a stack of thin metal sheets which are bonded together. The stack consists of alternating slotted and unslotted sheets. Each of the slotted sheets includes multiple parallel slots which form fluid flow channels when sandwiched between the unslotted sheets. Successive slotted sheets in the stack are rotated ninety degrees with respect to one another so as to form two sets of orthogonally extending fluid flow channels which are arranged in a crossflow configuration. The heat exchanger has a high surface to volume ratio, a small dead volume, a high heat transfer coefficient, and is suitable for use with fluids under high pressures. The heat exchanger has particular application in a Stirling engine that utilizes a liquid as the working substance.

Thermoacoustic Engines and Refrigerators

We ordinarily think of a sound wave in a gas as consisting of coupled pressure and displacement oscillations. However, temperature oscillations always accompany the pressure changes. The combination of all these oscillations, and their interaction with solid boundaries, produces a rich variety of “thermoacoustic” effects. Although these effects as they occur in everyday life are too small to be noticed, one can harness extremely loud sound waves in acoustically sealed chambers to produce powerful heat engines, heat pumps and refrigerators. Whereas typical engines and refrigerators have crankshaft‐coupled pistons or rotating turbines, thermoacoustic engines and refrigerators have at most a single flexing moving part (as in a loudspeaker) with no sliding seals. Thermoacoustic devices may be of practical use where simplicity, reliability or low cost is more important than the highest efficiency (although one cannot say much more about their cost‐competitiveness at this early stage).

Two-sensor power measurements in lossy ducts

Theory and measurements of the use of two adjacent pressure sensors to measure acoustic power flow of a simple harmonic sound wave in a duct are presented. This theory differs from the usual intensity-times-area formulation of this problem by including the phase shift between pressure gradient and velocity, which is caused by viscous drag on the gas at the duct wall. For high standing-wave ratios, the power obtained by this method differs significantly from the product of mid-duct intensity and duct area. These measurements confirm the method to an accuracy of 5%, even at high amplitudes where the acoustic flow is turbulent and the theory might not necessarily be valid.

A cascade thermoacoustic engine

A cascade thermoacoustic engine is described, consisting of one standing-wave stage plus two traveling-wave stages in series. Most of the acoustic power is produced in the efficient traveling-wave stages. The straight-line series configuration is easy to build and allows no Gedeon streaming. The engine delivers up to 2 kW of acoustic power, with an efficiency (the ratio of acoustic power to heater power) of up to 20%. An understanding of the pressure and volume-velocity waves is very good. The agreement between measured and calculated powers and temperatures is reasonable. Some of the measured thermal power that cannot be accounted for by calculation can be attributed to Rayleigh streaming in the two thermal buffer tubes with the largest aspect ratios. A straightforward extension of this work should yield cascade thermoacoustic engines with efficiencies of around 35-40% of the Carnot efficiency.

Design environment for low‐amplitude thermoacoustic energy conversion (DeltaEC)

The Los Alamos thermoacoustics code, available at www.lanl.gov/thermoacoustics/, has undergone extensive revision this year. New calculation features have been added to the original Fortran computational core, and a Python‐based graphical user interface wrapped around that core provides improved usability. A plotter routinely displays thermoacoustic wave properties as a function of x or tracks results when a user‐specified input variable, such as frequency or amplitude, is varied. The Windows‐like user interface provides mouse‐based control, scrolling, and simultaneous displays of plots and of several categories of numerical values, in which color indicates important features. Thermoacoustic phenomena can be calculated with superimposed steady flow, and time‐averaged pressure gradients are calculated. In thermoacoustic systems with toroidal topology, this allows modeling of steady flow caused by gas diodes (with or without time‐averaged heat transfer) and Gedeon streaming. Thermoacoustic mixture separatio...

Similitude in thermoacoustics

Similitude is applied to thermoacoustics without using the acoustic approximation. The equations which are important to thermoacoustics (continuity, motion, and heat transfer) are rewritten in dimensionless form, verifying that the list of dimensionless variables obtained from similitude is complete. Similitude is demonstrated in a thermoacoustic engine using helium, neon, and argon as working fluids, even for large‐pressure‐amplitude nonlinear behavior which differs significantly from predictions of linear thermoacoustic theory. Measurements are also presented for nitrogen and a helium–argon mixture, in order to reveal the influence of the specific‐heat ratio and Prandtl number. Implications of similitude for building scale models are discussed; dimensions, temperatures, or pressures can be scaled.