Go to the Acoustics' website Go to the Penn State Thermoacoustics website Go to Penn State's website Home

 

The FRANKENFRIDGE

"I believe we can make a refrigerator out of parts of others!"

This project to examine thermoacoustic refrigeration at high amplitudes has married a high power acoustic driver (up to 100 W of acoustic power) to the resonator of the STAR refrigerator that flew on the Space Shuttle Discovery (STS-42) in 1992. The effort that started in the winter of 1997 has produced a Master's Degree Thesis in 1998 by Matt Poese called "Performance Measurements on a High Amplitude Thermoacoustic Refrigerator." A list of publications and presentations that relate to this project, including the thesis, are listed below. This work was sponsored by the Office of Naval Research, PSU Applied Research Lab's Enrichment and Foundation Program and the Pennsylvania Space Grant Consortium.

A few other articles that relate directly to this project are also listed below.

Overview of the Frankenfridge Device

The Space Thermoacoustic Refrigerator (STAR) was designed and built by a team at the Naval Postgraduate School led by Steve Garrett. It has the ability to move about 5 Watts of heat and exhibited a peak coefficient of performance relative to Carnot (COPR) across the stack of about 20%. A good paper about the STAR project is "Thermoacoustic Refrigerator for Space Applications," referenced above.

The Shipboard Electronics Thermoacoustic Cooler (SETAC) sailed on the USS Deyo in 1994. It was designed and built at NPS by a team led by Steve Garrett. SETAC can move about 400 Watts of heat, which is about enough to cool the space of a standard dormitory sized refrigerator. The drivers for this refrigerator are a custom design that are 55% electric-to-acoustic efficient. Driving the STAR resonator with the SETAC driver allows us to study the effect of nonlinear effects (such as turbulent oscillating flow) on the performance of a thermoacoustic cycle. The study of performance at high acoustic amplitude is important because the power density of a thermoacoustic device is roughly proportional to the square of the acoustic pressure amplitude. This means if you can double the acoustic pressure amplitude in the thermoacoustic resonator, you get four times the power density! Large power density is important for refrigeration applications that have severe volume or weight restrictions, like automotive air conditioners, satellite and other space applications.

 
This picture shows the fully constructed apparatus. The approximate height of the entire apparatus is about 56 cm and the diameter of the driver (the widest dimension) is 23 cm.   The photo below shows the disassembled components of Frankenfridge. Starting from the left side, the exhaust flange insulator ring is on the left of the exhaust heat sink flange. These two components are leaning against the SETAC driver. The STAR resonator is to the right side of the photo. Inside of the yellowish fiberglass section of the resonator is the stack (7.8 cm long and a diameter of 3.8 cm) which is sandwiched between the hot heat exchanger (on the bottom side of the stack in this photo) and the cold heat exchanger (on the top of the stack in this photo.)

It is hoped that the Frankenfridge project will live on to support several thesis projects or other investigations. Frankenfridge is a well calibrated thermoacoustic device and its performance in its current configuration is thoroughly recorded. For this reason, the device provides a very good test bed for new thermoacoustic stack or heat exchanger designs. In fact, one such project to investigate a new combined stack/heat-exchanger package has taken place in the Physics Department at Penn State.

Details of Frankenfridge

Measurement of Exhaust Heat Flux

A water circulation loop coupled to the hot heat exchanger has been added to the resonator as part of the Frankenfridge project This addition increases the flexibility of Frankenfridge as a test bed for thermoacoustic components. This loop (soldered around a copper flange, shown in the previous picture, leaning against the driver housing) allowed us to make direct measurement of the exhaust heat flux.

Frankenfridge has a small electric resistance heater taped to the outside of the copper cold side duct just below the cold heat exchanger. In the STAR project, the entire resonator was insulated from the atmosphere by a vacuum insulation system which guaranteed that the amount of heat dissipated by the resistance heater was the amount of heat pumped by the refrigerator up the stack to the exhaust heat exchanger (which was thermally connected to the large aluminum driver housing of the STAR driver). Because of this near perfect insulation scheme, the First Law of Thermodynamics guarantees that the amount of energy introduced to the resonator by the driver plus the amount of heat pumped up the stack (the energy introduced by the resistance heater) is the exhaust heat flux. However, the vacuum insulation system is very cumbersome and a direct measurement of exhaust heat flux would eliminate the need for it. By measuring the mass flow rate through the copper tubing soldered to the outside of the flange and the temperature difference of the water on the inlet and the discharge of the tubing, the amount of heat absorbed by the water is known. This measurement of exhaust heat flux has proven to have a absolute resolution of 65 milliwatts.. The resonator section of Frankenfridge is insulated with only construction grade Corning Pink Fiberglass insulation, which works well for a small temperature difference (3-7 degrees Celsius) between Frank's cold side and the temperature of the room. In this regime, the exhaust heat flux measurement confirms that there is very little heat leak from the room to Frank's cold side. For lower cold side temperatures, the exhaust heat flux measurement allows us to know how much extra heat the fridge pumps from the room.

Heat Exchanger Model

In order for the STAR resonator to be space qualified, there could be no penetrations to the resonator. Consequently, the temperature sensors for the hot and cold side are mounted externally on the body of the resonator, near the hot and cold heat exchangers. This is problem because calculation of the stack Carnot coefficient of performance requires that the stack hot and cold side temperatures be known, which are different than the temperatures outside of the resonator body. In Frankenfridge, the stack temperatures are inferred based on the temperatures measured on the resonator body with a simple calculation of the thermal conduction resistance of each parallel heat exchanger fin.

DeltaE Model

The measured performance of Frankenfridge, that is expected to exhibit nonlinear performance degradation, is compared to a linear computer model that predicts Frankenfridge performance. This computer program is called DeltaE and is available through Greg Swift or Bill Ward at Los Alamos National Laboratory. If you would like to use the DeltaE model of Frankenfridge, you can have the model file. This file is a valid model file for DeltaE and is simply a text file.