A prototype Stirling-engine regenerator containing a matrix made of carbon-fiber-based composite materials has been developed. The concept underlying this development effort is one of exploiting the properties of composite materials (e.g., the anisotropy of thermal conductivity of carbon fibers and the tailorability of composite materials and structures) to reduce thermal and flow losses below those of previously developed regenerators containing metal matrices.

The regenerator in a Stirling engine is an internal heat exchanger for transferring heat between a working fluid and a flow-channel wall (which is also part of the regenerator). The fluid can be helium or another gas that has suitable thermodynamic properties and that does not react chemically with engine components. A typical regenerator is cylindrical in overall shape and includes one or more axial passage(s) containing a matrix -- an open, thermally conductive structure with many flow paths and large surface area for transfer of heat to and from the working fluid. ("Matrix" as used here is meant to be distinguished from "matrix" as used elsewhere to designate the nonfibrous or nonparticulate component of a composite material. Hereafter in this article, the terms "regenerator matrix" and "matrix material" will be used to avoid ambiguity.) Stated somewhat differently, the matrix provides a thermal connection between the gas and the heat capacity of the wall.

Figure 1. A Composite-Matrix Regenerator containing a matrix of radial carbon fibers offers advantages over current regenerators containing matrices of screens, balls, or foils.

Problems associated with making effective regenerators stem from limitations of materials of which they are made. Regenerator matrices are subjected to hot, oscillating gas flows and high temperature gradients. For high performance, a regenerator should be thermally insulating in the axial direction (along which a substantial thermal gradient can exist) and should exchange heat rapidly with the working fluid. The regenerator should contain minimum dead volume because dead volume reduces the engine compression ratio. The flow of gas through the regenerator matrix introduces drag and viscous losses, which should be minimized in order to maximize performance. Efforts at reducing some loss mechanisms tend to aggravate others.

Matrices in current regenerators have been made of various components, including steel wool, steel felt, stacked screens, packed balls, metal foils, and parallel plates. The matrix in a composite-matrix regenerator (CMR) could be made in any of a variety of composite materials and configurations; for example, regenerator matrices could comprise radially or circumferentially oriented thick fibers (see Figure 1). The fibers in the channels could be composites built up on thinner carbon-based fibers. The ends of the fibers would be embedded in cylindrical or otherwise-shaped axial-flow-channel walls.

Figure 2. This Prototype Composite-Matrix Regenerator is made of radially oriented carbon fibers bonded to an anisotropic carbon/carbon composite tube. Fabrication involved electrostatic flocking of the carbon fibers followed by rigidization (including thickening of the fibers) by chemical-vapor infiltration.

Relative to current regenerators containing packed balls, felts, or stacked screens, CMRs containing fibers across axial-flow channels offer the potential advantage of greater heat-transfer effectiveness for a given flow friction. The high surface area of the fibers enhances gas/solid heat transfer, making it possible to obtain adequate performance from fewer, wider flow channels than one might otherwise need; this creates an opportunity to reduce costs because fewer, larger channels can be fabricated more easily. Other advantages of CMRs containing fibers crossing axial-flow channels, relative to current regenerators, include smaller pressure drops and smaller dead volumes.

The lengthwise thermal conductivities of graphitized carbon fibers range from about 300 to about 1,000 Wm-1K-1 -- about 20 to 67 times the thermal conductivity of stainless steel. The transverse thermal conductivities of these fibers are only about 1/100 of their lengthwise conductivities. If these fibers were combined with low-thermal-conductivity matrix materials, the resulting composite materials would exhibit high anisotropy of thermal conduction. In principle, such fibers and matrix materials could be used to advantage in a regenerator in the following ways:

  • The fibers could be used to conduct heat between the flowing gas and the flow-channel walls.
  • The walls could be made of a composite material containing radially oriented fibers to utilize the thermal-conduction anisotropy to maximize radial conduction between the wall and the fibers in the channel while minimizing undesired axial conduction. Preferably, the matrix material in the wall would be one of high specific heat as well as low thermal conductivity.

Figure 2 shows a prototype CMR. A CMR of this type has survived limited endurance testing in a small Stirling engine, where it exhibited thermal performance among the best of a number of different regenerators that were tested.

This work was done by Timothy R. Knowles of Energy Science Laboratories, Inc., for Lewis Research Center. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com under the category, or circle no. 70 on the TSP Order Card in this issue to receive a copy by mail ($5 charge).

Inquiries concerning rights for the commercial use of this invention should be addressed to

NASA Lewis Research Center,
Commercial Technology Office,
Attn: Tech Brief Patent Status,
Mail Stop 7-3,
21000 Brookpark Road,
Cleveland, Ohio 44135.

Refer to LEW-16581.