COPPER-NIOBIUM, COPPER-VANADIUM, OR COPPER-CHROMIUM NANOCOMPOSITES, AND THE USE THEREOF IN HEAT EXCHANGERS
We propose here a class of new materials for high heat-flux applications including high flux heat exchangers, rocket engines, jet engines, gas turbines, space-plane wings, and fusion reactors. The materials are nano-composites formed from copper and a refractory metal, especially niobium, vanadium, or chromium, but also potentially silver, iron, tantalum, tungsten, or molybdenum. The copper plus refractory mix is fast-melted, e.g. by arc melting, and then fast-cooled and worked. When cast the component metals separate into a fractile metal-metal composite that should have excellent heat-transfer qualities. Working the material makes it a lot stronger by extending the fractile structures into micron, and submicron (nano-scale) filaments and sheets of metal-metal composite. The resulting strong, high thermal-conductivity material should be excellent for demanding heat exchange applications, especially those where the heat flux is so high that ordinary materials of construction would suffer from thermal creep: that is from large forces generated internally by the differential expansion caused by the heat flux. Typical heat exchanger surfaces that might use this material might be tubes or indented flat plates.
Of non-composite, ordinary materials, the ones that are most resistant to high thermal stress tend to be alloys of molybdenum and alloys of copper, see Table 1. These alloys excel in their thermal stress parameter, kσT/αE, combining high thermal conductivity, reasonably high strength, low young's modulus, and low thermal expansion coefficients. Applications that need these properties include any highly compact, high value heat exchangers, as in rocket and jet engines, fusion reactor first walls, hypersonic plane wings, or any other high thermal stress and shock applications.
Some other high temperature alloys appear attractive, and can be expected to exceed these two for very special applications. Based on their thermal stress parameters, other attractive materials should include Be, Mo—Re, Ta—W, and tungsten (W), but these materials suffer from being more expensive than Mo or Cu alloys, and/or suffer from being hard to fabricate and join.
SUMMARY OF THE INVENTIONA heat exchanger is provided that is made of a copper-refractory metal composite that defines a structure of the heat exchanger. A process of forming a refractory nanofilament filled copper matrix is provided that includes fast casting a molten mixture of a refractory copper mixture of more than 10% refractory to form a casting. The casting is then mechanical worked to form dendrites with mean diameters of less than 10μ (micron).
DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTSThis invention involves the use of nano-composites of refractory metals (particularly Nb, V, and Cr) dispersed within a copper matrix. In particular, nanocomposites containing 10-60% refractory component dispersed in the copper, and highly worked to form fractile nanofilaments and nano-thickness sheets dispersed in it. Wire of this material was developed as an intermediate step on the process to make niobium-tin superconducting wire, but a comparison of its predicted thermal shock parameter to those of Mo and copper alloys, Table 1, suggests that sheets and tubes of this nano-composite should be very useful for high thermal flux heat exchangers.
As Table 1 shows, the thermal stress parameter of Cu-40% Nb is predicted to be more than double that of molybdenum or copper alloy due to a higher critical strength, a lower Young's elastic modulus, and a higher thermal conductivity, predicted here based on a high electrical conductivity. Copper beryllium is a used in rocket-engine heat exchangers, a very high thermal stress application [2].
Although it is generally difficult to form nanometer-thickness wires and sheets in a metal matrix, the proposed nanocomposite essentially self-form when copper-niobium, copper-vanadium, or copper chrome is fast cast, e.g. following arc melting, and is then worked (deformed) to thin the filaments and sheets. It appears that the best nano-composites form with very pure, very ductile metals, e.g. pure niobium and copper at high copper contents (over 82% Copper), but ductile nano-composites also form with copper and chromium, though chromium is not ordinarily ductile. We have also been able to make nano-composites with 40-50 volume percent niobium or vanadium in copper, and that appears to be a technical first.
One key requirement to make these nano-composites is that the two components must be immiscible in each other in the solid phase; another is that the metals be very pure, and have a very low oxygen content of less than 2 atomic % and preferably less than 0.5 atomic % and most preferably between 0.001 and 0.3 atomic %. Upon fast casting from the melt, these immiscible mixtures are observed to form a fractal interconnecting networks of dendrites, 1μ to 100 nm in diameter. In low-oxide copper, these dendrites are very ductile. When the two-phase materials is now drawn or rolled the copper and refractory dendrites thin out, reaching about 10-50 nm without needing an anneal step. The resulting metal-metal composites can be quite strong [3,4] and should be quite thermally conductive. Because these composites self-form on processing, they are sometimes called “Deformation Processed Metal-Metal Composites” or DPMMCs as a shorthand [4].
The high thermal conductivity values in Table 1 are predicted by a simple rule of mixtures based on an assumption of interconnecting matrixes, and the conductivity of the pure materials. Measurements of the electrical conductivity seem to confirm this simple rule—at least for lightly annealed composite along the direction of draw [4]. Based on this, we can predict that the nano-composites will show the high thermal conductivity values in Table 1 when measured along the direction of rolling, and not-much lower conductivity when measured across the direction of rolling.
Table 1 also shows exceptional break-strengths for the Cu—Nb composite, and modest values of E, the Young's elastic modulus. These values go along with the composite's very high strain to break, and very high strain to initiate creep, about 3% for modestly worked material [3]. These high strengths, and higher, up to 2200 MPa have been measured for highly worked Cu-20% Nb, and for several other similar metal pairs including Cu-15% Nb, Cu-15% Ta, and Cu-15% Fe [4]. The values for Cu-40% Nb shown in Table 1 are projections based on an extrapolation between Cu-20Nb, Cu-10% Nb, and Cu-15% Nb.
A possible explanation for the very high strength of the highly worked composite is that the niobium ends up as filaments that are only 10-25 nm thick. This is the equivalent of 40-100 refractory atoms width. Similar strength has been seen in other metals when drawn down to filament sizes this narrow. The copper also appears to be hardened in the composite as the refractory prevents intercrystalline creep.
Cu-40% Nb, Cu-50% Nb, and Cu-45% V have been made for us by the Materials Preparation Center of Ames Lab. All these compositions appear to be reasonably ductile; and the early experiments confirm that the higher refractory compositions are a lot stronger than the low-refractory ones. We expect this pattern to show particularly at high temperatures, where there is the greatest need for new heat-exchanger materials. Cu-60% Nb has been made as part of the invention, but it has not been tested yet, and may not be as useful for heat exchanger applications as the thermal conductivity may suffer. It should be noticed that the measured Young's modulus for the modestly worked, Cu-20% Nb composite 128 MPa matches, reasonably closely to the value one would expect from the values for copper and niobium, following a simple rule composition-weighted average.
Theory of Thermal Stress:
A material's resistance to mechanical force is shown below in terms of the force per unit area, σ, the material's Young's modulus, E, and the strain ΔL/Lo:
F/A=σ=EΔL/Lo 1)
In SI units, s is measured in Pa, force in Newtons and area in meters-square. In SI, E is likewise measured in Pa, or N/m2. For an unconstrained metal, the linear thermal expansion is:
ΔL=LoαΔT or ΔL/Lo=αΔT 2)
where α is the coefficient of thermal expansion, and ΔT is the change in temperature, ° C. When the surface of a material is heated across its thickness, the hot surface is constrained to not expand more than the centerline does. Canceling the required elastic strain, ΔL/Lo in Equation 1 against the thermal induced strain, ΔL/Lo from 2, we find:
σ=αEΔT 3)
the maximal amount of heat transfer, Q*, through a material is the maximal ΔT* to the centerline times k, the thermal conductivity, and divided by the material thickness to the centerline, ∂.
Q*=ΔT*k/∂ 4)
Combining Equations 3 and 4 we see that the maximal heat transfer of the material times this material thickness, ∂, is
∂Q*=σTk/αE 5)
where σT is the maximal strength of the material at the use temperature.
The key materials parameter in Equation 5 is kσT/αE. This is seen to determine the robustness of a given thickness of material to thermal stress. For thin-walled tubes constrained at the ends, the thermal stress parameter can be shown to include the Poisson ratio:
M=2(1−ν)σTk/αE 6)
Where M is the thermal stress parameter and ν is the Poisson ratio.
Table 1 compares the parameter in Equation 5, kσT/αE, for a variety of possible first wall materials. It is seen that copper alloy, molybdenum, beryllium, and tungsten are the best of the simple materials in resistance to thermal stress. According to the predictions in Table 1, these composites should resist thermal far better than these, or any simple alloy at a given wall thickness based on kσT/αE. These advantages should extend to fairly high temperatures, important e.g. for use in gas turbines and fusion reactor first walls.
The high strength of the these materials may allow for the use of thinner walled heat-exchange surfaces than possible with lower-strength materials. Thus, for example, allowing for the use of exceptionally small wall thickness tubes in heat exchangers. According to equation 6, this should reduce thermal stress further, and should allow for more-compact heat exchangers.
REFERENCES
- 1. B. Badger, R. W. Conn, etc., A High-performance Non-circular Tokamak Power Reactor Design Study—UWMAK-III, UWFDM-150, University of Wisconsin, 1976.
- 2. A. M. Russell and K. L. Lee, Structure-Property Relations in Nonferrous Metals, Wiley 2005.
- 3. J. Bevk, J. P. Harbison, and J. L. Bell, “Anomalous Increase in Strength of In Situ Formed Cu—Nb Multifilamentary Composites” J. Applied Phys. 49 (1978) 6031-6038.
- 4. A. M. Russell, L. S. Chumvley, and Y. Tian, “Deformation Processed Metal-Metal Composites,” Advanced Engineering Materials, 2 (2000) p. 11-22.
Claims
1. A heat exchanger comprising a copper-refractory metal composite that defines a structure of the heat exchanger.
2. The heat exchanger of claim 1, where the refractory metal is niobium
3. The heat exchanger of claim 1, where the refractory metal is vanadium
4. The heat exchanger of claim 1, where the refractory metal is chromium
5. The heat exchanger of claim 1 where the composite is worked to a true-strain in excess of 2.
6. The heat exchanger of claim 1, where the composite is worked to a true strain in excess of 5.
7. The heat exchanger of claim 1 where the refractory metal content is between 10% and 60%.
8. The heat exchanger of claim 1, where the refractory content is between 20% and 50%
9. The heat exchanger of claim 1 as part of a rocket engine.
10. The heat exchanger of claim 1, where the exchanger serves as a fusion reactor first wall.
11. The heat exchanger of claim 1, where the exchanger is part of a jet engine.
12. The heat exchanger of claim 1 as part of a gas turbine system.
13. The heat exchanger of claim 1 where the exchange surface is in the form of a tube made of the composite.
14. The heat exchanger of claim 1 where the exchange surface is in the form a flat sheet.
15. Heat exchanger of claim 1, where the surface is coated with a layer of a metal of nickel, molybdenum, beryllium, or tungsten.
16. Heat exchanger made of claim 1 where the refractory metal is tantalum, molybdenum, tungsten, iron, silver, or a combination thereof.
17. A process of forming a refractory nanofilament filled copper matrix comprising:
- fast casting a molten mixture comprising a refractory copper mixture of more than 20% refractory to form a casting; then mechanical working said casting to form dendrites whose mean diameter is less than 10μ.
18. The process of claim 17 wherein said mechanical working is rolling or drawing.
19. The process of claim 18 wherein said mechanical working is rolling or drawing and occurs without annealing.
20. The process of claim 17 wherein said dendrites have a mean diameter is less than 500 nanometers.
21. The process of claim 17 wherein said dendrites have a mean diameter of 40 to 400 atoms width and are formed of copper, refractory, or a combination thereof.
22. The process of claim 17 wherein said molten mixture consists essentially of copper with more than 20% percent niobium, vanadium, or chromium or a combination thereof.
23. The process of claim 17 wherein said molten mixture has a low oxygen content.
24. The process of claim 17 where said molten mixture consists essentially of copper with more that 20% tantalum, molybdenum, tungsten, iron, or a combination thereof
Type: Application
Filed: Nov 17, 2010
Publication Date: May 19, 2011
Inventor: Robert E. Buxbaum (Oak Park, MI)
Application Number: 12/948,454
International Classification: F01P 9/00 (20060101); B22D 25/00 (20060101); F28F 7/00 (20060101);
