Characterizaton of Brazed joints of C-C composite to Cu-clad-Mo (5)

Thermal conduction in Brazed Joints: For heat rejection applications, the thermal resistance of the joined assembly relative to the resistance offered by the individual constituents is important. For one-dimensional steady-state heat conduction, the joined materials form a series thermal circuit with an effective thermal resistance, Reff, given from Reff=∑(△xi/Ki), where △xi and Ki represent the thickness and the thermal conductivity, respectively, of the ith layer. For the joints created in this work, △Xc-c=△xCu-Mo=0.25*10-2m, △xbraze ~100*10-6m, KTicusil=219 W/m-K, and KCusil-ABA=180 W/m-K. The value of the thermal conductivity of Cu-clad-Mo, KCu-Mo, varies with the clad layer thickness, and is taken from ref4; KCu-Mo varies from 138W/m-K to 235 W/m-K for 0 to 30% clad layer thickness. For C-C composites, KC-C is anisotropic and varies considerably; for example, for 2D and 3D composites, KC-C=60 and 190 W/m-K perpendicular and parallel to the carbon cloth at 500K, and for 1D composites, KC-C=300 W/m-K at 500K. Taking the average KC-C  to be 125 W/m-K for 2D and 3D composites, the effective thermal resistance of our joint assemblies can be computed for a range of clad layer thicknesses. The results shown in figure.8 indicate that Reff varies in the range 31.5 to 38.5*10-6 M2.K/W, and that there is insignificant difference between Ticusil and Cusil-ABA. Because the difference in selecting brazes to satisfy other criteria such as ductility and wetting characteristics without impairing the thermal conductivity and weight advantages of the joined materials.

calculated (a) strain energy and (b) effective thermal resistance in the C-C, Cu-clad-Mo joint

Fig.8-calculated (a) strain energy and (b) effective thermal resistance in the C-C, Cu-clad-Mo joint

Figure.8 also compares the Reff values of the joints to the Reff values of C-C and Cu-clad-Mo substrates of the same total thickness as the joined materials; the thermal resistance of the C-C block is about 40.8*10-6 m2.K/W and that of a Cu-clad-Mo substrate is 22.8*10-6 m2.K/W. The decrease in the thermal conductivity of our joints relative to an isolated Cu-clad-Mo substrate is compensated by a 39% decrease in the weight of the assembly. The Rule-of-Mixtures density of our joints is ~5,919 kg/m3 compared to a density of 9,937 kg/m3 for Cu-clad-Mo alone.

SImilar calculations for the thermal resistance of the joints can be made for 1D C-C composite joined to Cu-clad-Mo. For 1D C-C composites, the effective thermal resistance of the assembly wil be 19.9*10-6 m2.K/W, and for a C-C substrate of the same total thicknesss as the joined assembly, the thermal resistance will be 17.0*10-6 m2.K/W, which is only about 18% less than the effective resistance of the assembly. These simplified thermal considerations illustrate the potential benefits of joining C-C to Cu-clad-Mo to create light-weight heat rejection systems.

Carbon-carbon composites with either pitch+CVI matrix or resi-derived matrix were joined to copper-clad molybdenum using two Ti-containing acive braze alloys. Large-scale braze penetration of the inter-fiber spaces in the CVI C-C composites was observed. The SEM and EDS examination of brazed joints revealed good interfacial bonding in all C-C/Cu-clad-Mo joints, some diffusion and redistribution of alloying elements, and preferential segregation of Ti at the composite/braze interface. The distribution of microhardness across the joints was reproducible, consistent with the Ti content in the braze, and indicated sharp gradients at the Cu-clad-Mo/braze interface. The metallurgically sound composite joints produced in this work, and the projected benefits of reduced thermal stress and thermal resistance, suggest that C-C composite/Cu-clad-Mo joints may be attractive for potential applications in thermal management systems.


ADD: Yizhuang Economic Development Zone, Beijing 100176, China.
Fax: +86 10 80828912
Marketing center: +86-18618169878
Human Resources: +86-15313026852