In preparation for testing a lithium-helium heat exchanger at Sandia, unexpected rapid failure of the mild steel lithium preheater due to liquid metal embrittlement occurred when lithium at ~400 °C flowed into the preheater then at ~200 °C. This happened before the helium system was pressurized or heating with electron beams began. The paper presents an analysis of the preheater plus a discussion of some implications for fusion.
This paper describes the liquid metal integrated test system (LIMITS) at Sandia National Laboratories. This system was designed to study the flow of molten metals and salts in a vacuum as a preliminary study for flowing liquid surfaces inside of magnetic fusion reactors. The system consists of a heated furnace with attached centrifugal pump, a vacuum chamber, and a transfer chamber for storage and addition of fresh material. Diagnostics include an electromagnetic flow meter, a high temperature pressure transducer, and an electronic level meter. Many ports in the vacuum chamber allow testing the thermal behavior of the flowing liquids heated with an electron beam or study of the effect of a magnetic field on motion of the liquid. Some preliminary tests have been performed to determine the effect of a static magnetic field on stream flow from a nozzle.
In the authors initial high heat flux tests on small mockups armored with W rods, done in the small electron beam facility (EBTS) at Sandia National Laboratories, the mockups exhibited excellent thermal performance. However, to reach high heat fluxes, they reduced the heated area to only a portion ({approximately}25%) of the sample. They have now begun tests in their larger electron beam facility, EB 1200, where the available power (1.2 MW) is more than enough to heat the entire surface area of the small mockups. The initial results indicate that, at a given power, the surface temperatures of rods in the EB 1200 tests is somewhat higher than was observed in the EBTS tests. Also, it appears that one mockup (PW-10) has higher surface temperatures than other mockups with similar height (10mm) W rods, and that the previously reported values of absorbed heat flux on this mockup were too high. In the tests in EB 1200 of a second mockup, PW-4, absorbed heat fluxes of {approximately}22MW/m{sup 2} were reached but the corresponding surface temperatures were somewhat higher than in EBTS. A further conclusion is that the simple 1-D model initially used in evaluating some of the results from the EBTS testing was not adequate, and 3-D thermal modeling will be needed to interpret the results.