231 0
5
surface removal, 1.2 milslyear. Chemical analysis of four different samples taken from the bismuth drained from the loop did not indicate any increase in the carbon concentration. These results indicate that graphite to be used in chemical processing applications should have small pore entrance diameters (<I p) and should not have interconnecting porosity. Since removal of rare earths from MSBR fuels involves bismuth that contains up to 50 mole 7% lithium, we are testing the various potential containers in bismuth-lithium solutions. Static capsules containing two grades of graphite have completed a 500-hr test in bismuth containing 0.5 and 3.5 wt % lithium but have not been examined yet. Two all-metal thermal-convection loops will soon be placed in operation. The first of these was fabricated from gun-drilled low-carbon, low-oxygen molybdenum bar stock and contains molybdenum tensile samples in both the high- and low-temperature regions. The second was fabricated from T-111 (Ta-8% W-2% Hf) alloy tubing and contains T-1 11 tensile samples. The bismuth in each of these loops will contain 2.5 wt % Li and will operate at a maximum temperature of 700째C. A graphite loop is currently being designed and will be operated under conditions similar to those discussed above. We are planning to continue testing these three potential structural materials (Mo, T-1 11, graphite) in various concentrations of lithium in bismuth to determine their physical and mechanical property changes. One point of interest is the possible formation of lithium carbide or intercalation compounds with graphite. Future tests will also include the compatibility of braze alloys with bismuth containing lithium. 16.6 CHEMICALLY VAPOR DEPOSITED COATINGS
J. I. Federer Althoug fuel, most iron- and nickel-base alloys are d by liquid bismuth. Tungsten and molybdenum are resistant to liquid bismuth, but they are much more . Therefore, we are investigating the molybdenum coatings on iron- and a possible solution to this difficult em. The coatings are being applied apor deposition techniques using hydroof WF6 and MoF6 at about 600 and spectively. We have contin coatings on small sheet-type spec attempting to demonstrate the applicability of the coating process by coating a variety of test vessels and more complicated shapes.
Tungsten coatings do not adhere to stainless steels unless the steels are first plated with a thin layer of nickel. Nickel is readily applied to simple shapes by electrodeposition, but complicated shapes are difficult to plate by this method. As an alternate method, we are investigating electroless nickel plating from an acidified phosphate bath. In our first experiment, bend specimens of types 304 and 430 stainless steels measuring 10 X 'h X in. were electroless plated with nickel, which was then bonded to the base materials by heating to 800째C for 30 min. Duplicate specimens of each steel were coated with about 0.005-in.-thick tungsten, and another pair was coated with about 0.005.in.-thick molybdenum. The tungsten coated smoothly, but the molybdenum coating had numerous blisters. Moreover, when subjected to a bend test, all of the coatings were less adherent than had been previously observed when an electrodeposited nickel layer was first applied. The electroless nickel plate contained about 8% P, which may have affected coating adherence. We are continuing this investigation by testing nickel plates of lower phosphorus content. The largest object that we have coated was a 4'12 -in.-OD by 36-in.-long Monel vessel closed on both ends except for '4 -in.-OD tube extensions. The entire inner surface was coated with 0.006- to 0.010-in.-thick tungsten except for the ends, where the coating was thinner. A 4-in.-long section of the coated vessel was thermal cycled 25 times between 25 and 600째C. Examination after the 25 cycles revealed the coating was intact and there was no evidence of spalling. A dye penetrant revealed only a few edge cracks, probably caused by cutting the section from the original vessel. The entire section was slightly out of round after thermal cycling, possibly as a result of the stress produced from the difference in thermal expansion between tungsten and Monel. Similar results would be expected for tungsten coatings on other mate rials such as Inconel 600, Hastelloy N, nickel, and nickel-plated steels. We have deposited molybdenum on the inside surface of a nickel capsule and tested it in bismuth for 700 hr at 600째C. No attack of the coating nor the nickel substrate below was found.7 Figure 16.15 is a schematic representation of a Hastelloy N corrosion loop that is to be coated with tungsten so that it can be tested in flowing bismuth. Since the nickel-base alloy substrate would be readily attacked by bismuth, a severe test of the integrity of the coating will be obtained. The loop was modified as shown in Fig. 16.15 for coating. A cross member was removed to form a U-shape, thus providing both a gas inlet and outlet. The U-shape will be coated first and