| A liquid metal (LM) consisting of 1/3 weight fraction each of Pb, Sn, and Bi has been proposed as the bonding substance in the pellet-cladding gap in place of He. The LM bond eliminates the large DeltaT over the pre-closure gap. Because the LM does not wet either UO2 or Zircaloy, simply loading fuel pellets into a cladding tube containing LM at atmospheric pressure leaves unfilled regions (voids) in the bond. Calculations indicate that these void spaces lead to local fuel hot spots. Voids were eliminated during fabrication by first evacuating the rod loaded with solid alloy and a fuel stack, melting the alloy, pushing down the fuel stack to drive the LM into the gap, and finally applying at least 5 atm He overpressure. A 4-m long full-scale fuel rod using this fabrication technique was successfully demonstrated. A destructive examination showed that the bonding in 80% of the fuel rod remained completely intact. Some sections, however, contained small unfilled regions in the LM bond. Sufficient explanations are given to classify them as defects caused by rough handling of the fuel rod. Calculations showed that the pre-closure reduction in fuel temperature in the LM-bonded rod slowed fission-gas diffusion and so increased the time required to saturate the grain boundaries. Numerical calculations utilizing the NAG subroutines to solve the diffusion equation, with Speight's approximation to treat re-solution, showed that the delay afforded by LM-bonding could be as high as ∼ 1 year and as low as 2 days, depending on the fuel temperature history, linear heat rate, and fission gas diffusivity. FRAPCON and FRAPTRAN codes were modified to take into account the increased gap thermal conductivity. Three reactivity insertion accident cases were studied to determine the initial beginning of life gap size to avoid pellet-cladding mechanical interaction at burnup of 60 GWd/MTU. The largest gap size required is 340 mum. Application to commercial fuel manufacturing requires only minor modifications to the existing fabrication line. The most suitable NDE technique was X-ray radiography using a collimated X-ray beam to probe edge-on the region between the pellet surface and the cladding ID. |
