Application of a recursive identification technique to noise analysis for fuel performance study

A noise analysis system based on various digital signal processing techniques has been developed for fuel performance studies using test rigs at the R2 reactor in STUDSVIK. This system has been applied to the investigation of the thermal and mechanical behaviour of in-reactor fuel rods with different design parameters and under various fuel irradiation conditions. A recursive identification technique based on an adaptive lattice filter has been developed as a tool for the noise analysis, where the identification model is updated at each data sampling point in order to track time-varying process. The method has been applied to a time constant estimation for the dynamic response of the coolant temperature, i.e., fuel thermal (FT) time constant, and that of cladding elongation, i.e., mechanical response time constant, to a rod power change under both transient and steady conditions during the fuel irradiation. The results demonstrated the usefulness of the technique for studying the fuel behaviour where the thermal and mechanical states are slowly time-varying due to, e.g., fission gas release, creep, swelling and so forth. The paper also presents some results of the FT time constant estimation from various experiments which have been conducted at STUDSVIK. (author)