Coupled Thermal-Mechanical Model of the LORELEI Experimental Device with Multiple Interactions and Heat Sources during Controlled Transients and Accident Scenarios

Introduction: A primary component within the LORELEI experimental device [1] designed for single-rod Loss of Coolant Accident (LOCA) experiments in the French JHR research reactor, is a double-wall pressure flask with unique geometry and loading conditions containing the fuel rod and a peripheral heater. During the experiment sequence, the fuel power is controlled by moving the entire systems position relative to the reactor core. Heat transfer between the fuel, a surrounding heater (and instrumentation holder) and the double-wall flask result in significant axial and azimuthal temperature variations over the flask walls. The resulting temperatures are also influenced by the gamma heating of the structure materials, and the Zirconium cladding exothermal oxidation reaction and ballooning during the transient. In Addition, the system is also subjected to several notable external loads and constraints.

Methods: The Components Design-by-analysis (Nuclear ISE2 Safety class) initially relied on separate heat transfer and temperature-displacement analysis (uncoupled). This method however, produced relatively large displacements resulting in significant contact between the inner and outer flask walls, which in turn should have a strong effect over the resulting temperature field. For this reason, a fully coupled thermal and mechanical model of the LORELEI experiment device is developed, using a transient ABAQUS/Implicit solver with several tailored user-subroutines. The model employs a "simplified" approach for heat transfer over a gap (Radiation and Conduction/Convection) between cylindrical bodies using controlled (programmed) interactions, to accurately enable transient heat transfer between the deformable surfaces. Volumetric heat generation (nuclear, gamma, oxidation) is computed in user-subroutine HETVAL for each material model using several solution dependent variables (SDV`s), Common block parameters and spatial functions, defined in user-subroutines USDFLD and UAMP. The device velocity, position, and the heater power are controlled by two independent PID controllers using temperature sensors on the cladding and heater hot-spots.

Results and Conclusions: The fully-coupled simulation displays notable differences in the temperature variation over the flask walls, and helps examine the influence of the existing pre-loads, cladding ballooning, added design features (such as centering pins) and the desired controller functions. LORELEI design calculations and safety analysis are based on qualified models. At this stage in development, the model is employed in evaluating safety margins and possible optimization that would result from future qualification of the model. In this presentation, special challenges in the development of the model and important findings will be discussed.

[1] L.Ferry, D.Parrat, C.Gonnier, C.Blandin, Y.Weiss, A.Sasson. "The LORELEI Test Device for LOCA Experiments in the Jules Horowitz Reactor" WRFPM 2014