b'Heat SourceMultiphysics modeling and simulation determine heat source intensities when Identification Usingdirect temperature measurements are not possible.Inverse Optimization M odeling and simulation can accurately predict the temperature distribution in a structure from a known heat source. In this project, researchers examined the inverse problemdetermining the TOTAL APPROVED AMOUNT:properties of the heat source from a set of temperature measurements. The solution $144,000 over 1 year to these types of inverse parameter estimation problems can be written as a partial PROJECT NUMBER:differential equation constrained optimization problem. The partial differential 20A1052-022 equation describes the coupled physics of the processes, and modeling and simulation can identify the parameters of the partial differential equation that most PRINCIPAL INVESTIGATOR:closely fit the measurements. This work implemented the optimization algorithms Lynn Munday required for solving the inverse problem of reconstructing a heat source that CO-INVESTIGATORS: produces a known temperature distribution using gradient-based optimization. This Chandrakanth Bolisetti, INL project developed a proof-of-concept design for a general gradient-based partial Daniel Schwen, INL differential equation constrained optimization framework in our MOOSE-based Dewen Yushu, INL application named InverSe OPtimizatiOn and Design (ISOPOD).Zachary Prince, INL ISOPOD provides a MOOSE executioner interface to the Portable, Extensible Toolkit for Murthy Guddati,Scientific Computation (PETSc) Toolkit for Advanced Optimization. The PETSc Toolkit for North Carolina State University Advanced Optimization provides several Hessian, gradient, and gradient-free optimization algorithms that can be used in ISOPOD. ISOPOD provides gradient information to the Toolkit for Advanced Optimization by solving the adjoint of the original physics problem. The ISOPOD application implements this software in the context of existing MOOSE objects, including a new executioner, reporter, Dirac kernel, and MultiApp with transfers. Documentation and testing of ISOPOD is done within the MOOSE markdown and testing system with plans of merging this into the main MOOSE code base, providing users with an entirely new type of modeling and simulation capability. The multiphyisics nature of MOOSE will make this new capability extensible to parameterization of other multiphysics processes. For example, this work can be extended to other optimization problems, such as topology optimization and material design.INTELLECTUAL PROPERTY:ISOPOD is open-source software in GitHub.Steady-state temperature (a) and velocity (b) distributions in a molten salt45fast reactor obtained from the coupled multiphysics Griffin/Pronghorn simulation with nominal material properties.'