Verification and Validation of Results

Verification and validation are independent procedures that are used together to ensure that a code meets requirements and specifications, and that it fulfills its intended purpose. In simple terms, validation can be expressed by the query ”Are you building the right thing?” and verification by the query ”Are you building it right?”

In other words, verification determines whether the programming and computational implementation of the conceptual model (that is, the partial differential equation system) is correct by comparison to exact analytical results. For example, does the temporal and spatial rate of convergence match the theory? Verification is more concerned with mathematics than engineering. By contrast, validation determines whether the computational simulation agrees with physical reality by comparison to experimental results. For example, in quenching, does the temporal variation of the surface temperature match experimental results?

Because experimental data is not always available for complex multi-physics applications, you must use a building-block approach consisting of phases that involve successively more complex flow physics, geometry, and their interactions. When validating multi-physics problems combined with complex geometry and forcing, the initial validation problem should be as simple as possible and only involve single physics and simple geometry. For example, if you are interested in forced flow over a cooled turbine blade, you could start with isothermal flow over a flat plate. Once the simulation code has been validated using single physics and simple geometry, you can add more physics and increased geometrical complexity until the actual complete system is being simulated (that is, all of the relevant physics and geometry are present).