Verification and validation (V&V) in Computer-Aided Engineering (CAE) is very important to ensure reliability and accuracy of simulation models. It establishes that the model simulates the real-life physical situation accurately. Verification process confirms whether the implemented mathematical models and computational algorithms describe the real-life situation correctly whereas validation process determines whether the simulation model represents the real-life situation accurately.

Verification involves error-checking and code validation. It is generally divided into two categories: code verification and solution verification. Code verification ensures that the software solves the model equations as intended. It ensures the correct implementation of the numerical algorithms and correction of any programming errors. An example of code verification is comparison of simulation results with analytical solution. Solution verification procedure checks the results of simulation model for accuracy. It involves estimation and reduction of discretization errors. An example of solution verification is mesh-sensitivity study.

While verification ensures correctness in model building, validation compares simulation results with experimental observations. Validation ascertains that the model behaves as intended under realistic conditions. Validation includes carefully designed experiments that replicate the parameters used and conditions simulated in the model. Differences observed between the simulation results and experimental data are then analyzed to identify potential sources of error which may be a result of model simplifications or inappropriate material properties, boundary conditions etc.

Verification and validation procedure increases confidence in the simulation results and, consequently, predictive capabilities of the simulation model. This is essential for making reliable engineering decisions. V&V also helps in identifying and mitigating errors at early stages of development. However, V&V may be challenging for very complex simulation models which often involve multi-physics or multi-scale phenomena. Sometimes, it may be difficult to obtain high quality experimental data for validation and, thus, quantify uncertainty in models. Therefore, many industries have or are in the process of developing standards and regulations that govern the V&V processes to ensure safety and efficacy.