Text Box: Verification and validation consulting Consulting services   Introduction  Computational models and simulations need to be shown to be fit for their intended use. Sources of error, uncertainty and mistakes inevitably arise in the computational modelling process and need to be identified, and appropriate techniques applied to ensure their reduction, treatment and future avoidance. It is imperative that results generated by computational models are as reliable as possible and that, prior to application of the results, it is demonstrated that the model is indeed fit for its intended use, i.e. the validity of the model needs to be established. It must, with respect to accepted confidence limits, accurately replicate the behaviour of the simulated phenomenon. Model verification is first carried out to confirm that the computational model accurately represents the specified mathematical model, that the computed solution is of sufficient accuracy, and that the software used to process the simulation is sound. Following model verification, model validation is carried out to confirm, by some independent means, that the model, data, and solution, are indeed adequate for supporting the problem assessment conclusions. The stringency of the validation will largely be dictated by the risk associated with the end-use of the engineering application. Services offered  Compusis can assist in model verification and validation activities by carrying out a systematic model reliability assessment. A validation report will be produced, detailing the validation assessment and its conclusions. Definition of acceptance criteria Category of simulation importance Verification and validation acceptance criteria Results acceptance criteria Verification procedures Code verification Calculation verification Validation procedures Assessment of idealisation assumptions Assessment of data assumptions Results corroboration Experimental data Experiential data Alternative models Verification and validation report Procedures applied Assessment summary Validation conclusions

Summary

Computational models need to be shown to be fit for their intended use.

Sources of error, uncertainty and mistakes inevitably arise in the computational modelling process and need to be identified. Appropriate techniques need to be applied to ensure their control and reduction.

© Copyright 2010 Compusis

Consulting

 

Modelling and analysis consulting

 

Verification and validation consulting

 

Quality management consulting

 

 

 

Enquires

 

 

 

computational modelling, simulation and analysis

compusis

Consulting

Contact

Short courses

Home