FRS FREng FCGI
Distinguished Research Fellow
Mechanical Engineering Department
Imperial College London
Fusion of Test and Analysis - Validation and Verification procedures for Structural Dynamics
Structural Dynamics has major activities in Validation and Verification. This involves (i) ensuring that the FE Model which has been developed for the design of the structure is good enough for that purpose and (ii) demonstrating that the product meets the ultimate objective in terms of structural performance. Validation tends to be Analysis-led and Verification is Test-led. In fact, both activities involve both test and analysis and it is the combination (or Fusion) of Test and Analysis which of the underlying theme in this paper.
Model Validation is a well-developed technology in structural dynamics in which predicted and measured values of the primary modes of the structure are compared and differences used to adjust the FE model to bring prediction and measurement are within the specified accuracy. It is also necessary that the model is adequate as well as accurate, in the sense that there must be enough variables to be representative of the structure’s physics. A problem arises if the model excludes finer details of geometry and in such cases the model must be refined before validation. In other cases, the structure may exhibit nonlinear features which are excluded from the FE model, and here the model must be upgraded to include higher-order elements. Identifying the features of model inadequacy is referred to as verification of the model and this inevitably requires additional testing to identify the defective parts of the model. In Product Verification, physical tests must be conducted to demonstrate that the structure will meet its structural performance specification. This involves (i) prediction of the safe working life for normal operating environments and (ii) prediction of abnormal operation which may involve structural failure of some components. It has been normal industrial practice for many years to reproduce in a laboratory environment the dynamic loading that the structure experiences in service. These tests have been notoriously unreliable, with significant over- and under- testing being the norm. In the past 5-6 years there has been development of what is now referred to as ‘Smart Dynamic Testing’. These are tests ‘designed’ to achieve the requirement of the verification test specification using a mathematical model of the test structure and the test set up installation. By introducing a significant analysis element into the test design, order-of-magnitude improvements both in accuracy and cost-effectiveness are achieved.
David Ewins studied Mechanical Engineering at Imperial College and then undertook PhD research at Cambridge University. He was appointed Professor of Vibration Engineering at Imperial in 1983 and set up the Centre of Vibration Engineering in 1990 with support from Rolls-Royce. He has held Visiting Professor appointments in France, Switzerland, USA and Singapore, where he was Temasek Professor from 1999 to 2002. He served as Pro Rector (International Relations) at Imperial College from 2001 until 2005. He was Chair of the Scientific and Technology Advisory Board for EU Clean Sky Programme from 2010 to 2014 and is currently active collaborating with and consulting for aerospace and defence industries worldwide – Rolls-Royce, MOD, NASA - both in respect of strategic research initiatives and teaching industrial practitioners of Structural Dynamics.
Stay up to date with our technology updates, events, special offers, news, publications and training
If you want to find out more about NAFEMS and how membership can benefit your organisation, please click below.
Joining NAFEMS© NAFEMS Ltd 2024
Developed By Duo Web Design