BENCHMARK July 2019

The Structural Integrity Issue

In the BENCHMARK July 2019 Issue, you will find out about:

• The Application of Finite Element Based Limit Load Analysis
• Modeling Self-loosening in Bolted Joints
• Modelling of Bones
• FEA Puzzler: How Confident Are You?
• Exploring the Design Freedom of Additive Manufacturing through Simulation
• Predicting the Behaviour of Reactor Core Components
• How to Model and Assess Welded Structures with Finite Element Analysis
• FEA Based General 3D Crack Propagation Simulation
• The Evolution of Structural Integrity Assessment via FEA
• What is Simulation Governance and Management?
• ML for AM
• Weathering Simulation Challenges
• Excel for Engineers and other STEM Professionals

Welcome to a special issue of Benchmark Magazine, themed around structural integrity. This issue has been put together by the Computational Structural Mechanics Working Group, and particularly by Adam Towse the deputy chair, to give a flavour of the many ways that simulation can help engineers to identify when things might fall apart, and enable them to design systems that avoid anarchy being loosed if the centre does not hold.

Structural integrity is a concern for a wide range of sectors; civil engineering is probably the most obvious, but all industries need to be confident that their products and systems are going to operate safely and successfully over a reasonable lifetime. The fact that most of us can go about our daily business knowing that our house roof isn’t going to cave in on us if the wind blows, our train isn’t going to shake itself to bits on the journey to work, and our nuclear power stations aren’t going to explode is largely down to structural integrity assessment. Testing plays a vital part in demonstrating structural integrity, but for large-scale objects such as buildings and high-value objects such as aeroplane engines, testing to failure is either impossible or so expensive that only a small number of tests can be carried out. Similarly inspection and maintenance ensure that integrity continues to be assessed once the object is in use, but inspection is costly, time-consuming, sometimes subjective and occasionally so hazardous as to be impossible, and maintenance is most effective when it is scheduled and preventative rather than reactive. Calculation and simulation are important tools for filling the gaps in testing, for looking where inspection cannot see, and for identifying when maintenance needs to happen. Simulation can also provide useful integrity information at the design stage, where the object cannot be tested because it does not exist.

Historically, structural integrity was achieved through experience and rules of thumb, and was later codified as safety factors and handbook solutions. Many magnificent constructions such as aqueducts, cathedrals, castles and bridges survive today, centuries after they were first constructed, in part because their creators applied large safety factors and so built them to last. The problem with large safety factors is that they prioritise strength over material usage, so that cost of material and the mass of the final structure are neglected. As resources become scarcer, industry is striving to minimize its production costs, and we are driven to consider the life cycle assessment of structures, efficient use of materials becomes more important. As Finite Element (FE) analysis has become more widely accepted as a reliable computational tool, the role it plays in assessment codes has changed too. Two of the articles in this issue discuss how well-designed FE simulations can be used within structural integrity assessment codes to reduce conservatism and so make better use of resources

Simulation for structural integrity assessment can involve aspects that are challenging to model accurately. Modelling material damage is effectively a multi-scale multi-physics problem: damage is initiated at the atomic level and can occur through thermal, chemical, or mechanical processes but integrity assessment requires its effects to be calculated at a macro scale. Methods used for joining components, such as welding, can also require multi-scale multi-physics models as they change the chemical composition, and hence the thermomechanical properties, of the material in the region of the joint. Time scales are a further challenge because damage is often a cumulative process that may take years to reach a critical level, but the loading that drives the damage generally occurs over a much shorter time frame. This aspect is a particular challenge for fatigue modelling under cyclic loading, and for high-temperature or high stress applications where creep deformation may be a concern.

One of the most common causes of failure is crack propagation. Many different damage mechanisms lead to crack formation, but the presence of cracks may not be a problem if they do not grow. The breadth of situations that undergo catastrophic failure through cracking has led to a corresponding breadth of research into how to model crack growth. Most of us know that stress concentrations occur at crack tips so you have to be careful with FE meshing in that region, but in most applications that is only the beginning of the story. As an introduction to the topic we have an overview of the key challenges and solution approaches, with some practical examples for illustration. We also have a separate article giving a deep dive into a specific problem with unique challenges and a safety critical application: modelling of damage to graphite bricks in nuclear reactors.

Joining mechanisms lead to further assessment challenges. Bolts, rivets and screws involve complex contact conditions and stress localisations, often requiring parameters such as friction coefficients that can be difficult to quantify. Joint failure mechanisms can be more complicated than simple stress-related failure, as is discussed in an article on modelling self-loosening of bolts. Even when the failure mechanism is straightforward, the challenges of describing how the bulk material interacts with objects such as screws can be difficult, highlighted in an article on simulation of bone screws. The prevalence of this sort of challenge will only increase as we move to a world where additively manufactured parts with complex inner structures are mechanically joined to other materials.

Structural integrity is a major interest of the computational structural mechanics working group, so we’ve also included an overview of the group and its current activities, and we’ve highlighted a forthcoming publication on modelling and assessing welded structures. If these activities are of interest, or something in the articles sparks your interest, please get in touch: the more of us that are stopping things from falling apart, the better!

Louise Wright, Chairman NAFEMS Computational Structural Mechanics Working Group

Louise Wright is the Science Area Leader for Modelling at the National Physical Laboratory, the UK’s national measurement institute. Following an MA in Mathematics and an MSc in Mathematical Modelling and Numerical Analysis, she spent four years working with FE and CFD in industry before joining NPL in 1999. Her work uses FE and similar methods to support experimental design, interpretation of measurement results, and solution of industrial problems. She is interested in improving confidence in use of FE results in decision-making processes and works on uncertainty evaluation applied to finite element models.