Verification and Validation in Analysis

Understanding the Foundation of Engineering Analysis Quality

In the complex world of engineering analysis, the difference between a successful project and a catastrophic failure often lies in two critical processes: verification and validation. For engineering firms across Atlantic Canada, from the bustling ports of Halifax to the industrial centres of New Brunswick and the manufacturing facilities throughout Nova Scotia, these quality assurance methodologies form the backbone of reliable engineering practice.

Verification and validation (V&V) are frequently mentioned together, yet they serve distinctly different purposes in the engineering workflow. Understanding these differences—and implementing them effectively—can mean the difference between a structure that stands for decades and one that fails prematurely. For industries operating in the Maritime provinces, where harsh weather conditions, salt air corrosion, and extreme temperature fluctuations present unique challenges, rigorous V&V processes are not merely best practices; they are essential requirements.

Defining Verification and Validation: The Critical Distinction

Before diving deeper into implementation strategies, it is essential to establish clear definitions for these two interconnected but separate concepts.

Verification: Are We Solving the Equations Correctly?

Verification addresses the accuracy of the computational solution itself. It asks the fundamental question: "Are we solving the mathematical equations correctly?" This process ensures that the numerical methods, software algorithms, and computational procedures produce accurate solutions to the governing equations.

Verification encompasses two primary components:

• Code Verification: Confirms that the software correctly implements the intended mathematical models and algorithms without programming errors

• Solution Verification: Quantifies the numerical errors in a specific computation, including discretisation errors, iterative convergence errors, and round-off errors

For finite element analysis (FEA), code verification might involve testing the software against benchmark problems with known analytical solutions. Solution verification, meanwhile, examines mesh convergence studies, time-step sensitivity analyses, and iterative solver tolerance assessments.

Validation: Are We Solving the Right Equations?

Validation addresses the physics of the problem. It asks: "Are we solving the right equations to represent reality?" This process compares computational results against experimental data or real-world observations to determine whether the mathematical model adequately represents the physical phenomenon being analysed.

Validation activities typically include:

• Comparing simulation results with physical test data

• Assessing model assumptions against observed behaviour

• Quantifying the agreement between predictions and measurements

• Identifying limitations in the model's predictive capability

The V&V Framework: A Systematic Approach

Implementing effective verification and validation requires a structured framework that integrates seamlessly with the engineering analysis workflow. The American Society of Mechanical Engineers (ASME) V&V standards, particularly ASME V&V 10-2019 for computational solid mechanics and ASME V&V 20 for computational fluid dynamics, provide comprehensive guidelines that engineering firms throughout Canada have adopted.

Establishing the Validation Hierarchy

A well-designed validation programme follows a hierarchical approach, progressing from simple unit problems to complete system validation:

• Unit Problems: Simple configurations with well-characterised physics, typically involving single phenomena with controlled boundary conditions

• Benchmark Problems: More complex scenarios that combine multiple physical phenomena while still allowing detailed experimental characterisation

• Subsystem Validation: Testing of component assemblies under representative loading conditions

• Complete System Validation: Full-scale testing under operational conditions

For a typical structural analysis project in Nova Scotia—perhaps analysing a marine vessel component or an offshore platform element—this hierarchy might begin with simple material characterisation tests, progress through component-level testing, and culminate in full-scale load testing of the assembled structure.

Quantifying Uncertainty

Modern V&V practice recognises that both simulations and experiments contain uncertainties. Effective validation must account for:

• Input Uncertainty: Variations in material properties, geometric tolerances, and loading conditions (typically ±5% to ±15% for material properties)

• Numerical Uncertainty: Errors arising from discretisation, typically reduced to below 2% through proper mesh refinement

• Experimental Uncertainty: Measurement errors and test variability, often ranging from ±3% to ±10% depending on the measurement technique

Practical Implementation in Engineering Analysis

Translating V&V principles into everyday engineering practice requires systematic procedures that can be consistently applied across projects. For engineering firms serving the Maritime provinces, where projects range from agricultural equipment design to marine infrastructure analysis, adaptable V&V protocols are essential.

Verification Procedures for Finite Element Analysis

A robust verification procedure for FEA should include the following elements:

Mesh Convergence Studies: Systematically refining the mesh until the solution becomes independent of mesh density. Industry best practice suggests achieving less than 2% change in key results between successive refinement levels. For structural analyses, this typically involves examining:

• Peak stress values at critical locations

• Displacement at key reference points

• Reaction forces at support locations

• Natural frequencies for dynamic analyses

Element Quality Metrics: Monitoring element quality parameters such as aspect ratio (preferably below 5:1 for accuracy), skewness (below 0.85 for most elements), and Jacobian ratio (above 0.3 for reliable results).

Benchmark Validation: Running standardised test cases with known solutions. The NAFEMS (National Agency for Finite Element Methods and Standards) benchmarks provide an excellent resource, with over 40 documented test cases covering linear statics, dynamics, thermal analysis, and nonlinear behaviour.

Validation Strategies for Real-World Applications

Validation connects computational predictions to physical reality. For projects in Atlantic Canada, validation activities might include:

Strain Gauge Measurements: Installing strain gauges at predicted high-stress locations and comparing measured strains to FEA predictions. Agreement within 10-15% is typically considered acceptable for complex structures, while simpler configurations should achieve agreement within 5-10%.

Displacement Monitoring: Using dial indicators, LVDTs, or digital image correlation (DIC) systems to measure structural deflections. Modern DIC systems can achieve measurement accuracies of ±0.01 mm, enabling detailed validation of complex deformation patterns.

Modal Testing: Comparing predicted natural frequencies and mode shapes against experimental modal analysis results. Frequency agreement within 5% and Modal Assurance Criterion (MAC) values above 0.9 indicate good model correlation.

Industry-Specific Applications in the Maritime Context

The unique industrial landscape of Atlantic Canada presents specific V&V challenges and opportunities across various sectors.

Marine and Offshore Structures

Nova Scotia's significant marine industry, including shipbuilding, offshore energy, and port infrastructure, demands rigorous V&V for structures exposed to dynamic wave loading, corrosion, and fatigue. Analysis validation for these applications often requires:

• Comparison with Classification Society rules (DNV, Lloyd's Register, ABS)

• Correlation with sea trial measurements

• Long-term fatigue monitoring programmes

• Environmental load factor validation through met-ocean data comparison

Cold Climate Considerations

Maritime winters subject structures to temperature variations that can exceed 50°C annually (from -25°C to +30°C), creating thermal stresses that must be accurately modelled. Validation activities should include:

• Material property verification at low temperatures (particularly for fracture toughness, which can decrease by 30-50% below the ductile-brittle transition temperature)

• Thermal expansion coefficient validation

• Ice load analysis verification against CSA S6 and Transport Canada guidelines

Agricultural and Food Processing Equipment

The agricultural sector throughout Nova Scotia and the broader Maritime region relies on specialised equipment that must withstand demanding operational conditions. V&V activities for this sector include:

• Fatigue life validation through accelerated life testing

• Vibration analysis correlation with field measurements

• Hygiene design verification for food contact surfaces

• Corrosion resistance validation for chemical exposure scenarios

Documentation and Quality Assurance

Comprehensive documentation transforms V&V from an ad-hoc activity into a systematic quality assurance process. Professional engineering practice in Canada, governed by provincial engineering acts and reinforced by Engineers Nova Scotia guidelines, requires traceable documentation of analysis assumptions, methods, and results.

Essential Documentation Elements

A complete V&V documentation package should include:

• Analysis Plan: Defining objectives, methods, acceptance criteria, and V&V activities before analysis begins

• Model Description: Documenting geometry simplifications, material models, boundary conditions, and loading assumptions

• Verification Records: Mesh convergence plots, benchmark comparisons, and solver diagnostics

• Validation Evidence: Test data comparisons, correlation metrics, and uncertainty quantification

• Results Summary: Key findings, limitations, and recommendations for use of the analysis results

Software Quality Assurance

Engineering software used for analysis should undergo its own verification programme. Commercial FEA packages such as ANSYS, Abaqus, and SolidWorks Simulation maintain extensive verification test suites, but users should still perform application-specific checks. This includes:

• Documenting software version and build information

• Recording solver settings and convergence parameters

• Maintaining benchmark test libraries specific to common analysis types

• Implementing periodic software verification reviews

Emerging Trends and Future Directions

The field of verification and validation continues to evolve, driven by advances in computational capability, measurement technology, and regulatory requirements.

Digital Twin Integration

Digital twin technology, which combines real-time sensor data with computational models, is creating new paradigms for continuous validation. For infrastructure projects across Atlantic Canada, digital twins enable ongoing model calibration based on operational measurements, improving prediction accuracy over the asset lifecycle.

Machine Learning and V&V

Machine learning techniques are increasingly being applied to V&V activities, including automated mesh quality assessment, surrogate model development for uncertainty quantification, and pattern recognition in validation data comparison. However, these tools introduce their own verification challenges that must be carefully addressed.

Regulatory Evolution

Canadian regulatory bodies and industry standards organisations are continuously updating V&V requirements. The National Building Code of Canada, CSA standards, and provincial engineering guidelines increasingly reference V&V best practices, making systematic implementation essential for regulatory compliance.

Partner with Experts in Engineering Analysis

Implementing effective verification and validation procedures requires both technical expertise and practical experience. At Sangster Engineering Ltd., our team brings decades of combined experience in engineering analysis across diverse industries throughout Nova Scotia and Atlantic Canada. We understand the unique challenges posed by our regional environment—from marine corrosion to extreme temperature variations—and incorporate rigorous V&V practices into every analysis project.

Whether you require structural analysis for marine applications, thermal assessment for industrial equipment, or fatigue evaluation for mechanical systems, our systematic approach to verification and validation ensures that your engineering decisions rest on a foundation of validated, reliable analysis. Contact Sangster Engineering Ltd. today to discuss how our expertise in professional engineering analysis can support your next project with confidence.

Partner with Sangster Engineering

At Sangster Engineering Ltd. in Amherst, Nova Scotia, we bring decades of engineering experience to every project. Serving clients across Atlantic Canada and beyond.

Contact us today to discuss your engineering needs.