Mechanics & Materials
System reliability quantification using probabilistic analysis techniques.
The ability to quantify the uncertainty of complex engineered systems subject to inherent randomness in loading, environment, material properties, and geometric parameters is becoming increasingly important in design and certification efforts. Traditional design approaches typically use worst case assumptions and safety factors to certify a design. This approach is overly conservative, does not quantify the reliability, nor does it identify critical parameters or failure modes affecting the system performance.
A probabilistic analysis approach characterizes input variabilities using probability density functions and then propagates these density functions through the performance model to yield uncertain model outputs, which can be related to failure metrics such as fatigue life, rupture, dose, or stress intensity. The approach quantifies the reliability, can reduce over-conservatism, and identifies critical parameters and failure modes driving the reliability of the system.
Southwest Research Institute (SwRI) has been addressing the need for efficient probabilistic analysis methods and tools for more than two decades. Nearly all of the reliability technology developed and implemented by SwRI researchers is available in the NESSUS® probabilistic analysis software. NESSUS has been used to compute the reliability for components and systems in a range of application areas, especially where the performance models require multi-physics analysis packages and/or computationally expensive function evaluations. The reliability estimates are used alone or integrated into other traditional probabilistic risk assessment analysis packages. Some applications of probabilistic analysis include:
- Component and system reliability for existing and new designs
- Reliability-based optimization
- Reliability test planning
- Inspection scheduling
- Risk-based cost analysis
- Model verification and validation
Probabilistic analysis combines model input uncertainties with validated physical model to quantify reliability and corresponding sensitivities.
Probabilistic analysis methods and tools are becoming more mature with the majority of the effort of an analysis devoted to the probabilistic modeling. The engineers at SwRI bring a breadth and depth of knowledge to probabilistic modeling by uncertainty modeling and quantification:
- Identifying sources of errors and uncertainties
- Developing probability distributions for input variables
- Determining spatial and temporal variations
- Developing probabilistic load modeling
- Tailoring failure models for modeling uncertainty and obtaining appropriate system performance measures
- Creating system models (multiple failure mode and components)
The Mechanics and Materials Section team assists organizations with probabilistic and reliability analyses in different capacities:
- As the lead group that provides an end-to-end solution for the client
- As a partner with the client exploiting the expertise of all
- Develop tailored solution methods
- Perform comprehensive probabilistic modeling
- Integrate probabilistic and deterministic modeling
- Provide support and training for NESSUS
- As a consultant to provide NESSUS support
- Client maintains lead role
To complement these activities, the Mechanics and Materials Section also leverages these capabilities and expertise to:
- Develop tailored probabilistic analysis and design approaches
- Develop custom software tools
- Provide standard and specialized training for software usage, probabilistic modeling theory and design approaches (Probabilistic Design and Analysis short courses)
mechanics and materials • structural integrity • reliability assessment • mechanical behavior • mechanical characterization, fatigue life characterization • crack growth • corrosion fatigue • probabilistic mechanics • uncertainty modeling • bone fracture • bone properties