This presentation was made at CAASE18, The Conference on Advancing Analysis & Simulation in Engineering. CAASE18 brought together the leading visionaries, developers, and practitioners of CAE-related technologies in an open forum, to share experiences, discuss relevant trends, discover common themes, and explore future issues.
In order to ensure profitability for the entire product life cycle, engineers need to optimize preventive maintenance and downtime. Multi physics simulations have long been used optimize product performance during the initial development but their relatively large investments in time and effort have slowed their use with applications involving predictive maintenance and prognostic health monitoring. Reduced order models (ROMS), coupled with systems simulation show promise to deliver fast, accurate data. For example, full models which took hours on multiple CPUs can now be achieved using ROMS within minutes or even seconds on single CPU. These fast results can be extremely useful to predict the cause, location and even time of failures or turned into performance charts for field operators.
We will present heat exchangers as an example. They are commonly used in industrial settings including refineries and chemical plants. Companies spend multimillions of dollars on unscheduled maintenance and shutdown due to heat exchanger failure. We will show examples of how multi physics simulation of various failure modes was used to help determine the cause of the failures, location of failures, even be used for testing failure mitigation methods. Computational fluid dynamics (CFD) modelled fouling and corrosion on heat exchanger surfaces such as tubes, baffles and shell. Effects of operating conditions like inlet flow rates were also shown on the heat transfer characteristics and failure modes. The CFD results (thermal and hydrodynamic loads) were then transferred onto FEA solvers to analyse the fatigue life and buckling of the metal surfaces. These failure modes were then connected to the system level digital plant via the use of Reduced Order Models (ROM) to perform operational optimization. Multiple approaches to create and use ROMs were highlighted in the paper. These ROMs are not just input-output signal type models but complete three dimensional field view models.
|Date||7th June 2018|