When developing or improving products, engineers face the challenge of analysing complex simulations to test their designs. High-performance computing (HPC) is used to solve major problems in engineering by performing demanding operations such as simulations relatively quickly. Also, by processing more precise operations, HPC can avoid the errors that might be introduced by humans.
Engineers use HPC mainly for performing simulations: these digital models, comprising huge amounts of numbers and algorithms, can determine a product’s viability. When performing simulations, the experts can track the behaviour of the product by including several components. When the simulations are complete and the results have been analysed, engineers can successfully improve or develop new products.
The following use-cases demonstrate how HPC can be used effectively within the engineering sphere. All use-cases presented in this article were carried out within the Fortissimo project (funded from the EU H2020 research and innovation programme).
Mechanical Engineering: HPC-based design of non-circular gears
Non-circular gears (NCGs) generate a prescribed motion with great precision, regardless of the external factors. These indispensable items are produced in several shapes and can fulfill various types of special motion requirements. NCGs are used in different application sectors such as automation & robotics, defence & security, aerospace, transport, and energy.
The design, development and testing of NCGs is a complex process that requires a significant amount of expertise and expensive, computationally-intensive iterations. Additionally, engineers must limit the time from design to delivery as much as possible.
In this use-case, an HPC-based engineering workflow for NCGs design called CloudGear was developed. CloudGear software allows the producer to optimise and design NCGs in less than 3 minutes (previously, the whole process took 25 minutes, not including the set-up and results from delivery). NCGs can now be designed faster and with more advanced features. The designing process has been automated through the use of the new software and the entire manufacturing process is faster, saving the producer time and money.
Automotive: HPC-based high resolution modelling of magnets
The automotive industry is changing extremely fast. The technological evolution of automotive design and enhancements in motor technology have led to the use of more magnets throughout the automotive system with sensor magnets, holding magnets, motor magnets, and generator magnets all implemented.
The production of magnets is an expensive and demanding process. Using a hydraulic press, pressure is applied to magnetic powders until they solidify. The challenge in this process starts with the optimisation of the pressing tool so it can be used for longer and with lower material costs, but doing this requires the ability to automatically detect yielding of the tool under a given pressure. Optimisations are run through the HPC simulation process, which requires qualified engineering experts and are time consuming.
To tackle this problem a set of software services based on open-source solutions was developed, along with a computer model of the pressing tool and its performance during the magnet production process. By employing an easy-to-use application on the HPC system, the magnets producer can now run simulations that are much more cost-effective, faster, and more precise. The magnet producer reduced material costs by around 32% and the cost of making the pressing tool by 27%. These savings also enable the creation of new products and services based on the improved pressing tool.
Maritime and civil engineering: Optimisation of a multi-body wave energy device
Wave energy converter systems have the potential to become a significant source of affordable renewable energy, especially if they can also reduce wave impact on beaches. However, the strong dependency on each local environment and its unique wave-induced forces raises several technical challenges for modelling such systems.
CCell is a curved wave-energy converter that moves with the waves to extract their energy for electricity. The low-voltage charge produced by the CCell converters can then be used by BioRock formations to create large “coral” habitats by transforming seawater minerals into limestone rocks. Those coral formations could build natural breakwaters to protect coasts from erosion.
It would be very expensive to use a physical model to test such a complex system. Another barrier to testing CCells is the variability of natural factors, e.g. wind speed, and different types of seabed and beach formations. To model the system, OpenFoam software running on HPC was used to simulate CCells, with an accompanying easy-to-use graphical user interface allowing the simulations to be set up quickly.
These simulations brought impressive results. Set-up time reduced from two hours to less than one minute. A nine-fold reduction in costs was gained by reducing the use of physical modelling, while productivity increased by a factor of seven. The producer significantly speeded up its market entry by around nine to twelve months, and this time will now be dedicated to the evolution of future products and service offerings.
The future of HPC usage in the engineering sphere
The world is ever more digitised, connected, and complex, but the flood of data that is generated can be overwhelming. Producers gather huge amounts of data to optimise and develop products. Their factories use IoT, machine learning, big data and artificial intelligence to better understand customer needs and product usage; they try to optimise production by learning how a products’ elements work, how materials behave and much more. With the help of reliable and powerful HPC systems, this data gains new meanings – they are interpreted quickly, evaluated, and used in the next phase of product development.
Accessing HPC facilities on-premises or in an HPC cloud is faster, easier and cheaper than ever. The exploitation of HPC within engineering tasks for simulations, visualisations or data management enables more complex products, increasing their capabilities and value.
The Covid-19 crisis has greatly affected engineering applications, but digital technologies are creating new opportunities and ways of responding. For example, the use of digital twins increased by nearly 300 percent a few months after the implementation of coronavirus physical distancing measures. Newly-developed platforms for Intelligent operations can help companies react faster using systems that have been developed to create the interoperable, distributed, data-driven network of information that is the backbone of remote information sharing. “Cloud-first” solutions will be further developed and highly consumed by companies, so creating new digital opportunities.
—Tina Črnigoj Marc, Arctur
 P. Reynolds, “Intelligent Operations Management During a Pandemic,” [Online]. Available: https://www.arcweb.com/blog/intelligent-operations-management-during-pandemic. [Accessed 2020].
Image Rights indicated as follows: The images are also provided on the Fortissimo and FF4EuroHPC projects’ websites.
Mechanical Engineering: STAM, Noesis
Automotive: Magneti Ljubljana, XLAB
Maritime and civil engineering: Zyba