AsianScientist (Jan. 30, 2020) – Looking back on the past two centuries of industrialization, much has changed in terms of the economic value of a human being. In the 1800s, human labor used to matter most on the assembly line or manufacturing floor. Then along came machinery, and what workers could achieve with their hands became less important—and less valuable—than what they could achieve with their heads. Now, with digital technologies becoming an integral part of factories in the Fourth Industrial Revolution, even tasks that once solely belonged to the domain of grey matter are being taken over by intelligent systems.
A key addition to the ‘nervous system’ of modern industry is supercomputers, capable of performing thousands of trillions of calculations per second. The massive processing power of these devices promises to transform the way products are designed, engineered and maintained. Asia, home to more than half of the world’s supercomputers, is already benefitting from the use of supercomputers in a wide range of industries.
Crash test simulations
While no automobile manufacturer wants to see their product wrecked in a crash, they have a responsibility to ensure that their vehicles offer passengers an adequate degree of protection should a traffic collision occur. Crash testing is therefore routinely carried out by automobile manufacturers to ensure safety standards are met before a new model is introduced.
However, the very nature of crash testing means that dummy vehicles are built only to be destroyed, making the process expensive and time consuming. To reduce the number of dummy vehicles required, automobile manufacturers can tap on supercomputers for crash test simulations.
For instance, the K supercomputer in Japan is being used by Japanese carmakers to run crumple simulations of automobile chassis floors during collisions. This typically involves creating a digital representation of a car comprising more than a million components, then analyzing how these components behave under the extreme physical stresses of a crash. A wide range of parameters—such as chassis material and structure, as well as the speed of the car—can be modeled and monitored in silico, and adjustments can be made to the car’s design to optimize safety before actual crash testing is carried out.
Aerodynamics assessments
Since the Wright brothers built the world’s first successful airplane in 1903, the aviation industry has come a long way with larger aircraft capable of carrying heavier loads and traveling longer distances. Despite these improvements, aircraft manufacturers are not resting on their laurels and are seeking better airplane designs to reduce drag, which in turn translates to better fuel economy and lower operating costs.
Boeing and Airbus have dominated the aircraft manufacturing industry for decades, but a new contender has emerged in recent years—the Commercial Aircraft Corporation of China, Ltd., or Comac for short. Comac’s first model, the C919, is a narrow-body twinjet airliner that is 38.9 meters long, with a wingspan of 33.6 meters and a carrying capacity of 190 passengers.
The design of the C919 was informed by the Tianhe-2, ranked as the world’s fastest supercomputer from 2013–2016. Crucially, the Tianhe-2 was used for computational fluid dynamics simulations of the external flow field around the C919. According to personnel involved in the analyses, the highly precise calculations, which would have required two years on conventional computing systems, were completed in just six days on the supercomputer.
The C919 first took the skies in a test flight on May 5, 2017, but is still undergoing refinement at the time of writing. It is expected to make its commercial debut with China Eastern Airlines in 2021.
