Simulation is one of the most important factors in the product development process and is being increasingly deployed in the automotive industry. The process of digital, computer simulation has been in existence since the middle of the last century, and has been continually evolving. An important area in the evolution of simulation technologies is the requirement for the combined validation of multiple physical characteristics of products. This form of simulation, known as Multiphysics simulation provides well-rounded information of a product as it is being developed.
Auto Tech Review caught up with Dr Vineet Dravid, Managing Director, COMSOL India, to understand the evolution of Multiphysics simulation, which the company is known to be a pioneer of as well as its development along with the automotive industry megatrends. We also find how new-age technologies are being introduced into simulation solutions so as to offer more accurate and quicker forms of testing and validation for customers.
MULTIPHYSICS SIMULATION EVOLUTION
The method of coupling physical phenomena in computer simulation and the studying of these multiple interacting physical properties is called Multiphysics simulation. All that takes place around us can be described using physical laws, and they do not occur in isolation in the real world. The world and all that takes place within it is Multiphysics in nature. The benefits of a Multiphysics capable simulation tool are that the important aspects of any design can be captured accurately. The solutions offered by companies like COMSOL include tools to address various physical criteria pertaining to general, electrical, mechanical, fluid and chemical areas.
There has been significant improvement in simulation technologies, firstly because of the advancements in the supporting architecture, which is the hardware, noted Dr Dravid. At the advent of digital simulation, huge supercomputers housed in warehouses were carrying out verification and validation tasks. Even simple structural simulations used to take days to finish then, he pointed out. The industry has come a long way since. Hardware is cheap and easily accessible, making it easy to perform huge computations.
Alongside this, different physics could not be coupled and solved simultaneously. For example, if one wanted to couple structural mechanics and heat transfer, it was not possible to do since the numerical techniques did not exist, explained Dr Dravid. However, the technology today has advanced to a state, wherein there is a possibility to couple any physical phenomena with another.
The other aspect is that of usability, he observed. As with everything else, a lot of improvement has taken place with regards to ease of software usability and given this scenario, the level of expertise required to carry out simulation is not high. It has become easier to use simulation technologies that also have become more robust, which is definitely an advantage because one can get more accurate results faster without relying on an individual for expertise, he noted.
MULTIPHYSICS vs. TRADITIONAL SIMULATION
The usage of Multiphysics simulation versus other traditional forms of simulation depends on customer needs. Dr David said digitalisation has become an integral part of the whole product life cycle, especially with the requirement to speed up the development or design lifecycle.
As far as digitalisation needs are concerned, this is where a platform solution approach and the Multiphysics simulation come into play. There is a need to carry out experimentation since it can never be eliminated for product development. Dr Dravid said the idea is to capture the real life situation of experimentation on a computer and this can be done only by carrying out Multiphysics simulation.
When different physics are simulated in isolation, it will never be possible to capture the real-life phenomena using these discrete tools, he explained. There is a need to carry out all these simulations simultaneously, to achieve the objective of obtaining real-life accuracy on a computer.
FUTURE OF SIMULATION TECHNOLOGY
The future developments in the area of simulation technologies will rely heavily on the needs and requirements of customers. Dr Dravid said the priority of consumers today is to reduce the entire design cycle, for which it is inevitable that they have to resort to seamless integration of simulation technologies into their entire design as well as product life cycle, even post-sales service. Simulations will be an integral part of predictive maintenance and post-sales services in future, he said.
However, each company needs to figure out how they would integrate simulation into their design and maintenance philosophies. Dr Dravid noted that this integration of simulation is interesting since different companies
are embracing different strategies on this front. Companies that will end up being successful will be those that can integrate simulation seamlessly into their entire product life cycle from design to post sales, he asserted.
Inversely, simulation companies will have to adapt to this change from their customers. Simulation providers will have to come up with products that are robust, easy to use, and provide customers with accurate results for their users’ accurate results, in no particular order. The simulation solutions need to be fast, affordable and run in real-time.
In addition, artificial intelligence and machine learning are the buzzwords now, which will be relevant in the future. Dr Dravid said work has already begun on the method of integrating traditional and existing simulation strategies with machine learning. In terms of pure design itself, data gained from simulation can be combined with machine learning algorithms to provide features like extrapolating results for a larger range of testing and validation. While the use of such new-age technologies with simulation is in a nascent stage now, it is something that will catch on soon in a big way, he said.
AUTONOMOUS & ELECTRIC MOBILITY
There are no special processes or advantages that simulation can offer in the area of autonomous vehicles or electric vehicles, said Dr Dravid. For instance, in the case of an electric vehicle, there is no real difference in simulation from internal combustion engine vehicles, in terms of the design approach. It is a different technology where the workflow will remain the same; however, the challenges will be different, he noted. The same is true with the autonomous driving as well, where there will be new challenges for which simulation will be important.
The philosophy of testing will remain the same in the case of autonomous vehicles, with a larger degree of data required to be captured through predictive simulation. Of course, machine learning is going to be critical for autonomous driving, and there has already been a tremendous amount of progress in that aspect, Dr Dravid said. In the case of autonomous driving, machine learning will be applied for carrying out decision-making of the vehicle.
It is clear that the role of simulation technologies is only gaining more prominence in the entire product life cycle, rather than just the design cycle. Consequently, the importance of Multiphysics simulation is growing at a more rapid pace, with the requirement of integration of such technologies by customers into their product life cycle. In addition to this, the combination of simulation with new-age technologies of artificial intelligence and machine learning offer customers a range of validation solutions that will be more reliable and accurate than ever before.
It is also being seen that a large number of engineering service providers have transitioned to offering Multiphysics modelling capabilities in their tools, which also implies that this form of simulation technology is witnessing increasing importance and adoption across industries, especially automotive.
TEXT: Naveen Arul