Simulation has taken the centre stage of research, technology development and engineering in majority of the industrial sectors. Power of simulation has no bounds; it is transforming all facets of product life cycle, says Sameera Damle.

Any technology innovation and development, whether it is the control function, development of an aircraft, validation of autonomous car, development of personal protection equipment (PPE), planning of drug production, 5G network development, study of nuclear reaction, there is one thing in common - use of simulation.

In simple words simulation is the creation of a mathematical model that represents the physical world. Simulation models can be manipulated logically to analyse the behaviour of physical system & to understand how it works. The advancement in simulation technology allows us to model multiple physical systems, co-simulate them to understand their interactions and analyse the impact of one system on another. Digital twin technology enables us to build virtual system prototypes (replicas of physical system in virtual world) to optimise and validate system behaviour in real time. 

the centre stage of research, technology development and engineering. Today, the use of simulation is not limited to product design, it is being used in areas like ideation, manufacturing, operations & maintenance as well. Different industry segments like aviation, automotive, healthcare, energy, consumer goods, oil & gas, industrial equipment, chemical industry as well as defence & military have widely adapted simulation technology. 

The references to usage of simulation in automotive industry goes back to 1970s, long before the development of modern simulation tools. Initial applications included use of discrete simulation methodology in manufacturing to study new methods of manufacturing and material handling, in the design phase to create engineering drawings of components and system, analyse the design changes etc. Today, simulation is used for a broad variety of applications in automotive industry. 

from structural mechanics, fluid dynamics, and electromagnetics encompassing high frequency, signal & power integrity, electro-mechanical, electro-thermal, chip design, optical simulation, reliability engineering use simulation software for a variety of analyses. Multiphysics simulation is used to study complex system behaviour and optimize the system performance. 

Automotive OEMs & component suppliers use simulation technologies from component to complete system design. Today, simulation software providers are not just helping companies in the structural, fluids, electromagnetics, system simulation & software design but also working with R&D teams to build virtual prototypes, perform safety assessments, simulation data management, material information management and becoming a partner in building safer, reliable, cost effective and environment friendly products.  

The automotive industry is in the dawn of transformation today. New trends like Advanced Driver Assistance System (ADAS), Autonomous Vehicle (AV) along with shared mobility are set to change the way we are going to commute in future. The goal of making mobility environment friendly is driving the development of Electric Vehicles (EV) as well as bringing in stringent emission norms like BSVI (equivalent of Euro 6) on ICE vehicle. In addition to this the consumer expectation of more comfort features & lower operating cost are posing complex engineering challenges to the OEMs and the Tier1, Tier2 auto component suppliers. 

For an OEM, it is practically impossible to validate every scenario using prototype vehicles. Virtual prototypes & simulation models help OEMs not just in design & development but also for creation of complex scenarios to optimize the vehicle performance and validate edge cases. 

The auto component suppliers are expected to own much more responsibility in terms of component integrating in the system as well as fail safe operations. The component suppliers will have limited access or no access to proto vehicles. They rely largely on simulation technology. In many cases the suppliers are supposed to demonstrate the quality & reliability of product through simulation.

According to the WHO, road traffic crashes cost most countries 3% of their gross domestic product (Global status report on road safety 2018). The most common cause for the road accidents is human errors. In the United States more than 90 percent of crashes are caused by human errors says the report from National Highway Traffic Safety Administration. Several studies suggest that Advanced Driver Assistance Systems (ADAS) and Autonomous Vehicles (AV) have a potential to address this problem. Studies estimate significant reduction in road accident by using ADS & AV technology. On the contrary, there are questions on safe operation of autonomous vehicles. To harness the potential of ADAS & AV it is necessary to ensure that the autonomous vehicle perceives the environment around it, predicts possibilities and plans appropriate actions and actuates the vehicles controls. 

A report from RAND Corporation, an American non-profit global policy think tank, Autonomous vehicles would have to be driven hundreds of millions of miles and sometimes hundreds of billions of miles to demonstrate their reliability in terms of fatalities and injuries. To prove that autonomous vehicles would get into fewer serious crashes than human drivers would require a fleet of 100 autonomous cars traveling at 25 mph non-stop for 125 million miles which is 5.7 years of continuous driving. To provide same evidence for fatal crashes the fleet will have to be driven for 8.8 billion miles. It is practically impossible to run such fleet of vehicles. The only answer to this is simulation.

Perception and object recognition are crucial elements of the autonomous puzzle. It is extremely important to have precise perception to accurately recognize the objects and ensure safe operation. The artificial intelligence (AI) and machine learning (ML) algorithms support the perception functions to precisely recognize the objects. Training the machine learning algorithms is a complex task. The trained machine learning algorithms might easily recognize objects like road signs, traffic signals, human, car, truck, building; however it might struggle to recognize ‘edge cases’.

Edge cases are unusual situations which rarely occur but can lead to an accident since the autonomous vehicle is not able to handle the situation. The edge cases could occur because of several reasons like an unusual shape on the road, partially visible object, extreme weather etc. Baby stroller on road, man walking with a dog, dog behind a tree and flooding are also some examples of edge cases. 

In manual method, the developers would review the video captured while driving, label each unusual object or edge case then train the perception algorithm to identify the object. This is a humungous task and will take a lot of time and cost. Tools like Ansys SCADE Vision can make the edge case identification easy. It minimizes the cost of edge case resolution and perception system validation by automatically detecting the weaknesses in the systems. 

Solutions like Ansys VREXPERIENCE Driving Simulator provides the capability to simulate driving scenarios and enables testing of the autonomous stacks. Engineers model vehicle, road, traffic conditions, import map information, create scenarios, and test the ADAS & AV functions. Capabilities to create huge number of scenario variation will further enhance in validating the robustness of software. More accurate sensor simulation can be performed by using physics-based models. Closed loop simulation and virtual validation techniques like software in loop (SiL), hardware in loop (HiL) helps in reducing the testing time and creation of scenarios which might not be possible to create in real world. 

Though the power to wheel losses in an electric vehicle is much less compared to an internal combustion engine powered vehicle the EVs have a set of own challenges. Safety, range anxiety, availability of charging infrastructure, charging time, cost of the vehicle are some of the major aspects. These translate to safe operation of the vehicle, safety of battery, optimum performance of battery & motor as technical challenges to the engineer. 

Though batteries have been in use for decades it has been in stationary conditions. Usage of large high capacity batteries in motion pose safety risks. Safety aspects must be considered at subsystem level as well as system level, which means at battery level as well as complete vehicle level. Typical safety risk at battery level is thermal run over due to overheating, overcharging, cell degradation due to improper cell balancing etc. At a vehicle level several aspects like placement of the battery, impact of different environmental conditions, road profiles, impact of a crash, different stress & strain on the battery pack play a key role. 

State of Charge (SoC) which represents the available charge of the battery against its rated capacity and State of health (SoH) which represents the capacity of the battery are critical parameters. The Battery Management System (BMS) monitors parameters like SoC, SoH, and cell temperature and ensures proper operation of the battery. During the lifetime of battery, the health of the battery deteriorates due to physical & chemical changes. CFD modelling of cell electrochemistry can help in analysing aging behaviour. For example, Ansys Fluent includes a detailed electrochemistry model for optimizing lithium-ion battery cells, which may be used in the exploration of anode materials or to predict aging effects.

Analysing the impact of overheating, overcharging, cell imbalance, external forces, heat transfer between the cells etc with real batteries is practically impossible and pose higher safety risks. Deep analysis from the different aspects like structural, thermal, electrochemical & system are essential in battery research & development. Multiphysics simulation can help in coming up with the right battery design by optimizing trade-offs between cost, energy density, life cycle, operating temperature and safety. Accurate simulation of thermal behaviour can help in identifying the different triggers leading to thermal runaway. The reduced order models (ROM) brings the ability to create different scenarios in real-time enabling faster analysis. This can also be used for testing & validation of the BMS in real-time SiL & HiL system. 

Government policies on stringent emission regulations, consumer expectation of environment friendly & fuel-efficient vehicles is driving the green vehicle technology. Automotive OEMs and their suppliers are working on all fronts like design of more efficient engines, downsizing, light weighting, emission optimization, after treatment systems, aerodynamic design, electrical & electronics as well as software to make greener vehicles. 

Multiphysics simulation helps in designing reduced aerodynamic drag without sacrificing cooling and cabin quietness, reduced vehicle weight while still meeting strength and durability requirements, reduced rolling resistance in innovative ways, improved combustion efficiency of engines as well as in the deployment of effective exhaust after treatment systems.

Simulation also helps to shorten design iteration time, minimize hardware prototypes, and reduce the development time. Optimising the engine for emission and performance in the entire operating range is critical in engine development. Performing this using physical prototypes on an engine dyno or chassis dyno is time-consuming & expensive. Comprehensive emissions modelling and simulation tools help optimizing the engine and improve fuel efficiency. Innovative methods of building semi-physical models which combine physics-based models with statistical models are bound to further improve the optimization and speedup the development.

Use of simulation is pervasive in the automotive industry today. There are extensive applications of simulation beyond the ones discussed above. Electric motor design, analysis of material characteristics, optical simulation, interior lighting, acoustics & sound, human in the loop simulation are few such interesting use cases. 

Simulation enables virtualization which is transforming product development. Virtual prototypes are not just saving the development cost by reducing the number of prototypes but are helping in creating better designs, optimization, testing, verification, validation & maintenance of vehicles which otherwise would not be possible by using real physical prototypes. 

Power of simulation has no bounds; it is transforming all facets of product life cycle. Its role is pivotal in the ongoing automotive industry transformation. 

https://www.researchgate.net/publication/290738082_Simulation_Applications_in_the_Automotive_Industry 

https://www.who.int/publications-detail/global-status-report-on-road-safety-2018 

https://crashstats.nhtsa.dot.gov/Api/Public/ViewPublication/812384

https://www.curbed.com/2016/9/21/12991696/driverless-cars-safety-pros-cons

https://www.rand.org/content/dam/rand/pubs/research_reports/RR1400/RR1478/RAND_RR1478.pdf 

https://users.ece.cmu.edu/~koopman/lectures/Koopman19_SSS_slides.pdf 

https://www.fueleconomy.gov/feg/atv-ev.shtml 

Sameera Damle is currently working as Technical Account Manager with Ansys where he is responsible for the business with strategic enterprise customers. He comes with vast experience in the areas of automotive software development, simulation and validation of automotive controllers for Powertrain, BMS and Infotainment domains.