Competitive pressures and rising customer expectations are forcing the pace of innovation in electric vehicle (EV) design. Early market entrants now face many more challengers. Established automakers are responding with their own ranges. And start-ups are rethinking how people use personal transportation and building new forms of EV to suit. European market statistics reflect this trend: In 2015, battery and plug-in hybrid EVs accounted for just over 1% of new vehicle registrations, according to the European Alternative Fuels Observatory. By 2022, this figure was just over 21%.
EV designers are under pressure to increase vehicle efficiency because this directly affects key customer care-abouts such as vehicle range. No matter that many personal cars are mostly used for short journeys – ‘range anxiety’ puts some buyers off. EV makers also have internal concerns about issues such as access to scarce resources and the impact of regulation, which are driving demand for more efficient designs.
The fascinating thing about the EV market is that we can watch its very rapid evolution in real time. We’re already seeing EV designs that seem to have been optimised for a single figure of merit. These include huge pickup trucks, from companies such as Ford and new entrant Rivian, that can tow tonnes, sedans from Lucid with claimed ranges of more than 500 miles, and hyper-cars, such as the Rimac Nevera, that can accelerate from 0 – 60mph in under two seconds.
Mercedes-Benz has gone a step further with its Vision EQXX concept car, combining many of its most advanced technologies to create what it says is “the most efficient car Mercedes-Benz has ever built.” The car has a claimed range of 1000km (621 miles) thanks to very high aerodynamic efficiency, lightweight materials, a high energy-density battery, a low-friction powertrain, and other enhancements.
Efficiencies left on the table
The Vision EQXX is a nice example of what is possible if an EV vendor brings together the best available technologies. There are plenty of other examples of this kind of co-optimisation – if designers can access it.
For example, increasing the DC operating voltage of an EV’s battery pack will reduce the current that has to be carried in the powertrain, creating an opportunity to move from a copper wiring loom to a lighter, less costly aluminium alternative. The weight advantage can then be used to increase the vehicle’s range or traded off to enable the use of a smaller, lighter battery. Taking advantage of this potential efficiency gain depends on being able to source switching electronics and design PCBs that work reliably in the electric fields created by the higher operating voltage.
To extend the thought experiment, silicon carbide semiconductors can meet these higher voltage switching needs, but in turn, designers will have to validate the systemic impact of their higher switching frequencies on electromagnetic compatibility (EMC) issues with control electronics. Likewise, changes to bodywork to improve aerodynamics can affect the operation of road-sensing radar systems mounted in vehicle grilles. And so on and so on.
These are multivariate optimisation problems that present a real challenge to EV designers. If they can’t do rapid, accurate design exploration across multiple domains, they risk leaving all sorts of optimisations and subsequent efficiency gains on the table. But many existing tools and flows don’t help – they tend to involve designing in one domain, dumping the design data out to a database, ingesting that design into another tool, and then checking that it doesn’t violate any constraints of that domain. Co-optimisation, in which EV designers do more than just check that they are avoiding design violations, may be driven out by time-to-market pressures and the costs of ‘flip-flopping’ between domains during design exploration.
The solution to this issue is twofold. For single-discipline design issues, it is important to shift the analysis process earlier in the flow – a ’shift left’ – so that it becomes a continuing part of the design process rather than a check imposed at its end. For more complex, systemic issues, vendors need to offer tools that enable multiple aspects of a design to be explored holistically. For example, it should be possible to check signal-integrity issues during PCB layout, explore the performance impact of different battery cooling strategies, or analyse the effect of a change of material on the power-delivery characteristics of a wiring loom.
Vendors serving the EV market already undertaking this kind of design-and-analysis process to enable them to check their designs will have the reliability necessary. For example, STMicroelectronics wanted to show that its parts could be used in a stringent high-power application on a complex PCB reference design by understanding how the devices’ reliability would change with operating temperature. This led to a requirement to do electrothermal co-simulations of the design to estimate the effect of joule heating on the IR drop within it.
Cadence was able to offer an integration of its Celsius Thermal Solver with its Allegro PCB Designer platform to support the in-design analysis necessary to get the job done in a timely manner. The tools handled the design as one entity, rather than having to break it down into smaller pieces to accommodate the limitations of other solutions.
Cadence has more complex flows to tackle more complex issues. For example, its power-integrity analysis flow can handle issues such as power trees, IR drop analysis, and the loop inductance of decoupling capacitors (decaps); IC package analysis in terms of DC electrothermal performance and AC decap behaviour; multi-fabric analysis from source to PCB to interposer to thermal sink; and even system-level thermal analysis, including electronic cooling.
These tools all fit within a much wider Cadence offering. It includes system design platforms for analogue and custom IC design; RF and microwave design; and PCB and packages. These are matched with system analysis technologies for EMC signoff, signal and power-integrity analysis, and thermal and electronic cooling checks. Cadence also has a fluid-flow analysis package that is particularly relevant to automotive body design and the development of efficient electronics cooling systems.
Taking inefficiency off the table
EV development is a complex, multidisciplinary process that is being carried out under intense market pressure to achieve multiple forms of efficiency – the range figures that will attract consumers as well as the other efficiencies that EV makers need to thrive. Tools that can make analysis available earlier in the design process enable more effective design space exploration, ensuring design teams begin on the right path. Tools that enable co-optimisation of multiple aspects of designs can help achieve more efficient designs and do so more quickly than has previously been possible.
Cadence has a broad suite of point tools relevant to EV design, as well as an evolving portfolio of integrations between tools that shift analysis left within the design process and enable co-optimisation. Together, they give EV designers the power to pick up all sorts of efficiencies that they would otherwise have to leave on the table.
Author details: Ben Gu, corporate VP, Multiphysics System Analysis Group at Cadence