One of the well-known challenges in electric vertical takeoff and landing (eVTOL) aircraft design is solving aerodynamic problems that do not have decades of established and validated flight data to draw from. Rotor–airframe interactions, highly integrated lifting surfaces, and unsteady flow effects introduce uncertainty that is hard to quantify early in the design process.
However, a less-discussed challenge is proving, under extreme program timelines, which design ideas are actually worth building. In certification-driven programs, teams cannot afford to experimentally explore the full design space, yet relying too heavily on intuition or a limited number of simulations carries real risk. When each configuration takes days to analyze, relying on a limited number of simulations carries real risks:Compromised Trade Studies: Analyses collapse into single-point evaluations rather than comprehensive design explorations.
Premature Design Freezes: Engineers are forced to commit to a design based on incomplete data.
Certification Gambles: Blind spots in the aerodynamics can lead to costly redesigns during CAA (Civil Aviation Authority) or EASA (European Union Aviation Safety Agency) certification.
Ultimately, this tension between aerodynamic uncertainty and schedule pressure is where high-fidelity computational fluid dynamics (CFD) either becomes a crippling bottleneck or a massive competitive advantage.
Vertical Aerospace Valo eVTOL Aircraft
This is exactly the tension Vertical Aerospace, a global aerospace and technology company pioneering electric aviation, encountered while designing their Valo eVTOL aircraft. Tasked with delivering a category-leading 100-mile range and a near-silent noise signature, the team had to balance aggressive program timelines against the need to make high-confidence aerodynamic trade-offs.
As with any eVTOL development program, simulation-driven design ultimately transitions into physical validation through wind tunnel testing. This critical milestone places a higher bar on predictive fidelity, particularly for configurations where unsteady, interaction-driven flow physics dictate real-world behavior.
During this phase, Flow360’s results were benchmarked directly against experimental data. The physical testing yielded several key validations:
Excellent Agreement: The simulations matched the experimental data well within the required error margins.
Post-Stall Accuracy: Flow360 successfully captured the lift coefficient even in the post-stall regime—a notoriously difficult aerodynamic condition to model accurately.
Workflow Confidence: Vertical Aerospace confirmed that the exact same high-throughput CFD workflow used for rapid design exploration could be fully trusted when evaluating configurations against physical measurements.
This validation phase further separated Flow360 from the field, establishing Flow360 as both a design exploration engine and a credible predictor of real-world aerodynamic behavior.
In practice, eVTOL aerodynamic design requires a staged approach to balance rapid exploration with rigorous accuracy:
Broad Screening: Rapidly evaluating a wide set of design concepts using faster, lower-fidelity analyses to identify promising candidates.
High-Fidelity Validation: Assessing the shortlisted configurations using high-fidelity simulations—including unsteady cases—to verify behavior under realistic, interaction-driven flow conditions.
However, this strategy only works if both phases can be executed within strict program timelines. The workflow initially in place at Vertical Aerospace severely constrained how aggressively the team could iterate. Built around a legacy commercial solver with a decades-old codebase, the architecture lacked the key capabilities and scalability required for modern, high-rate eVTOL product development.
This legacy system introduced critical bottlenecks into the design cycle:
Costly Two-Week Iterations: Each design concept took on the order of two weeks to analyze end-to-end, forcing premature down-selection simply to stay on schedule.
Sluggish Unsteady Analysis: Once a design direction was chosen, running higher-fidelity, unsteady cases took roughly five days per configuration.
Delayed Aerodynamic Insights: Because of these slow turnaround times, crucial high-fidelity effects entered the decision process too late—often after key design directions had already been locked in.
At this point, it was clear that the existing workflow was actively limiting how thoroughly the design space could be explored under real program timelines. The team needed a way to dramatically increase analysis throughput so that trade studies could cover more concepts early, and incorporate high-fidelity, unsteady behavior at exactly the right moment in the process, without sacrificing accuracy.
This realization, combined with proven validation of its accuracy, speed, and robustness, drove Vertical Aerospace's decision to fully transition their workflow to Flexcompute Flow360.
More than just a solver swap, Vertical Aerospace modernized their entire aerodynamic process. This fundamentally changed the role CFD played, turning simulation from a potential bottleneck into a powerful mechanism for expanding the design space under real program timelines.
To achieve this, the workflow was completely rebuilt for automation, scale, and accuracy:
Automated Execution: Simulation setup and execution were streamlined through Flow360’s Python API.
Efficient Meshing: Automated meshing handled highly complex eVTOL geometries with minimal manual intervention.
Advanced Rotor Modeling: Rotor models, including blade element theory (BET) disk approaches, were seamlessly integrated.
GPU-Native Performance: Leveraging Flow360's architecture enabled large-scale trade studies to run with unprecedented efficiency.
Visualization of flow around Vertical Aerospace's aircraft using Flow360
Combined, these upgrades removed long-standing legacy bottlenecks and made broad design sweeps a practical part of day-to-day engineering work. The impact on productivity was immediate:
10X Design Throughput: The team successfully evaluated ten times as many design variations within the exact same timeframe.
Accelerated Configuration Analysis: The turnaround time to evaluate a single configuration was reduced from two weeks to just one day.
Rapid Unsteady Analysis: Unsteady simulations involving complex propeller-airframe interactions in takeoff and transition flight conditions, which previously took up to five days, could now be completed in one to two hours.
This exponential leap in speed made it practical to routinely incorporate complex, unsteady physics directly into trade studies, drastically improving the fidelity of the data used to guide design down-selection.
With absolute confidence in the final design direction, the program could successfully shift its focus from exploration speed to predictive accuracy.
The impact of the modernized CFD workflow through Flow360 extended across the full design cycle for Vertical Aerospace. This transformation delivered critical value at every stage:
Broader Early Exploration: The team could realistically evaluate a massive design space during initial concept trade studies.
Confident Down-Selection: High-fidelity, unsteady data allowed engineers to lock in key design directions with absolute certainty.
Sustained Predictive Accuracy: Trust in the digital models was maintained seamlessly through physical wind tunnel validation.
More broadly, this experience highlights why CFD platform choice matters. Selecting a solver is not just a procurement decision around licensing or familiarity. It shapes how quickly teams can iterate, how much of the design space they can realistically explore, and how early high-confidence decisions can be made. In programs where early choices compound downstream, iteration velocity becomes a strategic lever, not just an efficiency metric.
Partnering with Flexcompute enabled Vertical Aerospace with faster exploration without sacrificing fidelity, giving the team greater confidence as they progressed from concept trade studies to physical validation. In that sense, the CFD platform became part of how the program advanced, not just a tool used along the way.