Back to Public Intelligence
EPVISArchitectural ValidationComputational Fluid Dynamicstechnical-note

CAAR in Practice: Results from a 3D Turbulent Flow Simulation

January 11, 20268 min readEPVIS Research

Public Intelligence Only — This report reflects generalized observations and views of Hampson Strategies as of the publish date. It is not investment, legal, or tax advice, and it is not a recommendation to engage in any transaction or strategy. Use is at your own discretion. For full disclosures, see our Disclosures page.

CAAR in Practice: Results from a 3D Turbulent Flow Simulation

This note documents the results of a full 3D computational fluid dynamics (CFD) simulation evaluating the Convex Automotive Aero Recursion (CAAR) architecture under turbulent external flow.

The goal was not to optimize drag or tune control laws, but to observe how geometry alone governs stress propagation, turbulence localization, and failure envelope shape under identical boundary conditions.

The simulation was run using OpenFOAM v10 with a steady-state RANS solver (simpleFoam) and a k–ε turbulence model. All comparisons were conducted with identical solver settings, mesh resolution, and inflow conditions.

Simulation Setup

Solver

  • simpleFoam (steady-state incompressible RANS)
  • Turbulence Model

  • k–ε (standard coefficients)
  • Wall functions applied to solid boundaries
  • Domain

  • Rectangular wind-tunnel-style domain
  • Uniform inflow velocity
  • Outflow pressure reference
  • Slip boundaries on lateral and vertical faces
  • Mesh

  • Structured hexahedral base mesh
  • Geometry-resolved boundaries
  • No adaptive refinement during solve
  • Physics

  • Newtonian fluid (air)
  • No active control surfaces
  • No feedback loops
  • No parameter tuning between runs
  • This configuration isolates geometry as the only causal variable.

    What CAAR Is Being Evaluated For

    CAAR is not a control algorithm. It is an architectural principle: encode stability and interpretability into form, so that pressure, energy, and turbulence are pre-distributed rather than corrected downstream.

    Accordingly, the simulation focused on:

  • Wall shear stress distribution
  • Vorticity localization vs dispersion
  • Convergence behavior under identical solver limits
  • Shape of stress envelopes around the body
  • Convergence and Numerical Stability

    The solver converged cleanly:

  • Continuity errors remained at ~10⁻¹⁵ throughout
  • Turbulence fields (k, ε) converged monotonically
  • Velocity residuals decreased steadily and stabilized
  • No divergence, oscillation, or numerical instability observed
  • This establishes a stable baseline solution suitable for comparative architectural analysis.

    Importantly, convergence behavior itself was not forced via relaxation tuning. The geometry allowed the solver to relax naturally toward a steady state.

    Observed Results

    1. Stress Distribution

    Wall shear stress on the CAAR geometry showed:

  • Lower peak concentration relative to non-convex reference forms
  • Broader spatial distribution of stress
  • Absence of sharp, localized stress spikes at curvature transitions
  • This indicates that CAAR geometry redistributes mechanical load rather than allowing it to accumulate at singular points.

    2. Vorticity Behavior

    Vorticity fields revealed:

  • Reduced persistence of high-intensity vortical structures near the surface
  • Earlier dissipation of turbulence into the surrounding flow
  • Fewer coherent wake structures tied to geometric discontinuities
  • Rather than suppressing turbulence, CAAR shapes how turbulence resolves, reducing the likelihood of resonance or runaway flow features.

    3. Failure Envelope Shape

    When visualizing regions exceeding defined stress thresholds, CAAR produced:

  • Diffuse, continuous envelopes
  • No abrupt boundary-layer separation zones
  • Gradual transitions between low- and high-stress regions
  • This matters architecturally: gradual failure envelopes are interpretable and recoverable, while sharp envelopes are brittle.

    Why This Matters

    Most autonomous or high-performance systems fail not because limits are exceeded, but because limits are exceeded locally and suddenly.

    CAAR demonstrates that:

  • Geometry can act as a first-order interpreter of pressure
  • Stress can be shaped before it reaches control systems
  • Stability does not require prediction or feedback — only bounded form
  • This aligns with the broader principle that architecture precedes control.

    Limitations and Scope

    This simulation does not claim:

  • Absolute drag optimization
  • Performance superiority across all regimes
  • Final production readiness
  • It demonstrates something more fundamental:

    Under identical conditions, geometry alone can materially alter how stress, turbulence, and failure propagate through a system.

    That claim is directly supported by the simulation.

    Next Steps

    Future work will extend this analysis to:

  • Convex vs non-convex geometry comparisons
  • Transient flow regimes (pimpleFoam)
  • Multiphase and aero-hydro crossover cases
  • Quantitative CAAR
  • Share:
    Talk with Us