SRS
Scale-Resolving Simulation at Scale
Benchmarking Turbulence Transition with the 3D Taylor–Green Vortex
- Aerospace
- Energy
- Automotive
What This Simulation Does
The 3d-taylor-green example is the canonical benchmark for transition to turbulence and turbulent decay. A smooth initial vortex breaks down into fine-scale turbulence; the decay of kinetic energy and growth of enstrophy test a solver's implicit-LES / DNS fidelity with a well-documented reference answer.
- Compressible Navier–Stokes on a triply periodic box; the energy cascade develops from a single smooth mode
- Vortex stretching & energy cascade — the physics that governs every turbulent industrial flow, in a controlled, reproducible setup
- Integral diagnostics — volume-integrated kinetic energy and enstrophy logged over time for direct comparison to DNS references
Key Parameters
- Grid: hexahedral elements, polynomial order 3, single precision; triply periodic box on [−π, π]3
- Fluid: ideal gas γ = 1.4, μ = 6.25 × 10-4, Pr = 0.71, reference pressure Ps = 111.607
- Scheme: Rusanov flux, LDG viscous; adaptive RK34 with PI controller
- Time: adaptive (Δtmax = 2 × 10-3), tend = 20; energy and enstrophy logged to integral.csv
The core transferable physics: A smooth flow transitions to turbulence and decays through the energy cascade. How accurately a solver reproduces kinetic-energy decay and peak enstrophy is the single best measure of its scale-resolving (LES/DNS) fidelity — the quality that determines whether a turbulent wake, jet, or combustor prediction can be trusted.
What Makes This Capability Unique
DNS-referenced fidelity
Kinetic-energy decay and enstrophy have well-documented DNS references, so scale-resolving accuracy is measured against a known answer.
Implicit LES without a model
High-order flux reconstruction acts as implicit LES — resolving the cascade with numerical dissipation tuned to the physics, no ad-hoc subgrid model.
GPU-native scale resolving
PyFR is built for streaming GPU architectures, making order-3 LES of a 3D turbulent field affordable in the cloud.
Solver qualification baseline
Provides an objective baseline for choosing mesh density and polynomial order before an expensive production LES campaign.
Domain Applications
Select a domain to see how this simulation applies, with industry-specific scenarios and ROI.
The Problem
Turbulent wakes, jet noise, and separated flows demand scale-resolving simulation, because RANS turbulence models systematically fail on massively separated and transitional flows. But LES is only trustworthy if the underlying solver reproduces the energy cascade correctly — which is exactly what this benchmark certifies.
- Jet-noise prediction depends on resolved fine-scale turbulence; an under-resolved solver mis-predicts the spectrum.
- High-lift and buffet at the edge of the envelope are transitional/separated flows where RANS is unreliable.
Aeroacoustic and high-lift redesigns driven by wrong turbulence predictions cost $10M+ per program.
Applications
| Application | How this simulation maps |
|---|---|
| LES solver qualification | Kinetic-energy decay and enstrophy vs. DNS certify the solver before a production LES campaign |
| Jet-noise source fidelity | Verifies the solver resolves the cascade that produces the acoustic source spectrum |
| Transition prediction | Benchmarks the solver's ability to capture laminar-to-turbulent breakdown |
| Mesh/order selection | Sets the minimum resolution needed for a target turbulence fidelity |
Quantifiable Business Value
Scenario: An aeroacoustics group qualifies its LES resolution on the Taylor–Green benchmark, avoiding an under-resolved production run that would have produced a wrong jet-noise spectrum and a needless nozzle redesign.
A qualified LES baseline is reusable across every future aeroacoustic and separated-flow study.
Recommended Next Steps
Compare against DNS references
Run the case and compare the kinetic-energy decay and peak enstrophy against published DNS to certify your solver's LES fidelity.
Study resolution sensitivity
Vary mesh density and polynomial order to find the minimum resolution that meets your turbulence-fidelity target.
Scale to your turbulent flow
Carry the qualified settings into a production LES of your wake, jet, combustor, or cooling flow.
Ready to Run This Simulation?
Run this example on SRS's cloud platform. No installation, no infrastructure management — just results.
For questions or to schedule a technical briefing, contact the SRS simulation team.
Scale-Resolving Simulation (SRS)
© 2026 SRS, a NumericalAI product by Empirisch Tech GmbH (empirischtech.at). All rights reserved. |Privacy Policy |Terms of Service |Executive brief |FAQ