The research was led by Emmanuel Gillyns of the von Karman Institute for Fluid Dynamics and ENGIE Laborelec, together with Sophia Buckingham, Jeroen van Beeck and Grégoire Winckelmans of the von Karman Institute and UCLouvain. Their starting point was a problem that has long frustrated wind energy engineers: the air that flows off a turbine blade, known as the wake, governs how much wind reaches the turbines downstream, how heavily their structures are loaded, and ultimately how much electricity a farm produces. Getting the wake wrong in a simulation means getting the economics of wind power wrong.

To attack the problem, the team first designed and built a brand-new model-scale wind turbine, christened TWIST, short for Turbine for Wind-tunnel Investigation and Scaled Testing. The machine has a rotor diameter of 0.80 metres and a hub height of 0.61 metres, small enough to fit in a laboratory yet large enough to avoid the misleading physics that plagues tinier desktop turbines. Its blades were 3D-printed at UCLouvain from a ductile PLA+ plastic, using the SD7037-092-88 airfoil profile, a shape specifically chosen because it performs well at the low Reynolds numbers that small-scale models inevitably experience.

Scaling down a wind turbine is not simply a matter of shrinking it. The researchers had to preserve the key dimensionless numbers that control wake behaviour, including the tip speed ratio, which sets how fast the blade tips sweep through the air, and the thrust and power coefficients, which determine the velocity deficit and the energy extracted from the flow. The TWIST turbine operates at a rotor Reynolds number above 35 times ten to the fourth, a threshold above which wake statistics become largely independent of scale, meaning the miniature wake genuinely mimics what a full-size machine would produce. Blockage effects from the tunnel walls were carefully minimised, with the rotor occupying only about 8.4 percent of the test section cross-section.

The experiments took place in the L1-B wind tunnel at the von Karman Institute, a closed-loop facility with a test section two metres high, three metres wide and roughly twenty metres long. Crucially, the team did not test their turbine in smooth, artificial air. Instead, they generated a turbulent boundary layer on the tunnel floor using roughness elements, a grid and a fence, replicating the atmospheric conditions that real turbines face in the field. Hot wire anemometers then measured wind speeds along vertical lines at three locations, sampling at 20 kilohertz for five minutes per point, with uncertainties meticulously quantified and propagated through every calibration step.

On the computational side, the team ran large eddy simulations using Nek5000, an open-source, high-order spectral element code developed at Argonne National Laboratory. Large eddy simulation resolves the dominant turbulent structures directly, but modelling the full geometry of rotating blades remains computationally prohibitive, so researchers rely on the actuator line method, which represents the blades as lines of force rather than solid surfaces. The catch is that the method’s accuracy depends heavily on its implementation, and conventional versions smear out the physics near the blade tip and root, where forces change most abruptly.

Here lies the study’s central innovation: a continuous actuator line method, or cALM, that represents blade forces as polynomials on the simulation mesh nodes rather than as discrete point sources. The approach avoids spanwise averaging along the blade, applies its Gaussian force spreading only in the plane perpendicular to the blade, and samples the wind velocity independently at every grid point. This not only improves the physical fidelity of the tip vortex, the spiral of fast-spinning air shed from each blade end, but also slashes computational overhead by keeping every node independent and perfectly parallelisable. The same actuator principle was extended to the nacelle and tower, with the nacelle treated as a bluff body with an empirically tuned drag coefficient.

The team also invested enormous effort in making the simulated inflow match the wind tunnel conditions, using a modified recycling and rescaling technique in a co-simulation of two computational domains, plus a cleverly designed forcing term built from thirty closely spaced sine waves to boost turbulence where the experiment demanded it. The result was striking: the base flow error between simulation and experiment fell to a root mean square error of just 0.11 metres per second, with a mean absolute percentage error of 1.55 percent. That level of agreement means any remaining differences when the turbine is switched on can be attributed to the turbine model itself rather than to mismatches in the incoming wind.

When the results were compared, the cALM captured the essential features of the near-wake at 1.41 rotor diameters downstream, including the velocity deficit and the wake boundary, and remained consistent with measurements further out at 4.35 diameters. Most impressively, the simulation reproduced the sharp velocity gradient between the free stream and the wake, a zone where tip vortices dominate. With tip spreading disabled and a finely refined mesh, the simulated vertical shear in the transition zone came far closer to the experimental value than conventional smoothing would allow. Deviations persisted near the blade root, where the geometry transitions from airfoil to ellipse to circle and the flow becomes genuinely three-dimensional, stretching the assumptions of any actuator line model beyond its limits.

The study also probed how sensitive the wake is to blade pitch angle, running simulations at plus one and minus three degrees around the experimental operating condition, and examined how the flexible printed blades actually deform under load. Camera measurements revealed mean tip deflections of 7.5 millimetres at the top of the rotation and 6.2 millimetres at the bottom, more than the 2.9 millimetres predicted by a structural model that assumed a solid blade, yet still only about two percent of the blade span, too small to drive the observed wake discrepancies. The bigger picture is one of growing confidence: by pairing a purpose-built experimental turbine with a sharper, faster simulation method, the researchers have delivered both a reference dataset for the community and a demonstration that wind farm performance can be predicted with greater physical grounding, a quiet but meaningful step toward squeezing more clean energy from the same patch of sky.

Subject of Research: Experimental and numerical validation of wind turbine wake modeling using a model-scale turbine and large eddy simulation with a continuous actuator line method

Article Title: Investigating wake reproduction of a model-scale wind turbine: experimental measurements versus large eddy simulation with actuator line

Article References: Gillyns, E., Buckingham, S., van Beeck, J., & Winckelmans, G. (2026). Investigating wake reproduction of a model-scale wind turbine: experimental measurements versus large eddy simulation with actuator line. Wind Energy Science, 11(9), 3531-3553. https://doi.org/10.5194/wes-11-3531-2026

Image Credits: AI Generated

DOI: 10.5194/wes-11-3531-2026

Keywords: wind energy, wind turbine wake, large eddy simulation, actuator line method, wind tunnel experiments, Nek5000, atmospheric boundary layer, tip vortex, blade deformation, turbulence modeling, renewable energy, TWIST turbine

The core problem the team set out to solve is generalization. In recent years, data-driven turbulence modeling has flourished: researchers train neural networks on high-fidelity simulation or experimental data, and the networks learn corrections to classical models such as the k-epsilon or Spalart-Allmaras closures. These learned models can dramatically improve accuracy—but usually only within the flow regime they were trained on. The reason is architectural. Most existing approaches use dense, monolithic networks in which every parameter participates in every prediction. Such networks tend to entangle the physics of different regimes into a single set of weights, so a model tuned to attached boundary layers may produce nonsense when asked about separated flow, and vice versa. The result is a patchwork of specialized models that engineers must manually select and swap as conditions change—an approach that is fragile, labor-intensive, and poorly suited to real-world applications where flows routinely traverse multiple regimes.

Liu and colleagues’ answer is a framework they call the factorized-gating mixture of experts, or FG-MoE. Mixture-of-experts architectures are not new to machine learning; they have powered large language models and other large-scale systems by dividing computation among specialized subnetworks, with a component called a router or gating network deciding which expert handles each input. What distinguishes FG-MoE is how it builds and controls its experts. Instead of one opaque router making high-dimensional decisions, the framework trains a set of regime-specific experts—each specializing in a distinct physical regime of turbulence—and blends their outputs using spatially varying probabilities. At every point in a flow field, the gating mechanism estimates how strongly each expert should contribute, producing a smooth, location-dependent combination rather than a hard switch between models.

The truly novel ingredient is the factorization of the gating function itself. Conventional routers are high-dimensional: they take the full set of local flow features and map them onto expert-selection probabilities through a dense layer, which is computationally expensive and notoriously difficult to interpret. FG-MoE replaces this monolithic router with a set of lightweight, physics-guided binary attributes—simple yes-or-no questions about the local flow state, such as whether a particular regime indicator is active. The overall gating probability is then approximated as a product of low-dimensional binary factor probabilities. This product-form decomposition slashes the complexity of the gating computation, because the network no longer has to learn an arbitrary mapping through a vast parameter space. Instead, it composes a handful of interpretable, physically motivated decisions into a final expert weighting.

Interpretability is not merely an aesthetic bonus here; it is central to why the approach matters for engineering. When a conventional neural network corrects a turbulence model, engineers often cannot tell why it made a particular prediction, which breeds distrust in safety-critical applications like aircraft certification or nuclear reactor design. With FG-MoE, the binary factors can be tied to recognizable flow physics, so the gating mechanism effectively explains itself: a given point in the flow is assigned to an expert because specific, legible conditions hold there. The authors argue that this transparency, combined with the reduced gating complexity, makes the framework a practical pathway for engineering turbulence modeling—one that decomposes complex turbulent flows into distinct physical regimes and then leverages the factorized gating strategy to keep the whole system interpretable, efficient, and scalable.

Scalability arrives in an unusually elegant form. In most machine learning systems, adding new capability means retraining the entire model, an expensive and disruptive process. FG-MoE is designed as a plug-and-play architecture: new experts can be added to the expert library without retraining the whole model. If engineers encounter a flow regime the current system handles poorly—say, a novel heat-transfer configuration in a gas turbine—they can train a dedicated expert for that regime and slot it into the existing framework, with the factorized gating mechanism automatically learning when to call on the newcomer. This modularity mirrors how human engineering knowledge actually accumulates: not by rewriting every textbook each time a new phenomenon is discovered, but by adding a chapter and cross-referencing it with the rest.

The evidence for the framework’s effectiveness comes from extensive numerical experiments spanning canonical benchmarks to complex, real-world engineering applications. Canonical cases—the well-instrumented standard flows that the turbulence community uses to judge models—allow head-to-head comparison against established baselines. The engineering applications then test whether the approach survives contact with the messy, multi-regime reality of industrial flows, where separation, reattachment, pressure gradients, and wall effects coexist in a single domain. Across this spectrum, the authors report that FG-MoE delivers consistently high-fidelity predictions across diverse flow regimes, in contrast to monolithic dense networks whose accuracy degrades sharply outside their training distribution. The consistency is the headline result: rather than excelling in one regime and failing in others, the expert-based system maintains its accuracy as the physics changes.

The implications reach across aerospace, energy, and environmental technology, the three domains the authors single out as being critically limited by turbulence model accuracy. In aerospace, better cross-regime turbulence prediction could sharpen estimates of drag, heat loads, and buffet onset, reducing the safety margins that currently inflate fuel consumption and design cost. In energy, gas turbines, wind farms, and fusion devices all involve flows that transition abruptly between regimes, and improved models could improve efficiency and reliability. In environmental applications, from urban wind comfort to pollutant dispersion, the ability to blend regime-specific knowledge smoothly across a domain could make simulations more trustworthy. Because the framework is modular and interpretable, it also offers a natural vehicle for incorporating new experimental or simulation data as it becomes available, without discarding the accumulated expertise encoded in existing experts.

There are, of course, caveats worth keeping in mind. The study is computational, resting on numerical experiments rather than new wind-tunnel measurements, and the quality of any data-driven turbulence model ultimately depends on the fidelity and breadth of its training data. The factorized gating relies on physics-guided binary attributes, which means someone must choose sensible regime indicators—a task that leans on domain expertise even as it reduces the burden on the learning algorithm. And while plug-and-play expansion avoids full retraining, each new expert still requires its own training data and validation. The published version of the paper is also an early-release, peer-reviewed accepted manuscript that the journal notes remains subject to further edits before the final Version of Record, so some details may be refined.

Even with those qualifications, the work represents a notable conceptual shift. For twenty years, data-driven turbulence modeling has largely pursued bigger, denser networks trained on ever-larger datasets, in the hope that raw capacity would yield universal closures. Liu, Luo, Zhang, and He make the opposite bet: that turbulence is best learned by dividing it—by training specialists, wiring them together through a gating mechanism that is deliberately simple, physically grounded, and mathematically factorized, and by making the whole assembly extensible. If the framework’s cross-regime performance holds up as it is tested against more industrial flows and, eventually, experimental data, the humble turbulence model may finally be on its way to becoming something engineers have long wanted: a system that knows which kind of chaos it is looking at, and admits it.

Subject of Research: Machine learning framework for cross-regime turbulence modeling using a factorized-gating mixture of experts

Article Title: Factorized-gating mixture of experts enables cross-regime turbulence modeling

Article References: Liu, H.-C., Luo, Q., Zhang, X.-L., & He, G. (2026). Factorized-gating mixture of experts enables cross-regime turbulence modeling. Communications Engineering. https://doi.org/10.1038/s44172-026-00802-5

Image Credits: AI Generated

DOI: 10.1038/s44172-026-00802-5

Keywords: turbulence modeling, mixture of experts, machine learning, fluid dynamics, aerospace engineering, neural networks, gating mechanism, data-driven modeling, flow regimes, interpretability, computational fluid dynamics, Chinese Academy of Sciences

The research, led by Xiaolong Zhang of Nanjing University of Aeronautics and Astronautics together with colleagues including Hao Wang of Rongtong Aeroengine Technology and corresponding author Ning Ge, tackles a problem that has challenged engineers since the earliest scramjet experiments of the 1990s. When a shock wave strikes a boundary layer—the thin layer of slowed air hugging a surface—it can thicken that layer dramatically, or even force it to separate from the wall entirely. For film cooling, this is a double-edged sword: a thicker boundary layer might seem to offer more insulation, but separation bubbles create recirculating hot gas that can strip away the coolant film and expose the wall to searing mainstream temperatures. Understanding precisely where and how this happens is essential for designing cooling schemes that survive in hypersonic propulsion systems.

To capture these interactions, the team employed a hybrid turbulence modeling strategy called the Stress-Blended Eddy Simulation, or SBES, coupled with a fifth-order weighted essentially non-oscillatory scheme known as WENO-ZQ. The approach is a compromise between computational necessity and physical fidelity. Full large-eddy simulation, which resolves nearly all turbulent eddies, is extraordinarily expensive for the complex geometries and high Reynolds numbers of engine-relevant flows. Standard Reynolds-averaged methods, meanwhile, average away the very unsteady structures—vortex rings, shear-layer instabilities, separation bubbles—that govern film cooling performance. SBES blends the two: it runs in RANS mode near the wall where turbulence is dominated by small-scale stresses, and switches smoothly to LES mode in the outer regions where large, energy-containing eddies matter. The fifth-order WENO-ZQ scheme, developed originally by Zhu and Qiu, provides the shock-capturing sharpness needed to resolve discontinuities without spurious oscillations.

The validation of this numerical toolkit was itself a significant contribution. The researchers demonstrated that the SST–SBES combination could accurately reproduce the vortex structures and their temporal evolution within the film cooling domain—structures such as the counter-rotating vortex pair that lifts coolant off the wall, the horseshoe vortices that wrap around the jet as it emerges, and the hairpin vortices that trail downstream. These vortical structures are the machinery of mixing: they determine whether the cool air stays pressed against the wall where it protects the surface, or gets churned into the hot mainstream where it is wasted. A turbulence model that cannot capture them cannot be trusted to predict cooling effectiveness, which is precisely why the team invested effort in establishing the method’s resolution and accuracy before drawing physical conclusions.

The geometry at the heart of the study is the 7-7-7 shaped hole, a design whose cross-section expands laterally and vertically in a stepped fashion, resembling a dustpan laid into the surface. Shaped holes of this kind were developed to combat the fundamental weakness of cylindrical holes: the strong counter-rotating vortex pair that lifts the coolant jet away from the wall. By diffusing the jet exit, the shaped hole reduces the jet’s vertical momentum and spreads the coolant laterally, improving coverage. But shaped holes were largely characterized in subsonic conditions. What happens when a shock wave impinges upstream of such a hole, in a supersonic mainstream, remained poorly quantified—especially at different blowing ratios, the ratio of coolant momentum to mainstream momentum.

The team simulated a supersonic flat-plate configuration with the dustpan-shaped hole, examining blowing ratios of 0.5 and 0.65. The first major finding concerns the shock–boundary layer interaction itself. When the shock strikes the incoming boundary layer upstream of the hole, it thickens the layer and can induce a separation bubble—a pocket of recirculating flow detached from the wall. Interestingly, the blowing ratio, which changes the character of the jet emerging from the hole, had only a limited effect on the mainstream pressure distribution. In other words, the shock system upstream behaves largely as if the cooling jet were not there, decoupling the external aerodynamics from the internal cooling problem in a way that simplifies how engineers can think about the two phenomena.

The second key finding is more subtle and concerns the region near the hole’s trailing edge. At the lower blowing ratio of 0.5, the jet does not fill the expanded dustpan exit uniformly. Instead, aerodynamic bulges form within the jet—localized protrusions of high-speed coolant that act like small compression surfaces, generating compression waves of their own. These waves, the study found, intensify localized boundary layer thickening near the trailing edge of the hole, compounding the disturbance already created by the upstream shock. At the higher blowing ratio of 0.65, however, the jet has enough momentum to fill the expanded exit more completely, smoothing out the bulges and alleviating this additional thickening. The practical implication is striking: simply increasing the coolant supply can suppress a secondary shock-generating mechanism, though at the cost of more compressed air diverted from the compressor.

Perhaps the most elegant results come from spectral analysis—the turbulence equivalent of listening to the flow’s frequency signature. The researchers placed monitoring points upstream, midstream, and downstream of the interaction and examined the power spectra of the pressure fluctuations. Upstream and at midstream locations, the spectra displayed a distinctive −7/3 power-law scaling, a slope steeper than the classical Kolmogorov −5/3 law of inertial-range turbulence. The team attributes this anomalous scaling to two mechanisms working together: the compression imposed by the shock wave, which compresses turbulent eddies and steepens their spectra, and the shear-layer instability driven by the interaction between the jet and the mainstream. The −7/3 slope has precedent in pressure spectra of turbulent free shear flows, and its appearance here ties the shock-perturbed cooling flow into a broader family of compressible turbulent phenomena.

Downstream, however, the story changes. The spectra at far-field monitoring points transition toward the classical −5/3 inertial subrange slope, indicating that the violent, shock-dominated disturbances have decayed and the flow has relaxed into a state where turbulent mixing between the jet and the mainstream has saturated. In physical terms, the shock’s fingerprint fades with distance, and the jet-mainstream system settles into ordinary turbulent mixing. This spectral evolution gives engineers a diagnostic tool: by measuring pressure fluctuations at a surface, one can infer how far downstream the shock’s influence on the cooling film persists, and where the protective layer has either recovered or been fully consumed by mixing.

The study, funded by China’s National Science and Technology Major Project, arrives at a moment when hypersonic flight programs worldwide are pushing film cooling technology into regimes where shock interactions are unavoidable rather than exceptional. In a scramjet combustor, shock trains and expansion fans are everywhere, and cooling holes sit directly in their path. The finding that higher blowing ratios can suppress jet-induced compression waves offers a concrete design lever, while the demonstrated fidelity of the SBES–WENO-ZQ framework provides a validated virtual testbed for exploring hole geometries, spacing, and coolant conditions without costly wind-tunnel campaigns. As shaped holes migrate from subsonic turbines into supersonic combustors, studies like this one are building the quantitative foundation that turns a fragile cool blanket into engineering that can be trusted at Mach 5 and beyond.

Subject of Research: Shock wave interaction with shaped-hole film cooling in supersonic boundary layers

Article Title: SBES-Based Study on Aerodynamic Impacts of Upstream Shock Waves on Film Cooling from Dustpan-Shaped Holes

Article References: Zhang, X., Wang, H., Wu, X., & Ge, N. (2026). SBES-Based Study on Aerodynamic Impacts of Upstream Shock Waves on Film Cooling from Dustpan-Shaped Holes. International Journal of Aeronautical and Space Sciences. https://doi.org/10.1007/s42405-026-01272-x

Image Credits: AI Generated

DOI: 10.1007/s42405-026-01272-x

Keywords: film cooling, shock wave, SBES, dustpan-shaped hole, supersonic flow, boundary layer separation, turbulence modeling, WENO scheme, spectral analysis, scramjet, gas turbine, aerodynamics

The research, carried out by Baptiste Moreau, Alfonso Caiazzo, Thomas Amon, Long Chen and David Janke, set out to evaluate an open-source lattice Boltzmann solver called OpenLB against the established finite volume solver OpenFOAM. Conventional computational fluid dynamics tools used in agriculture solve the Navier–Stokes equations with finite volume or finite element methods, approaches that have been extensively validated over decades and are embedded in mature commercial and open-source packages. Their weakness is computational cost: as turbulence develops and domains grow, simulations can run for hours or days, which makes them poorly suited to applications where predictions are needed quickly, such as interpreting data streaming from barn-integrated sensor networks.

The lattice Boltzmann method takes a fundamentally different route to the same physics. Instead of discretizing the Navier–Stokes equations directly, it evolves a pseudo-compressible formulation on a fixed Cartesian lattice, tracking population densities that represent particles moving along a discrete set of velocities, commonly nineteen directions in three dimensions, the so-called D3Q19 model. Each time step consists of a collision step, in which the densities relax toward an equilibrium distribution described by the Bhatnagar–Gross–Krook operator, and a propagation step, in which information moves to neighboring nodes. The crucial property is locality: the collision step, which carries most of the computational burden, depends only on values at a single lattice node. Combined with a structured grid and explicit time-stepping, this makes the method exceptionally well suited to massive parallelization on modern hardware, including graphics processing units.

To test whether this promise holds for agricultural flows, the team built a benchmark that others can easily reproduce. They used a 1:100 scaled model of a typical agricultural building, 0.25 meters wide and 0.13 meters high, mounted perpendicular to the flow in the atmospheric boundary layer wind tunnel at the ATB Potsdam. The model, originally part of a dairy farm study, was elevated on a smooth table to minimize the influence of the tunnel’s thick boundary layer, producing an inflow with a nearly constant vertical velocity and turbulence profile. With the barn height as the characteristic length and an inlet velocity of 13 meters per second, the flow reached a Reynolds number of roughly 114,000, firmly in the turbulent regime. Velocity measurements were collected along fourteen vertical sample lines upstream and downstream of the model.

On the computational side, the barn geometry was supplied as an STL file and embedded in a domain about three meters long, 0.98 meters high and two meters wide. OpenLB generates a uniform lattice whose spacing is set by a refinement parameter, which the researchers varied from coarse grids of about 115,000 cells to finely resolved lattices exceeding 50 million cells. No-slip conditions on the ground and barn surfaces were enforced with a second-order interpolation scheme due to Bouzidi, a uniform inlet velocity was ramped up smoothly to suppress nonphysical oscillations, and slip conditions closed the top and sides. Turbulence was handled with the Smagorinsky large-eddy subgrid model, implemented in the lattice Boltzmann framework by adding an eddy viscosity to the relaxation time, with a Smagorinsky constant of 0.15 chosen after a preliminary sensitivity study.

The accuracy assessment compared time-averaged velocity components against the wind tunnel data using a battery of error metrics, including mean relative error, mean absolute error, root mean square error and a symmetric mean absolute percentage error. For the streamwise velocity, the simulation reproduced the qualitative trends of the experimental profiles along essentially all sampling lines, and about half of all measurement points showed relative errors below 13 percent. In regions dominated by the interaction with the building, such as lines immediately beside and behind the barn, agreement was partly very good, with errors below 5 percent. Larger deviations appeared for the vertical velocity component just upstream of the model, where the solver underestimated the measurements, possibly because a single constant Smagorinsky coefficient overestimates turbulent viscosity in the nearly turbulence-free inflow. The authors note that dynamic subgrid models and local grid refinement could improve these regions, and they emphasize that the achieved accuracy is consistent with previous validation studies of open-source solvers against comparable wind tunnel data.

The efficiency results are where the study becomes striking. Measured in million lattice updates per second, the CPU-based parallelization sustained roughly 100 MLUPs on refined meshes with 20 cores, at the upper end of values reported in the literature, and showed strong scaling with per-processor throughput declining only from 6.2 to 5.0 as core counts grew. Against OpenFOAM on the same grid, the lattice Boltzmann solver was about 23 times faster when both used 40 CPU cores, delivering a time ratio of 27 compared with 620 for the finite volume solver. The finite volume code’s scalability also broke down beyond roughly 35 cores, whereas OpenLB continued to benefit from additional processors. Moving to the server’s NVIDIA A100 GPUs brought a further speedup of about a factor of ten over MPI, reaching nearly 2,500 MLUPs on the finest meshes, a performance in line with the best lattice Boltzmann implementations reported elsewhere.

Combining these effects, the validated benchmark case achieved an overall acceleration of approximately 46-fold relative to the OpenFOAM reference, while for the most refined case examined, about 22 million cells, the combined CPU and GPU gains suggest a potential speedup approaching 230-fold, a figure the authors present as an indication of scalability rather than a validated result. True real-time simulation, where computation keeps pace with simulated time, was not achieved: GPU runs showed time ratios between 10 and 100. Even so, the researchers point out that multiple GPUs could close much of the remaining gap, and that the approach is already accessible far beyond supercomputing centers, since graphics cards with around 6,000 CUDA cores suitable for such computations are available for roughly 600 euros, raising the prospect of transient airflow simulations on a standard laptop.

The authors are careful to delineate the limits of their findings. The benchmark used a closed building, a nearly uniform inflow profile and an incompressible turbulent flow on a structured Cartesian lattice, precisely the class of problems for which the lattice Boltzmann method is best suited. More realistic scenarios involving internal ventilation, complex geometries, turbulent approaching flows or strongly coupled transport processes may show smaller efficiency gains and require case-by-case evaluation. A key constraint of the current solver version is its multi-block approach, which does not yet support local grid refinement, forcing uniform fine resolution across the entire domain and straining GPU memory in multiscale problems; a forthcoming release with local refinement is expected to be assessed in future work. Nevertheless, the study concludes that the open-source framework delivers accuracy comparable to established finite volume solvers at a fraction of the computational cost, positioning lattice Boltzmann methods as a credible foundation for the fast, transient dispersion simulations needed to quantify and mitigate emissions from naturally ventilated livestock buildings.

Subject of Research: Performance assessment of an open-source lattice Boltzmann framework for accelerated turbulent airflow simulation around agricultural buildings

Article Title: Accelerating airflow simulations for agricultural applications: Performance assessment of an open-source lattice Boltzmann framework

Article References: Moreau, B., Caiazzo, A., Amon, T., Chen, L., & Janke, D. (2026). Accelerating airflow simulations for agricultural applications: Performance assessment of an open-source lattice Boltzmann framework. Results in Engineering, 32, Article 112822. https://doi.org/10.1016/j.rineng.2026.112822

Image Credits: AI Generated

DOI: 10.1016/j.rineng.2026.112822

Keywords: lattice Boltzmann method, OpenLB, computational fluid dynamics, livestock housing, natural ventilation, GPU computing, OpenFOAM, turbulence modeling, wind tunnel validation, agricultural emissions, high performance computing, near-real-time simulation