Wind load prediction on single tree with integrated approach of L-system fractal model, wind tunnel, and tree aerodynamic simulation

Abstract

In this work, we adopt the integration of the L-system fractal tree generation, 3D printed wind tunnel modeling, and computational fluid dynamics (CFD) simulation approach to model the wind effect on a single tree. We compare the agreement between CFD simulations and wind tunnel measurements of rigid branched structures resembling trees. First, fractal tree mesh models based on species growth and branching patterns are developed to represent tree species for wind–tree modeling. Subsequently, a scaled-down fractal tree is generated with 3D-printing and subjected to tunnel testing with load cell and particle image velocimetry measurement data under the wind speed of 10 m/s and 15 m/s. Finally, CFD based on Reynolds-Average Navier–Stokes (RANS) simulation with a full closure model and Large Eddy Simulation (LES) using appropriate momentum sink and turbulence source terms for the volumetric tree is carried out. We use both the volume-average porous media and the volume-splitting discretized zones (split number 10 × 10 × 10) to reproduce the momentum sink effect in the numerical simulation. Three tree species, namely, Peltophorum pterocarpum (yellow flame), Khaya senegalensis (African mahogany), and Hopea odorata (ironwood), are tested, and a reasonable agreement of drag force prediction and velocity profiles is obtained when comparing the CFD simulation results with wind tunnel data. The RANS modeled drag force results exhibit 20% of over-prediction, while the normalized velocity profiles display a good match of velocity decay at the tree leeward sides. On the other hand, LES produces much better results with only 3% discrepancy with the experimental results. A comparison of experimental results among the tree species is also carried out. Due to the actual random wind direction, tree slenderness representation, and structural flexibility issues, the current methodology still has the limitation for validation with urban on-site measurement. Nonetheless, this integrated approach is the first step in establishing modeling tool applicability to examine the effect of the forest structure and composition on wind loads.