2023 - 2026
Anna Bagnara
Stefano Torre
Edoardo Ruffini
David Hankin
Camila Aldereguía Sánchez
Federica Tubino
+2 more
2023 - 2026
Thunderstorm Loading on Tall Buildings
ERIES-TLTB
Dataset Description
This Project aims to experimentally investigate the impact of thunderstorms on tall buildings. Thunderstorm and synoptic profiles are simulated in order to compare the effect of the two phenomena. Thunderstorm profiles are reproduced using two special grids: a passive one and an active one. Wind-induced loads on two building models, the benchmark CAARC building and a specific tall building (Project X), are measured through High Frequency Force Balance (HFFB) and High Frequency Pressure Integration (HFPI) techniques. The dataset corresponds to the flow velocity measurements of different profiles, as well as forces and pressures measured in HFFB and HFPI tests performed in the "Giovanni Solari" Wind Tunnel at University of Genoa.
Specimens
1. Special Grids
2
The simulation of thunderstorm wind properties is achieved using two grids specially designed within the project, here referred to as passive and active grid.
The passive grid consists of rectangular modules that span the full width of the wind tunnel, each containing square openings of varying sizes. The basic purpose of this device is to slow down the flow velocity at the upper part of the test chamber. The laboratory setup includes six modules—arranged in pairs—with small, medium, and large openings, corresponding to porosities of approximately 41%, 46%, and 52%. Positioned 4 m downstream of the wind tunnel inlet, the grid occupies about two-thirds of the test section height. This passive grid is capable of reproducing the characteristic nose-shaped mean wind speed profile, but it does not account for non-stationary flow features.
The active grid consists of four independently controlled rotating blades that redirect the incoming flow through coordinated motion. Each blade rapidly rotates to predefined angles and subsequently returns to its initial position, generating short-duration fluctuations in wind velocity and introducing transient characteristics into the flow.
The blades are manufactured from carbon fiber and incorporate a core structure spanning 1.7 m across the wind tunnel. Each blade has a width of 48.5 cm and is mounted on two steel frames that act as lateral supports. These frames also contain the driving modules, which are operated through independent signal channels integrated into the control system.
In addition to rotational motion, the blades can be independently repositioned along the vertical direction, allowing flexibility in reproducing different target flow profiles depending on experimental requirements.
1. Replication of the target profiles using the passive grid
Thunderstorm target profiles are defined according to the Wood and Kwok model [1].
In an initial phase, these profiles are reproduced in the wind tunnel using a passive grid by combining modules in different configurations, allowing a wide range of nose heights to be achieved. However, the attainable turbulence intensity with this setup is limited to approximately 15%. To account for the inherent uncertainty in key thunderstorm parameters, a set of wind speed profiles is selected for testing, covering variations in both nose height and turbulence intensity. Specifically, two nose heights are considered (100 m, TS1; and 60 m, TS2), together with two turbulence intensity levels (10%, LT; and 15%, HT).
For comparison with classical synoptic wind conditions, mean wind speed and turbulence intensity profiles are also generated based on the formulations and terrain categories defined in EN 1991-1-4 [2]. Two terrain configurations are considered: S-LT (roughness length 0.003 m) and S-HT (roughness length 0.05 m). In addition, a reference profile (S-REF, roughness length ≈ 0.2 m) is reproduced to enable comparison with previous studies on the CAARC building. These synoptic profiles are simulated using conventional devices, including spires, fences, and cube panels.
[1] Wood, S. and Kwok, K., An empirically derived estimate for the mean velocity profile of a thunderstorm downburst, Proceedings of the 7th Australian Wind Engineering Society Workshop, Auckland, 1998.
[2] EN 1991-1-4, Eurocode 1: Actions on structures – Part 1-4: General actions – Wind actions, European Committee for Standardization, 2005.
Instrumentation
The wind tunnel's reference flow velocity is measured using a pitot-static tube, positioned 20 cm from the test chamber roof and upstream of the model to avoid interference. The signal is filtered, digitized, and saved at 100 Hz. The velocity profile is measured using a Cobra Probe positioned at different heights. The Cobra probe is capable of measuring the three velocity components with a sampling frequency of 2 kHz. The accuracy of measurements is generally within +/−0.5 m/s and +/−1° pitch and yaw angle, up to about 30% turbulence intensity. The Cobra probe range of measurement is from 2 to 100 m/s within a 45° cone of its x-axis. This probe can be placed in the wind tunnel through a motorized 3-degree of-freedom traversing system placed above the second test chamber. The Cobra probe is located approximately 3.2 m downstream from the passive grid, at the center of the turning table.
2. Replication of the target profiles using the active grid
Among the tested cases, the thunderstorm profile TS1-LT (zm = 100 m, Iu ≈ 10%), previously generated under stationary conditions with the passive grid, is selected for further investigation using the active grid. The profile is reproduced under stationary conditions by fixing the blade angles, ensuring consistency with the passive grid results. The angles of each blade (from the top to the bottom) was 60°, 15°, 30° and -10°.
Instrumentation
The wind velocity at different heights is measured with the Cobra probe. The Cobra probe is capable of measuring the three velocity components with a sampling frequency of 2 kHz. The accuracy of measurements is generally within +/−0.5 m/s and +/−1° pitch and yaw angle, up to about 30% turbulence intensity. The Cobra probe range of measurement is from 2 to 100 m/s within a 45° cone of its x-axis. This probe can be placed in the wind tunnel through a motorized 3-degree of-freedom traversing system placed above the second test chamber. The Cobra probe is located approximately 2.5 m downstream from the active grid, at the center of the turning table.
2. Model 1 - CAARC HFFB model
1
The model is built at a geometric scale of 1:400 with dimensions of 76.2 mm in depth, 114.3 mm in width, and 457.2 mm in height. The force balance model is designed to optimize both stiffness and lightness to avoid resonance effects at low to moderate frequencies. To achieve this, a specialized high-modulus carbon fiber "sting" is utilized, with a mass of 205 g, an outer diameter of 30 mm, and a height of 394 mm. The tower envelope is constructed using high-density polymeric foam. The resulting composite structure— combining the carbon fiber spine and foam—achieves a natural frequency of approximately 100 Hz. The model is rigidly fixed to a 3.8 mm aluminum base plate through the spine to enable connection to the high-frequency response force balance. The model is tested in isolation.
1. HFFB tests using the passive grid
On CAARC HFFB model, the six components (three forces and three moments) of the resultants of the fluctuating wind loads at the base of the building model are measured. Due to the rectangular symmetry of the building, tests are carried out for 19 wind directions (0° - 180°) at 10° increments. Tests are conducted under six distinctic flow conditions (TS1-LT, TS1-HT, TS2-LT, S-LT, S-HT, S-REF). The tests first assess whether the force and moment coefficients remain consistent across different Reynolds numbers. For a fixed building position, forces and moments are measured at seven wind tunnel speeds, and mean coefficients are calculated. The model scale sampling frequency (360 Hz) and the duration of the acquisition (143 s) are set in order to guarantee a full-scale cut-off frequency of at least 2 Hz and a suitable length of the records allowing for a reliable statistical characterization of the wind loads.
Instrumentation
The wind tunnel's flow velocity is measured using a pitot tube, positioned 20 cm from the test chamber roof and upstream of the model to avoid interference. The signal is filtered, digitized, and saved at 100 Hz. The forces and moments are measured using a six-component high-frequency piezo-electric force balance in conjunction with signal conditioning (Kistler type 9257B Q02). The model is connected to a turntable that rotates automatically with a step of 10 degrees.
3. Model 2 - CAARC HFPI model
2
Model 2 (CAARC HFPI) is built at a geometric scale of 1:400 with dimensions of 76.2 mm in depth, 114.3 mm in width, and 457.2 mm in height. The pressure model is developed using rapid-prototyping techniques, thus achieving a high degree of precision (to a model scale tolerance of 0.1 mm). It is 3D printed in six parts using the Selective Laser Sintering, a popular additive manufacturing process. A total of 280 pressure taps are installed across the facades comprising 10 rings distributed vertically, with two of them coinciding with the heights of the nose examined in this study. The distribution of these taps follows the layout proposed by Dragoiescu et al. [3], incorporating the standard CAARC locations along with additional taps positioned near the building edges. The pressure transducers are connected to the surface pressure taps via 1.4 mm bore tubing of constant length. The model is tested in isolation.
[3] Dragoiescu, C., Garber, J. and Kumar, K.S., A comparison of force balance and pressure integration techniques for predicting wind-induced responses of tall buildings, Structures Congress 2006: Structural Engineering and Public Safety, 2006, pp. 1-10.
1. HFPI tests using the passive grid
Fluctuating wind pressures are measured on the CAARC HFPI model for 19 wind directions (0° - 180°) at 10° increments, considering the rectangular symmetry of the building. Tests are conducted under five flow conditions (TS1-LT, TS1-HT, TS2-LT, TS2-HT, S-LT, S-HT). The pressure transducers measure the differential pressure between the surface pressure taps and a reference static pressure. This reference pressure corresponds to the static pressure within the test chamber during wind tunnel operation and is obtained by averaging measurements from 12 pressure taps distributed along the roof and walls near the turntable. These taps are connected via a manifold directly to the pressure transducers. The pressure transducers are connected to the surface pressure taps via 1.4 mm bore tubing of constant length. In order to correct for magnitude and phase distortion in the measured pressure signals, purposely designed (and length-specific) numerical filters are applied in the time domain during post-processing of the acquired wind tunnel data. The model scale sampling frequency and the duration of the acquisition are set in order to guarantee a full-scale cut-off frequency of at least 2.5 Hz and a suitable length of the records allowing for a reliable statistical characterization of the wind loads.
Instrumentation
The wind tunnel's flow velocity is measured using a pitot-static tube, positioned 20 cm from the test chamber roof and upstream of the model to avoid interference. The signal is filtered, digitized, and saved at 100 Hz. The fluctuating pressures are measured using a simultaneous multi-channel low-range pressure scanning system. The system uses Sensor Techniques low range SLP004D differential transducers. The transducers are arranged in blocks of eight, each sharing a reference pressure. This configuration enables calibration in the laboratory as well as periodic verification of calibration throughout the tests.
2. HFPI tests using active grid
The model is also tested using the active grid under thunderstorm flow conditions. Experiments are conducted for five wind directions, ranging from 0° to 90° in 30° increments; an additional test at 180° is included to verify symmetry. Surface pressure measurements are collected along the building facades. HFPI test was performed with fixed blade angles, resulting in stationary flow conditions to enable direct comparison between results obtained using the passive and active grids.
Instrumentation
The wind tunnel's reference flow velocity is measured using a pitot-static tube, positioned 20 cm from the test chamber roof and upstream of the model to avoid interference. The signal is filtered, digitized, and saved at 100 Hz. The fluctuating pressures are measured using a simultaneous multi-channel low-range pressure scanning system. The system uses Sensor Techniques low range SLP004D differential transducers. The transducers are arranged in blocks of eight, each sharing a reference pressure. This configuration enables calibration in the laboratory as well as periodic verification of calibration throughout the tests.
4. Model 3 - Project X HFPI model & surround model
1
Model 3 (Project X HFPI) is built at a geometric scale of 1:300. The full-scale building, reaching approximately 165 meters in height, stands atop a podium. The tower has full-scale dimensions of 20.5 meters in depth and 25 meters in width, approximately. It features two symmetrical protrusions that extend along its height on the largest facades. Towards the top, the cross section narrows and takes on a spiky shape. The pressure model is developed using rapid-prototyping techniques, thus achieving a high degree of precision (to a model scale tolerance of 0.1 mm). The model is 3D printed in ten parts via Selective Laser Sintering and instrumented with a total number of 365 pressure taps. Taps are concentrated near geometric discontinuities such as edges and corners, where flow separation typically occur. Additionally, uniform coverage is ensured across major surfaces (facades, roof and podium). The surrounding area is modelled out to a 240-meter radius from the center of the site. The nearby buildings are constructed from foam blocks detailed enough to replicate the wind flows at the proposed development location.
1. HFPI tests using special passive grid
The fluctuating wind pressures are measured on the Project X HFPI model for 36 wind directions (0° - 350°) at 10° increments. The pressures are measured for two target profiles: the worst-case thunderstorm mean wind speed profile from the experiments carried out on CAARC building (TS1-HT) and a single reference synoptic profile (S-ProjectX), corresponding to an effective roughness length of 0.5 m, that can be considered on the average as representative of the construction site for all wind directions. The model scale sampling frequency and the duration of the acquisition are set in order to guarantee a full-scale cut-off frequency of at least 2.5 Hz and a suitable length of the records allowing for a reliable statistical characterization of the wind loads.
Instrumentation
The wind tunnel's reference flow velocity is measured using a pitot-static tube, positioned 20 cm from the test chamber roof and upstream of the model to avoid interference. The signal is filtered, digitized, and saved at 100 Hz. The fluctuating pressures are measured using a simultaneous multi-channel low-range pressure scanning system. The system uses Sensor Techniques low range SLP004D differential transducers. The transducers are arranged in blocks of eight, each sharing a reference pressure. This configuration enables calibration in the laboratory as well as periodic verification of calibration throughout the tests.
Project Metadata
Rights
Creative Commons Attribution 4.0 International.
CC BY 4.0
2 sessions
0 downloads
2 views
0 metadata
4 file previews
Feedback
We are always looking to improve the quality of our data and metadata. If you have any feedback or suggestions, please let us know.
Send Feedback