ERIES-FLAMBeRG

2024 - 2025

Wind Eng.

Stanislav Pospisil

Arsenii Trush

Piotr Gorski

Marcin Tatara

Karel Dejmal

Olivier Flamand

+1 more

2024 - 2025

Wind Eng.

Wind tunnel testing of FLexible Advanced aeroelastic Models of BRidGe cables

ERIES-FLAMBeRG

Wind loads

Aerodynamic forces

Dataset Description

An experimental campaign was carried out at the Jules Verne Climatic Wind Tunnel – Dynamic Unit SC1 at the Centre Scientifique et Technique du Bâtiment (CSTB) in Nantes, France.

The project aimed to improve wind testing facilities and develop an experimental methodology for laboratory simulation and investigation of airflow conditions around cables with climatic and technological roughness in particular- the study of Wind Induced Vibrations (WIV) of bridge cables. Advanced flexible aeroelastic models with the closest possible representation of the real shape of the cables, including the characteristic shape of cables covered with glaze ice, spiral pattern of the helical strake, etc… allow to achieve higher natural modes (up to 4 and higher) with the ability to fine-tune natural frequencies for testing the realistic Reynolds number range, as well as ensuring the 3-dimensional nature of cable-wind interactions in controlled laboratory conditions. The project involves the use of the close to real cable diameter for the model (110 mm) as well as a large ratio of diameter to length (1:40). It is also necessary to have space for placing fastening and tensioning equipment and cables. Thus, key parameter required for carrying out the tests is a wind tunnel cross section where the model stand can be placed. Within the framework of the project, 2 models with different surface roughness were tested:

1. smooth model (as reference) ;

2. model of cable with asymmetrical ice accretions – this model was tested at 3 different angles of attack on the ice "edge".

The steel core pretension for both models was set at 2 levels tuned for the frequencies of first natural mode f = 5 and 10 Hz. All models configurations were tested at 2 orientations with yaw angles 0° and 15°.

Bridge Cables

Ice Accretion

Self induced vibration

Vortex Shedding

Specimens

1. Test setup

After the assembly and installation of the rig on the turntable in the Aerodynamic working section of the Jules Verne climatic wind tunnel at CSTB in Nantes, France, 6m wide and 5m high for a length of 12m, the corresponding model was installed on it and configured, the yaw angle was set by rotating the entire rig on the turntable with the replacement of the acrylic wind screens with the corresponding ones.

Using the main tensioner, the frequency of the first oscillation mode was adjusted, the damping was estimated by free decay tests without wind flow. Each configuration was tested at wind speeds of approx. 0.9 from the critical ViV speed for the first oscillation mode and up to the speed corresponding to the end of the lock-in range for the maximum achieved oscillation mode with a change in speed in two directions: 1) increasing velocity from minimum to maximum to obtain upper branch and 2) decreasing velocity to obtain lower branch of response in lock-in range.

1. Experimental stand

To provide orientation (setting of the yaw angle) and pre-tension of the cable, a testing stand has been developed. The stand consists of two parts, each of them has beam (1), console (2), mechanical tensioners (3-4), transportation brackets (5) and 2 sets of flat Plexiglas screens. The base plates are fastened to existing HALFEN Framing channels (6) on the turn table (two positions, corresponding to yaw angles of 0 and 15 degrees).

Stand has mechanical threaded tensioners on both sides: on one side, the tensioner provides tension on the cable (3), ensuring structural integrity of the main (vertical and horizontal) eigenmodes and allows to achieve the required frequency. On the opposite side of the stand is tensioner (4) for strings which are responsible for the torsional rigidity of the model and the corresponding eigenmodes. The frame and tensioners are made of construction steel.

Tensioner consists of :

1.tensioner base plate;

2.tensioner guides / threaded rods with nuts;

3.tensioner support nuts;

4.the movable end of the steel rope core;

5.fixed ends of torsion strings.

Instrumentation

To monitor cable tension during installation and vibrations, a strain gauge force sensor MEGATRON type KM1603 K 50kN was attached to one end of the stand.

Data collection from accelerometers and force sensor was carried out using recording system based on the NI PXI architecture, in particular PXIe-6349 Multifunction I/O Module and and NI PXIe-4330 module. Acquisition frequency was 1000 Hz for shorter measurements or 200 Hz for long- indicated in the tables for each file.

Video monitoring system

Monitoring of the overall vibration form was recorded using a 2D digital video system. The cameras were installed to record both vertical vibrations (the plane of the frame coincides with the plane of vibrations) and horizontal deviation from the frontal drag (the camera is installed above the model). To improve visibility, position matching and further analysis square reflective targets with 10 mm side size were placed on top and leeward side of the model.

Wind flow measurement system

Free stream wind velocity was measured simultaneously during experiments using a Cobra probe and a Prandtl tube. To obtain Strouhal frequency two Cobra probes were installed in the wake behind the accelerometer installation points at a distance of 550 mm from the model’s axis. Data from flow velocity probes were recorded by the wind tunnel internal control system. .Aqusition frequency for Cobra probes was was 2048 Hz for shorter measurements or 1028 Hz for long. Data from the Prandtl tube and the air characteristics in the tunnel were taken at a frequency of 300 Hz. During post-processing, all wind flow data were synchronized by a trigger signal and resampled to the PXI system frequency using anti-aliasing filter.

2. Flexible smooth cable model

1. Flexible smooth cable model testing on WIV- low pretension level (first mode frequency 5 Hz), yaw angle orientation 0°.

In this setup, measurements for each wind speed were saved in a separate file.

pdf

Instrumentation

Model

The length of a flexible cable model consisting of a supporting core in the form of a small-diameter steel cable, and a shell consisting of alternating small sections made using 3D printing from rigid and flexible plastic models will be 4.4 m, diameter D = 0.11 m. Model with Ice accretions has basic diameter of circular cable 0.11 m with prominent asymmetrical parts. The length-to-diameter ratio≈40.

The internal structure of the smooth cylinder model with steel rope core and accelerometers

1. Stiff 3D printed part

2. Flexible 3d printed silicone shell.

3. Slots for miniature accelerometers from Bruel and Kjaer type 4374

4. Channel for accelerometers cables and installation

5. Strings for model torsion and orientation

6. Channels for torsional strings

7. Load-bearing core made of a steel rope

8. Channel for steel rope

9. Channel for steel rope fixing.

The dynamic responses were measured in with pairs of accelerometers Bruel and Kjaer type 4374 installed orthogonal to each other at 2 key points to measure acceleration along 2 axes (in the plane normal to the model axis). At following locations cable’s response was obtained for all 4 first eigenmodes. The gain level of the signal from the Bruel Kjaer conditioning amplifier Nexus 2692-A-0I4 was adjusted depending on the expected amplitude of oscillations in order to maximumly use the range and resolution of the A/D board.

2. Flexible smooth cable model testing on WIV - high pretension level (first mode frequency 10 Hz), yaw angle orientation 0°.

In this setup, measurements for each wind speed were saved in a separate file. After completing the measurement of the upper branch, it was found that due to stronger vibrations than in the case of low tension, 2 accelerometers were damaged at the point where the signal cable was attached to the sensor body. In this regard, in order to preserve the equipment and further conduct the study, it was decided to limit the test program by excluding further tests with high tension of the model.

pdf

Instrumentation

Identical to previous one

3. Flexible smooth cable model testing on WIV - low pretension level (first mode frequency 5 Hz), yaw angle orientation 15°.

In this setup, measurements for all speeds and both branches were saved in 1 file in order to reduce the time for setup operations at acquisition systems.

pdf

Instrumentation

Identical to previous one

3. Flexible model of the cable with ice accretions

The procedure was the same as in the experiment with the smooth circular shape model

1. Flexible iced cable model testing on WIV, angle of attack 90°- low pretension level (first mode frequency 5 Hz), yaw angle orientation 0°.

In this setup, measurements for each speed in upper branch of ViV phenomenon were saved in a separate file. Lower branch measurements for all speeds were saved in another file in order to reduce the time for setup operations at acquisition systems.

mp4

avi

mp4

avi

mp4

pdf

Instrumentation

Except for the characteristic rough surface, the internal design of the model is similar to the smooth cable model.

2. Flexible iced cable model testing on WIV, angle of attack 90°- low pretension level (first mode frequency 5 Hz), yaw angle orientation 15°.

In this setup, measurements for each speed were stored in a common file with a range indication.

avi

pdf

Instrumentation

Identical to previous one

3. Flexible iced cable model testing on WIV, angle of attack 0°- low pretension level (first mode frequency 5 Hz), yaw angle orientation 0°.

No significant ViV were obtained due to coupling of rotational and vertical oscillation modes and caused high structural damping. It was impossible to promptly change the design feature of model that caused this phenomenon. Further tests of model with angle of attack 0° and different yaw angles were canceled.