数据采集​​协议确定嵌入式灵敏度功能

The overall goal of this procedure is to determine the embedded sensitivity functions of a structure. The procedure will be demonstrated on a residential-scale wind turbine blade. This method can help answer key questions related to structural health monitoring such as how a structure's response will change due to damage at a certain location.

The main advantage of this technique is that it provides a method for modeling a structure based on experimentally measured data, eliminating the need for knowing specific parameters for mass, stiffness, and damping. Begin with designing a test fixture to replicate realistic boundary conditions. In this example, bolt locations are prepared to match the mounting locations of the blade.

Fabricate the fixture from steel to minimize the fixture's contribution to the dynamic response of the test specimen. Assemble the fixture by first bolting the blade to the custom T-bracket and then clamping the fixture to a steel table. Now, identify and mark a grid of impact locations on the blade.

Use a marker or wax pen to mark and number 30 points that span the entire blade. Then, measure out the point's relative position to use for the visual representation of the test results. Next, prepare single-axis 10 millivolts per g accelerometers.

Choosing the proper sensitivity of your accelerometer is critical in getting good signal to noise ratios. Also, be sure that the frequency range of the accelerometer is sufficient to capture the frequency range of interest of your specimen. Calibrate each sensor.

Attach a sensor to a handheld shaker that can output a single frequency force of 9.81 meters per second squared. Measure the sensor's response for two seconds of shaking. The force output is provided in the software readout.

Multiply the RMS amplitude by 1000 to determine the calibration factor for the accelerometer in millivolts per g. The next step is to prepare an impact hammer which has a sensitivity of 11.2 millivolts per newton. Make certain the hammer can excite the test specimen in both amplitude and frequency.

Then, attach a nylon tip to the hammer which will not compromise its function. Lastly, connect the hammer to the data acquisition system via a BNC cable. Now, identify sensor locations on the blade and attach the accelerometers with super glue.

Choose locations at points m and n on either side of the damaged location. Then, mount a third accelerometer at location k. Data from this sensor will be used to validate the results of the embedded sensitivity function analysis.

Open the data acquisition GUI. First, enable double hit detection. Then, set the sampling frequency to 10, 240 hertz.

The usable frequency range is half the sampling frequency. Third, set the sample time to one second. Fourth, select the hammer channel as the trigger channel and set the trigger level to 10 newtons.

Fifth, set the pre-trigger length to 5%of the total sample time. The pre-trigger data is data collected and stored in a buffer before the data acquisition system is started. It's important to retrieve and save this data so that the entire impact event is captured.

Sixth, select the H1 FRF estimator which assumes that there is noise on the response channels and no noise on the force channel. Lastly, enter the accelerometer and hammer information including calibration factors and identification notes. Then, save the settings for record keeping and for use in future tests.

Once the super glue used to attach the sensors has fully cured, impact point one with the hammer. When the amplitude of the impact force exceeds the chosen trigger level, the data acquisition system will be triggered and data, including the selected amount of pre-trigger data, will begin recording. During data acquisition, monitor the channels in the software to avoid clipping and double impacts.

Also, observe the coherence plot to further evaluate the quality of the acquired data. Do not window the data during acquisition. Impact point one four more times with consistent impact amplitudes.

Then, repeat this process for each selected point on the blade. After impacting all the points, repeat the process entirely on the damaged blade. The data from the damaged blade is only needed to evaluate the effectiveness of the embedded sensitivity functions.

It is not necessary to determine the embedded sensitivity functions themselves. Similar to frequency response functions, the embedded sensitivity functions have peaks near the natural frequencies of the structure. The higher the value of the functions, the more sensitive the location is to damage between points m and n.

Take, for example, the amplitudes of the function near 142 Hertz. It is clear that sensor locations corresponding to the squares in the first and third columns are most sensitive to the damage. Note that these locations were determined from data acquired with a healthy blade.

The difference between the frequency response functions determined from the healthy blade and the frequency response functions determined from the damaged blade shows that the embedded sensitivity functions are very effective at predicting the locations on the blade which are most sensitive to damage, as evidenced by the similarity between these two plots. After watching this video, you should have a good understanding of how to acquire the data necessary to determine a structure's embedded sensitivity functions. Although demonstrated on a wind turbine blade, this procedure is applicable to any structure whose response can be measured with an impact hammer and an accelerometer.