Generating Noise Inputs for Shake Table Testing

Shake tables are widely used in structural and mechanical engineering research to simulate dynamic loads, including earthquakes, vibrations, and random noise inputs. One common requirement in laboratory testing is generating noise-based input signals to study how structures respond to broadband vibrations. This blog post will guide you through the process of generating noise signals and inputting them into a shake table, with a focus on achieving controlled displacement amplitudes.

Understanding Noise Inputs for Shake Tables

Noise inputs refer to random or controlled vibration signals that can be applied to a test structure using a shake table. Unlike sinusoidal or earthquake simulations, noise-based inputs provide a broad spectrum of frequency content, making them useful for:

• Structural modal identification
• Fatigue testing under random loads
• Simulation of real-world environmental vibrations
• Testing damping characteristics of structures

Key Considerations When Generating Noise Signals

Before applying a noise signal to a shake table, consider the following factors:

1. Desired Displacement Amplitude: If you aim to achieve a maximum vibration amplitude (e.g., 1 cm), you must carefully scale your input signal. Displacement is related to acceleration and frequency through fundamental vibration equations.
2. Frequency Content: White noise provides a flat frequency spectrum, whereas filtered noise can be tailored to a specific range (e.g., low-frequency dominant vibrations).
3. Shake Table Limits: Ensure that your generated input signal does not exceed the physical displacement, velocity, or acceleration limits of your shake table.

Methods for Generating Noise-Based Inputs

There are multiple approaches to generating noise signals for shake tables:

1. Using MATLAB or Python for Signal Generation

Both MATLAB and Python (with libraries like NumPy and SciPy) can generate noise signals in a format compatible with shake table controllers.

• MATLAB Example:

fs = 1000; % Sampling frequency in Hz  
t = 0:1/fs:10; % Time vector for 10 seconds  
noise_signal = 0.01 * randn(size(t)); % Generate white noise scaled to desired amplitude  
csvwrite('noise_input.csv', noise_signal); % Save the signal as a CSV file  

• Python Example:

import numpy as np  
import pandas as pd  

fs = 1000  # Sampling frequency in Hz  
t = np.linspace(0, 10, fs*10)  # Time vector for 10 seconds  
noise_signal = 0.01 * np.random.randn(len(t))  # Generate white noise  

# Save the noise signal to a CSV file  
pd.DataFrame(noise_signal).to_csv('noise_input.csv', index=False, header=False)  

These signals can then be uploaded to the shake table control software.

2. Using EASYTEST Software for Signal Generation

EASYTEST is the primary software used by most of our shake tables for control and signal processing. It provides a user-friendly interface to generate various types of signals, including:

• White noise and filtered noise
• Sine sweep signals for frequency response analysis
• Custom waveform inputs based on experimental requirements

How to Use EASYTEST for Noise-Based Testing:

1. Open EASYTEST and navigate to the signal generation module.
2. Select the “Random Noise” option and configure the amplitude and frequency range.
3. Specify the duration and sampling rate for the test.
4. Load the generated signal into the shake table controller and run the test.

3. Sine-Sweep Testing for Structural Identification

Before applying a noise input, it is often helpful to conduct a sine-sweep test to identify the resonance frequencies of the test structure. EASYTEST can also be used to generate sine-sweep signals that gradually increase or decrease in frequency over time. This helps in fine-tuning the noise signal to focus on critical frequency ranges.

Implementing the Noise Input on a Shake Table

Once the noise signal has been generated, follow these steps to apply it to your shake table:

1. Convert the Signal Format: Ensure the signal is in a format supported by your shake table control system (CSV, TXT, or direct software input).
2. Scale the Input Properly: If a displacement of 1 cm is required, ensure the noise amplitude is scaled appropriately.
3. Load the Input into EASYTEST or Shake Table Controller: Import the file and preview the waveform.
4. Run a Test Simulation: Before running the actual experiment, conduct a short-duration test to verify that the desired displacement is achieved.
5. Analyze Results: Use accelerometers or displacement sensors to confirm the input and response of the structure.

Conclusion

Generating noise-based inputs for shake table testing is a powerful way to simulate real-world vibration conditions. Whether using MATLAB, Python, or the EASYTEST software, researchers can create controlled random vibration signals tailored to their experimental needs. By understanding the relationship between frequency, displacement, and acceleration, users can ensure precise control over the shake table’s motion.

For users of our shake tables, we highly recommend using EASYTEST for signal generation and control. If you have any questions about generating noise inputs or using EASYTEST, feel free to reach out to us at support@quakelogic.net.

Seeing is Believing!