This setup uses a Sparkfun ADXL377 as acceleration sensor which sends data to an Arduino nano. The Arduino logs the data to an SD card in binary format. The data on the SD card can be evaluated using python scripts.
The sensor was attached to a motorcycle license plate. Vibrations in the license plate can lead to structural damage and should be mitigated. For this, first the frequencies of the vibrations should be evaluated. The following picture shows the setup of the system. In the background the accelerometer is mounted to the back of the license plate. In the foreground, the system can be seen.
The results of the frequency analysis are shown in the following figure. A short ride with the motorcycle is recorded by the accelerometer. The above plot shows the time series of the total acceleration and the below plot shows the short time Fourier transform (STFT) up to 250 Hz. As can be seen in the STFT, there are several distinct frequencies at which the license plate is excited. These often have harmonic dependencies between them. The modulation of the frequencies correspond to changes in motor RPM. Higher motor RPMs show higher frequencies as expected and also higher vibration amplitudes.
Since the frequencies show a lot of modulation, a passive way to mitigate the vibrations needs to cover a broad frequency range. Alternatively, active measures can be taken to reduce the vibrations such as dynamic shakers.
Interfacing with the Arduino is done with the platformio library. The included Makefile can be used to upload the code to the Arduino Nano by running
make upload
The SdFat library is used to log data to the SD card in a very fast way using the AnalogBinLogger as an example. This produces an binary file on the SD card but after the logging finished the Arduino can convert this file to CSV.
For the post processing the following important python modules are used:
| Python Module |
|---|
| SciPy |
| Numpy |
| Matplotlib |
The components used are:
- Arduino Nano
- Sparkfun ADXL377
- SD Card Breakout
- Push Button
- LED
- 9V battery as power supply (many options available)
| ADXL377 Pin | Arduino Pin |
|---|---|
| 3.3V | 3.3V |
| GND | GND |
| Z | A1 (analog 1) |
Note that the sensor needs a capacitor from the analog output to ground as a low pass filter. This capacitor can be chosen based on the desired bandwidth (see datasheet).
| SD Card Breakout | Arduino Nano |
|---|---|
| 5V | 5V |
| GND | GND |
| MOSI | MOSI (11) |
| MISO | MISO (12) |
| SCK | SCK (13) |
| CS | 10 |
The Arduino Nano is powered through a 9V battery. Note that there is a switch between the battery and the Arduino to be able to shut down the whole system. This switch can be mounted anywhere in the power line.
| Arduino Nano | Battery |
|---|---|
| Vin | + |
| GND | - |
In order to have as little effect of voltage changes of the battery on the accelerometer, it is advisable to use decoupling capacities. For further information consult the datasheet of the ADXL377.
For interaction with the system, there is a button with is mounted between GND and digital pin 7. Additionally, a LED is mounted with the proper resistor between GND and digital pin 5. The button is used to start and end the logging process while the LED indicates logging activity. Do not turn off the power while the LED is still on. The LED remains on after the button is pushed to turn off logging since the binary log file is converted to CSV format which takes some time especially for longer periods of logging.