Towards a hybrid approach for diagnostics and prognostics of planetary gearboxes

# Folders and files
- g1_crack: Gear 1, 8 different damage states (crack lengths) tested at different operating conditions.
- .MAT files contain the measured data. G1 refers to gear 1, P refers to the testing phase or degree of damage (crack length) tested at.
- G1_microscope_images is a folder with photos of the gear with different crack lengths tested.
- g2_crack: Gear 2, 6 different damage states at different operating conditions.
- .MAT files contain the measured data. G2 refers to gear 2, P refers to the testing phase or degree of damage
(crack length) tested at.
- G2_microscope_images is a folder with photos of the gear with different crack lengths tested.
- missing_tooth: Data for a planet gear with a missing planet gear tooth.
This dataset's original purpose was to see if it is possible to grow a crack in one of the gear teeth in situ (It
was not possible).
There is, therefore, in addition to the missing gear teeth, one of the gear teeth that is cracked (crack depth ~20%
of gear width).
However, the vibration response is dominated by the excitation due to the missing tooth.
- compare_healthy_with_missing_tooth: Two example datasets measured in healthy and damaged condition.
Provides simple way to get exposed to the diagnostics problem for planetary gearboxes. This dataset is valuable for learning how about how a fault manifests in the vibration response of a planetary gearbox.
However, there are independent variables (see below) that are not kept constant between the healthy and damaged cases. Therefore, these datasets should only be used as a convenient introduction to the planetary gearbox diagnostics problem,
and not for publication.

-Healthy.h5 is the vibration response of the gearbox in healthy condition, but the gearbox housing is not yet modified
as is the case with the missing_tooth.h5 dataset (See dissertation).
The vibration signature of the housing may therefore be slightly different from the missing tooth dataset. However, this
likely has negligible influence considering the severity of the damage.
-The datasets in this example are chosen specifically to have similar measured RPM values.

# Information
Data is gathered for a planetary gearbox with a single planet gear to simplify detection of the damage.
The shaft key, as detected by the magnetic pickup, is aligned with the planet gear.
The planetary gearbox is the Bonfiglioli 300-L 1-5.77-PC-V01 B·E, with 13 sun gear teeth, 24 planet gear teeth and
62 ring gear teeth. Gear ratio:5.77.

For more information, please consult the dissertation "Towards a hybrid approach for diagnostics and prognostics of
planetary gearboxes" or contact Douw Marx at douwmarx@gmail.com.

# Folders
-h5: Data in .h5 format
-raw: Data in original format as measured
-load_data_with_python: Simple script to easily load the .h5 data with Python.

For more information, please consult the dissertation "Towards a hybrid approach for diagnostics and prognostics of
planetary gearboxes" (https://repository.up.ac.za/) (https://repository.up.ac.za/handle/2263/78157) or contact Douw Marx at douwmarx@gmail.com.

# Attribute Information:
All channels measured, in order:
1. 'Time': Time [s]
2. 'Acc_Carrier': Carrier side accelerometer [m/s^2]
3. 'Acc_Sun': Carrier side accelerometer [m/s^2]
4. 'Tacho_Carrier': Zebra tape tacho encoder on carrier-side
5. 'Tacho_Sun': Zebra tape tacho encoder on sun-side
6. '1PR_Mag_Pickup': Magnetic pickup
7. 'T_amb': Ambient temperature [Celsius]
8. 'T_oil': Oil temperature [Celsius]
9. 'Torque': Motor current proportional to applied torque


# Code
For data processing code see https://github.com/DouwMarx/Hybrid_Approach_To_Planetary_Gearbox_Prognostics