Granular Jamming Soft Robotics - Grip Strength and Shock Absorption

Brief description

The dataset contains two subsets related to the performance of different grains in two soft robotics test scenarios.

The first shows the performance of various different grains when used in a "universal" granular jamming gripper to grip various objects, with retention force measured by a load cell.

The second shows performance of the same grains when used as a shock absorbing component to reduce damage to robots during falls, etc.

The data aims to guide researchers interested in granular jamming soft robotics in choosing their components to maximise performance.
Lineage: Data was collected in lab N2 on CSIRO's Pullenvale site, Brisbane. Data collectors were members of CSIRO's Robotics and Autonomous Systems Group.

Gripping experiments:

A 12.5cm (inflated) latex balloon was filled with the grain of choice. This was attached to a Dremel workstation drill press via a 3D printed adapter. A Thomas vacuum pump (model 107CDC20 H) was then connected to the balloon gripper. A 15mm sphere was placed in the load cell as a test object. A flat platform (50mm x 50mm) was attached between the object and load cell to replicate the action of picking an object off a flat surface.

Testing involved lowering the gripper onto the object in an unjammed state. Once in position, the vacuum was activated and the gripper slowly raised till it completely cleared the test object. A vacuum regulator was used to ensure the pressure was consistently measuring -0.6bar. The vacuum was then deactivated and the gripper manually reset by giving it a small shake. Each combination of grain shape and size was tested 10 times per test object.


Shock experiments:

An analysis of granular shock absorption properties was conducted by placing an IMU (Radioland Technology NRF52832 Beacon) inside a 3D printed disc adapter. A silicon sleeve was then filled with a chosen grain type and stretched over the disc to keep the contents in place. The sleeve was filled to a marked point in order to maintain a constant volume.

Once the wireless IMU had been initialised, the sleeve was manually lifted to a marked height of 30cm and dropped. This was repeated 5 times per grain type with a 2 second pause prior to and after dropping the vessel.

To interpret the data, raw data is extracted from a .csv file using a python script. The recorded data is cyclical and as a result the script records the maximum value between two minimums as a test once a spike is detected.
-For retention force tests the data is already in the appropriate units and no further processing is required.
-For shock tests, the magnitude of the IMU data is converted to G's using a scaling factor of 0.244/1000. This corresponds to the beacons output and sensitivity of 16 bits and +-8G's respectively.