Final Design

Final Integrated Design

How it Works

The final design consists of three components: 3D printed prosthesis, jamming sheet, control and sensor electronics

The entire system is controlled by a Arduino Mega 2560 that is used to provide Bang-Bang feedback pressure control of the jamming sheet. This is done by controlling pumps using a small dual motor driver (TB6612FNG) and receiving sensor information from both a force sensitive resistor (FSR) sensor and an Adafruit MPRLS ported air pressure sensor. The FSR sensor is responsible for monitoring contact pressure at a pain point between the jamming sheet and residual limb, the reading of which can be compared to a predetermined reference pressure value as the threshold of allowable contact pressure. A Bang-Bang feedback controller involving the air pressure sensor, solenoid valve, pumps, and jamming sheet is then used to modulate stiffness of the jamming sheet. This is done by controlling the air pressure measured inside the pump system to within an allowable range of pressures. This system allows the use of either a contact-based or air pressure-based feedback control, with the potential for future connection and adaptation to a custom pressure sensing system being developed by the CHEI Lab.

The jamming sheet consists of layers of PET sealed inside a customized silicone membrane and connected to a pump system. In the final design, 16 layers of PET sheets cut into a hexagon pattern were used, with each hexagon connected via serpentine springs and layered such that the springs of different layers do not interfere when the springs elongate. This pattern enables 3D deformation of a 2D sheet without significant stresses, excess material, or permanent deformation at the high displacements. By inducing a vacuum inside the silicone membrane, one can modulate the stiffness by controlling the pressure. The pressure difference between the inside of the silicone membrane and the atmosphere changes the friction between the layers of the PET, thereby changing the yield stress of the entire soft robotic jamming sheet. This allows precise control of the stiffness of the soft robotic jamming mechanism, which allows controlled deformation of the jamming sheet when a load from an expanding residual limb is applied, thus control of the pressure applied to the residual limb.

The final system of electronics and jamming sheet is integrated into a cutout in the socket of a 3D printed nylon prosthesis at a predetermined pain point. The final jamming sheet along with FSR is then adhered to the socket cutout, where it interacts with the residual limb, providing support and additional comfort. Electronics including the pressure sensor, valve, pumps, and Arduino board are intended to be stored below the socket pocket of the prosthesis in a miniaturized version and are connected to the jamming sheet and FSR through a strip of wires and one tubing. This project stored the electronics on an external mounting platform for testing and diagnostic purposes, leaving miniaturization to a future engineering team at the request of the sponsor.

What is Layer (or Laminar) Jamming?

Multiple layers of thin sheets with high friction coefficients
Slips when pressure inside silicone membrane equals atmospheric pressure
Layers jam (lock together) when pressure inside is less than atmospheric pressure

Electrical Diagram of Vacuum Control System

Performance

Following quantitative and qualitative material and prosthesis testing, it was found that the soft robotic jamming mechanism consisting of 12 layers of specially designed PET layer jamming sheets is able to reach a stiffness ratio of ranging from 4 to 7 times the unjammed stiffness at -60 kPa gauge pressure, as seen below.

Material Properties of Membrane with 12 Jamming Sheets and Tested in the Instron.

When vacuumed to -60 kPa gauge pressure (yellow), the layer jamming membrane was able to achieve up to 600% increase in stiffness.

Results of Instron PET spring tensile test. Two different thicknesses of PET sheets were evaluated, with a 75% cut spring selected for the final design.

The 75% length serpentine spring (cut 75% of the width or vertical component of the rectangular area that contains the retracted serpentine spring) was chosen due to its high yield displacement and satisfactory stiffness. The spring pattern used can be seen below in both the stretched and unstretched state:

Leaking Test

For the airtightness of the entire system (vacuum system and the soft robotic mechanism), leaking tests were conducted where the vacuum system holds a vacuum of absolute pressure 45±3 kPa for 400 seconds, and the air pressure registered by the air pressure sensor was recorded. The entire system only requires ~3 seconds of actuation every 5 minutes to keep the pressure within this range. This suggests that the system has very good sealing, given that a slight change in volume in the tubing and membranes can lead to a significant change in pressure.

Balloon Test

This test uses a rubber balloon to simulate a 10% increase in the volume of residual limb. A rubber balloon was inserted into the prosthesis and was gradually inflated with water. The test measurement began once the FSR attached on the inner side of the socket pocket started to measure a difference in the contact pressure. This test was repeated for three vacuum pressures (101 kPa, 70 kPa, and 45 kPa absolute pressure) to evaluate the effect of vacuuming on contact pressure. This test verified that the soft robotic jamming mechanism could accommodate over 10% volume increases, and that there was a clear relationship between the amount of vacuum pressure held and the contact pressure applied at various volume increase, despite an error of 5% in the measured loading by the FSR.

(Left) Setup of the balloon test. A syringe was used to measure and control volume change in the balloon. (Right) Results of 1 iteration of the balloon test with curve smoothing to reduce the noise in the sensor readings. Other iterations of this test also suggested that there was a more significant difference in contact pressure between 45 kPa and 70 kPa than indicated in this graph.

Page updated

Report abuse