A biomimicking design for mechanical knee joints

Area of Science:

• Biomechanical engineering
• Robotics
• Prosthetics design

Background:

• Existing prosthetic knee joints often require large actuators and lack the natural compliance and sensory feedback of human knees.
• Constant moment arm designs in prosthetic joints present limitations in efficiency and adaptability during dynamic movements like stair ascent.

Purpose of the Study:

• To develop a bioinspired bicondylar knee joint that replicates human knee mechanics for improved prosthetic performance.
• To reduce actuator size and force requirements through a variable moment arm mechanism.
• To integrate artificial ligaments with stretch sensors for accurate joint angle estimation.

Main Methods:

• Designing a novel bicondylar knee mechanism mimicking elastic, rolling, and sliding elements of the human knee.
• Implementing a variable transmission system where the moment arm changes with joint angle.
• Integrating mechanical 'ligaments' with embedded stretch sensors to replicate neurosensory and compliant functions.

Main Results:

• The bioinspired joint achieved a 12% reduction in peak actuator force for stair ascent compared to constant moment arm designs.
• Demonstrated experimental validation of using ligament stretch to accurately estimate bicondylar joint angle.
• The new design addresses critical limitations in weight and power for robotic limbs.

Conclusions:

• The bioinspired bicondylar knee joint offers a more efficient and adaptable prosthetic solution by replicating human knee biomechanics.
• The variable moment arm and integrated sensing system overcome key challenges in prosthetic joint design and control.
• This research advances the development of lighter, more powerful, and sensorially capable robotic limbs.