Design and Evaluation of Magnetic Hall Effect Tactile Sensors for Use in Sensorized Splints

Dominic Jones, Lefan Wang, Ali Ghanbari, Vasiliki Vardakastani, Angela E. Kedgley, Matthew D. Gardiner, Tonia L. Vincent, Peter R. Culmer, Ali Alazmani2020 · DOI

TL;DR

This paper presents a novel design and evaluation of magnetic Hall effect tactile sensors for integration into medical splints. By combining finite element simulations with experimental validation, the researchers developed a compact, multi-axis sensor capable of measuring both normal and shear forces at the splint-skin interface, offering a significant advancement for monitoring and improving splinting treatments.

What Problem Does This Paper Address?

Splints are widely used in medicine to immobilize and offload arthritic joints, aiding in pain reduction and recovery. However, the effectiveness of different splinting techniques and their biomechanical impact remain poorly understood due to a lack of detailed force and pressure measurements at the splint-skin interface. Existing methods often rely on single-axis sensors or are too bulky to integrate seamlessly. This paper addresses the need for a compact, multi-axis tactile sensor that can accurately measure forces and pressures experienced at this interface, providing crucial data to inform clinical practice and optimize splint design for improved patient outcomes.

Methods

The researchers employed a multi-faceted approach combining finite element (FE) simulations with experimental validation.

  • Magnetic Field Simulation: FE models were used to simulate the magnetic field generated by different magnet geometries (radius and thickness). This helped identify an "effective workspace" for the sensor, defining the area where the Hall effect sensor could reliably detect magnetic field changes.
  • Mechanical Simulation: FE models were also used to analyze the mechanical response of hyperelastic elastomer materials under various loading conditions. This allowed for the optimization of elastomer geometry (thickness and radius) to achieve desired force ranges and sensitivities.
  • Sensor Design Optimization: By integrating the insights from both magnetic and mechanical simulations, the researchers iteratively refined the sensor design parameters, focusing on magnet size, elastomer geometry, and their interaction.
  • Prototype Fabrication: A prototype sensor was manufactured based on the optimized design, utilizing Ecoflex 00-30 elastomer and N42 neodymium magnets.
  • Experimental Validation: The prototype sensor was mechanically loaded using a custom three-axis loading platform. Its performance was compared against a commercial load cell (Nano17) to evaluate its accuracy in measuring normal and shear forces.
  • Calibration: A neural network (two-layer feed-forward network with a hyperbolic tangent activation function) was trained using a Levenberg-Marquardt backpropagation algorithm to characterize the non-linear relationship between the measured magnetic field and the applied load.
  • Splint Integration: Four calibrated sensors were integrated into a commercially available hand splint to demonstrate their practical application in a real-world scenario.

Datasets

Not discussed in the paper. The paper focuses on simulation and experimental data generated by the researchers themselves.

Key Insights and Ideas

  • Magnet Geometry is Crucial for Measurable Area: The radius of the magnet has a significantly greater impact on the size and shape of the "measurable area" (the region where the Hall effect sensor can detect magnetic field changes) than its thickness. This allows for initial selection of the magnet radius to define the required range of motion, followed by thickness adjustment to position the sensor's working area.
  • Optimizing Magnet-Sensor Distance for Full Range: Placing the magnet such that its minimum distance to the Hall effect sensor (under maximum normal force) aligns with the sensor's saturation point ensures the entire sensor range is utilized at the highest possible resolution.
  • Elastomer Geometry Dictates Working Area and Force/Shear Ratio: The elastomer's geometry is key to defining the sensor's "working area" (the range of magnet movement during normal use). The ratio of elastomer thickness to radius directly influences the normal-to-shear pressure ratio, offering a pathway to optimize for specific shear interaction requirements in splinting.
  • Simulation Accelerates Design: The use of FE simulations significantly accelerated the sensor design process, allowing for a detailed exploration of the design space and selection of appropriate parameters to meet system requirements before physical prototyping.
  • Multi-axis Sensing is Achievable and Necessary: The developed sensor successfully measures forces in three axes, which is critical for understanding the complex interactions at the splint-skin interface, where both normal pressure and shear forces play important roles.

Key Results

  • Sensor Design: A 4 mm radius, 3 mm thick sensor with a 0.5 mm thick elastomer layer above the magnet was identified as meeting defined sensing requirements.
  • Pressure Range: The prototype sensor exhibited a pressure range of 45 kPa in the normal direction and 6 kPa in the shear direction.
  • Simulation vs. Experiment: Experimental validation showed consistency with simulated results, with mean normal and shear pressure ranges of 44.82 ± 0.08 kPa and 6.11 ± 0.03 kPa, respectively. This represented a difference of +2.88% for normal and +5.31% for shear compared to simulations.
  • Calibration Accuracy: The calibrated sensor demonstrated comparable accuracy to a commercial load cell, with Root Mean Square (RMS) errors of 0.36 (11%) for the x-axis, 0.15 (4.5%) for the y-axis, and 1.35 (4.8%) for the z-axis.
  • Splint Integration: When integrated into a hand splint, the sensors measured pressures of around 20 kPa and shear stresses of around 5 kPa during hand closing actions, consistent with literature values.

Limitations and Future Work

  • Elastomer Viscoelasticity: The paper acknowledges that viscoelastic effects (creep and hysteresis) in the elastomer can introduce inconsistencies in the calibration, particularly in the x-axis. Further refinement of the calibration methodology is needed to mitigate these effects.
  • Sensor Size: The current prototype sensors are relatively large. Future work will focus on miniaturizing the system by optimizing sensing electronics and the elastomeric layer.
  • Shear Loading Regimes: The understanding of typical shear loading regimes at the splint-skin interface is currently limited. Preliminary testing is planned to gain a better understanding of these regimes, which will inform future sensor design for improved shear measurement.
  • Broader Applications: While the study focuses on splinting, the developed sensor technology has potential applications in other medical fields, such as grip sensing and other orthotic interface measurements.

Summary updated

BibTeX
@misc{jones2020design,
  author = {Dominic Jones and Lefan Wang and Ali Ghanbari and Vasiliki Vardakastani and Angela E. Kedgley and Matthew D. Gardiner and Tonia L. Vincent and Peter R. Culmer and Ali Alazmani},
  title = {{Design and Evaluation of Magnetic Hall Effect Tactile Sensors for Use in Sensorized Splints}},
  year = {2020},
  doi = {10.3390/s20041123},
}

Loading related content...

Figure