Session: 06-05 Implants and Biomedicine

Paper Number: 167631

167631 - Finite Element Simulation for the Design of Smart Piezoelectric Force Sensing Total Knee Replacement 

Introduction: In vivo load-sensing smart knee replacements offer a means for tracking postoperative joint contact force magnitudes and locations, enabling better assessment of patient outcomes, refinement of surgical techniques, and validation of biomechanical models. For widespread clinical adoption, it is critical that such sensing technology be implemented in a platform-agnostic manner to accommodate diverse implant designs and geometries. This requires an optimization framework capable of tailoring implant-specific sensor configurations to minimize error between sensed and ground truth joint kinetics and kinematics. A key component of such a framework, and the topic of this study, is the development of finite element (FE) simulations that resolve implant-specific tibiofemoral contact mechanics and predict the output of embedded load sensors under realistic, six-degree-of-freedom (6-DoF) loading. The resulting contact mechanics and predicted sensor output enable sensor optimization based on realistic tibiofemoral interactions.

Methods: In this study, we develop finite element models of a commercially available cruciate retaining (CR) total knee replacement system (Triathlon, Stryker, Mahwah, NJ, USA) with integrated piezoelectric sensors for compartmental load and tibiofemoral contact location sensing. Implant geometries are obtained through 3D scanning and are imported into ANSYS (ANSYS, Inc., Canonsburg, PA, USA) for FE simulation. The femoral component is modeled as a rigid body, and the tibial component and piezoelectric sensors are assigned the material properties of ultra-high molecular weight polyethylene (UHMWPE) and PZT-5A (APC 850, APC International, Ltd., Mackeyville, PA, USA), respectively. The FE model resolves simulated ground truth tibiofemoral contact mechanics under 6-DoF loading similar to a normal gait cycle and comprising femoral flexion, axial compression, anterior-posterior (AP) loading, and internal-external (IE) torque. Loading and motion constraints are consistent with ISO 14243-1: Implants for Surgery — Wear of Total Knee-joint Prostheses. The model also resolves loads transferred to the piezoelectric sensing elements – the sum of which in each compartment gives the sensed medial and lateral joint loads. The sensed loads are then used to estimate contact locations in each compartment via an equilibrium of moments. A total of 20 uniformly distributed load steps are simulated across the gait cycle. Flexion and axial load are applied to the femoral component, while AP load and IE torque are applied to the tibial bearing. To achieve equilibrium, medial-lateral (ML) and abduction-adduction (AA) motions are handled as reactionary degrees of freedom by the tibial and femoral components, respectively.

Results: Pressure distributions across the tibial contact surface reveal expected biomechanical features such as IE rotation during the stance-to-swing transition and femoral rollback during flexion. Sensed medial and lateral loads display high accuracy, with slight under- and overprediction during the stance phase, respectively. Error in total sensed axial load remains <= 0.5% at all points. Localization error remains < 2 mm and 2.5 mm in the medial and lateral compartments, respectively. The results confirm that loads are transferred appropriately and are meaningfully captured by the piezoelectric sensors, and that sensing and localization accuracy can be quantified, thereby establishing an FE modeling framework that enables future optimization of the piezoelectric sensing system to minimize localization error.

Conclusion: The FE simulation successfully captures reasonable tibiofemoral contact mechanics under realistic 6-DoF loading, and demonstrates the ability to assess the accuracy of the piezoelectric sensing system. The resulting model framework provides a critical foundation for implant-specific optimization of sensor configurations in platform-agnostic smart knee replacements. Future work will involve empirical validation of the developed modeling framework.

Presenting Author: Brandon Hines Tennessee Tech University

Presenting Author Biography: I am a Ph.D. candidate in Mechanical Engineering specializing in dynamic and smart systems. My interests include classical and biomechanical applications of dynamic modeling, vibration analysis, finite element analysis, structural health monitoring, smart materials and systems, and machine learning.