Summary:
UCLA researchers in the Department of Mechanical Engineering have developed a novel bidirectional thermal linear transducer for active vibration dampening and high precision manufacturing equipment.
Background:
Metamaterials are engineered structures that demonstrate properties not commonly found in natural materials. To date, most research focuses on passive metamaterial, or materials with predefined characteristics tailored to specific applications. However, passive metamaterials are limited in that they cannot dynamically respond to external forces. In structural applications such as vibration dampening, they are typically optimized for a narrow frequency band, dissipating energy through their intrinsic properties. As a result, multiple customized solutions may be required to address different frequency ranges, increasing complexity and limiting adaptability.
In contrast, active metamaterials are capable of sensing and responding to external stimuli by adjusting their stiffness and geometry in real time. This adaptability enables them to more effectively mitigate stress and strain in structural systems. A promising approach to achieving this responsiveness involves mechanical neural networks (MNNs)—actuator-driven systems inspired by artificial neural networks (ANNs). MNNs enable dynamic control of metamaterial properties, paving the way for more versatile and resilient structures. Thus, there is a growing need for active metamaterial solutions that go beyond static design to offer intelligent, tunable responses in real-world environments.
Innovation:
Researchers at UCLA have designed and fabricated a mechanical neural network powered by meso-scale thermal linear transducers. These transducers feature a flexible geometry that enables reliable, controllable linear motion—an essential function for tuning the structure of active metamaterials The system incorporates onboard circuitry that modulates thermal input to actuate the transducers, offering a compact and scalable method of actuation and mechanical control. Additionally, integrated sensors enable active monitoring of the transducer, allowing metamaterials to dynamically respond to external stimuli such as vibration, stress, or temperature changes. This novel thermal linear transducer system overcomes the static limitations of passive metamaterials by enabling adaptive behavior through an MNN architecture. This innovation significantly enhances the functionality and versatility of metamaterials. . Potential applications span smart building systems, advanced aerospace components, and resilient defense structures, where adaptable materials are critical for performance and safety.
Potential Applications:
● Active Vibration Dampening & high precision manufacturing equipment
○ Semiconductor Manufacturing
○ Aerospace/Defense
○ Medical devices
■ Surgical equipment/robotics
○ Automotive
○ Research Equipment
○ Wind turbine and energy infrastructure
○ Solar panel design
○ Smart Insulation
○ Seismic retrofitting infrastructure
○ Environmental monitoring and response systems
Advantages:
● Active Metamaterials
○ Increased functionality
○ Closed loop control
● Compact
● Scalable
● Meso-scale
Development-To-Date: Fabrication of invention complete.
Related Papers:
[1] JMEMS Letters. 1pt Response Speed Characterization of a Thermally Actuated Programmable Metamaterial C Luo, JB Hopkins, MA Cullinan - Journal of Microelectromechanical Systems, 2023. https://colab.ws/articles/10.1109%2Fjmems.2023.3332595
[2] Desired Stiffness Verification on Programmable MEMS Metamaterial C Luo, J Hopkins, M Cullinan - 2024 IEEE 37th International Conference on Micro …, 2024. https://www.researchgate.net/publication/378408528_Desired_Stiffness_Verification_on_Programmable_MEMS_Metamaterial
[3] Design and fabrication of a three-dimensional meso-sized robotic metamaterial with actively controlled properties C Luo, Y Song, C Zhao, S Thirumalai, I Ladner… - Materials Horizons, 2020 https://pubs.rsc.org/en/content/articlelanding/2020/mh/c9mh01368g
[4] Lee, Ryan H., Erwin AB Mulder, and Jonathan B. Hopkins. "Mechanical neural networks: Architected materials that learn behaviors." Science Robotics 7.71 (2022): eabq7278. https://www.science.org/doi/10.1126/scirobotics.abq7278
[5] Mechanical Neural Networks: architected materials that learn behaviors, Jonathan Brigham Hopkins, Erwin A.B. Mulder, Ryan Hansen Lee, US20240028882A1
Reference:
UCLA Case No. 2024-235
Lead Inventors:
Jonathan B. Hopkins, Pietro Sainaghi