Linear extrapolation method based on multiple equiproportional models for thermal performance prediction of...

Area of Science:

• Thermal engineering
• Materials science
• Computational physics

Background:

• Growing demand for ultra-large photothermal (PT) and electrothermal devices necessitates efficient performance prediction.
• Traditional finite element method (FEM) models are computationally intensive for ultra-large arrays, limiting optimization.
• Periodic boundary conditions can introduce significant errors in simulating local heating on large periodic arrays.

Purpose of the Study:

• To develop a computationally efficient method for predicting the thermal performance of ultra-large periodic arrays.
• To overcome the limitations of traditional FEM for large-scale thermal simulations.
• To provide a facile and efficient prediction strategy for optimizing PT transducers and other thermal engineering applications.

Main Methods:

• Proposed a linear extrapolation method based on multiple equiproportional models (LEM-MEM).
• Constructed several reduced-size FEM models for simulation and extrapolation, avoiding direct modeling of ultra-large arrays.
• Fabricated and tested a PT transducer with over 4000 × 4000 pixels to validate the LEM-MEM predictions.

Main Results:

• LEM-MEM demonstrated high predictability for thermal performance in PT transducers.
• The maximum percentage error in average temperature prediction was within 5.22% across four different pixel patterns.
• The fabricated PT transducer exhibited a response time within 2 ms.

Conclusions:

• LEM-MEM offers a significant reduction in computation consumption for thermal performance prediction in ultra-large arrays.
• The method provides valuable design guidance for optimizing PT transducers and similar devices.
• LEM-MEM is applicable to a wide range of thermal engineering problems requiring efficient prediction strategies for large-scale systems.