Scholarlink

<p>Enhancing the β-phase content in polyvinylidene fluoride (PVDF) composites is crucial for improving the piezoelectric performance of sensors and actuators. However, conventional optimization approaches depend on repeated experimental trials, which are costly, time-consuming, and often limited in their ability to capture complex parameter interactions. To overcome these challenges, a Gaussian Process (GP) surrogate model was implemented to predict β-phase content in PVDF/barium titanate (BTO)/multi-wall carbon nanotubes (MWCNTs) composites via phase immersion-direct ink writing (PI-DIW). The GP framework was chosen for its ability to model non-linear responses, quantify prediction uncertainty, and guide efficient exploration of the design space. The surrogate model identified optimal conditions of 26.62 wt.% BTO, 7 wt.% MWCNT, 103℃, and 7.45&#xa0;kV/mm. Under these parameters, the composites achieved an experimentally validated β-phase fraction of 84.44% with 97.6% model accuracy. This study demonstrates a novel integration of additive manufacturing and machine learning optimization, providing a powerful framework for accelerating the design of high-performance piezoelectric polymers.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of the beta phase content of PVDF/BTO/MWCNT composites fabricated by phase immersion-direct Ink writing

  • 期刊论文

  • 发表日期：

    2025年10月18日

查看全文
  • Aaron Rodriguez,
  • Abdiel Cruz,
  • Stephanie Gonzalez,
  • Stephanie Vargas,
  • Alexis Lopez,
  • Sabina Arroyo,
  • Yirong Lin,
  • Anabel Renteria

摘要

Enhancing the β-phase content in polyvinylidene fluoride (PVDF) composites is crucial for improving the piezoelectric performance of sensors and actuators. However, conventional optimization approaches depend on repeated experimental trials, which are costly, time-consuming, and often limited in their ability to capture complex parameter interactions. To overcome these challenges, a Gaussian Process (GP) surrogate model was implemented to predict β-phase content in PVDF/barium titanate (BTO)/multi-wall carbon nanotubes (MWCNTs) composites via phase immersion-direct ink writing (PI-DIW). The GP framework was chosen for its ability to model non-linear responses, quantify prediction uncertainty, and guide efficient exploration of the design space. The surrogate model identified optimal conditions of 26.62 wt.% BTO, 7 wt.% MWCNT, 103℃, and 7.45 kV/mm. Under these parameters, the composites achieved an experimentally validated β-phase fraction of 84.44% with 97.6% model accuracy. This study demonstrates a novel integration of additive manufacturing and machine learning optimization, providing a powerful framework for accelerating the design of high-performance piezoelectric polymers.