Augmented Tube Models for Blends of Star and Linear Polymers

非技术总结该奖项支持计算研究和教育，以开发快速、准确的聚合物流动行为模型。合成聚合物的性质可以通过改变其分子结构来调节。这使得聚合物可以用于从包装到汽车以及从织物到外科缝合线的各种应用。聚合物通常在液态或熔融状态下加工，其中分子结构对流动特性有很大影响。分子结构和流动特性之间的这种关系通常是用一种叫做管模型的理论来研究的。标准管模型的失败可以追溯到它简化了聚合物分子之间的“多体相互作用”的方式。这些相互作用由称为滑移链模型的模拟模型更准确地描述，不幸的是，该模型的计算成本很高。该项目旨在通过开发增强管模型，将管模型的优点(速度)与滑链模型的优点(精度)结合起来。如果成功，这项研究将为混合理论和模拟提供一个模板，预计这将对材料表征产生深远的影响，甚至在聚合物行业之外。该奖项还通过提供材料建模、马尔科夫链蒙特卡罗和统计分析方面的现代培训，支持佛罗里达州立大学的少数民族学生的教育。由研究团队开发的软件和数据集将公开提供。技术总结该奖项支持计算研究和教育，以开发快速、准确的聚合物流动行为模型。管模型是研究聚合物熔体纠缠的一种流行的平均场理论。不幸的是，即使对于相对简单的体系，如松弛时间相差很远的单分散聚合物的二元混合物，它也始终失败。另一方面，对于这样的系统，不同的滑链模型都非常成功，它们都明确地将链之间的纠缠视为滑链。这主要是因为在这两种模型中，对一种称为约束释放的松弛机制的处理方式不同。在滑链模型中，约束释放的描述是“自然的”，而在管模型中，约束释放可能很复杂，并且仍然不准确。滑移环节模型的一个缺点是其计算成本。该项目寻求将管模型(速度)的强度与滑移链模型的强度(约束释放的自然和准确表示)相结合。主要的研究目标是(I)探索结合这些优点的增强管模型的发展，以及(Ii)在根据流变学的实验测量推断聚合物混合物的组成的计算密集型反问题上测试这些增强型模型。该奖项支持开发一种名为ecoSLM的快速滑动链接模型，并使用星形和线性聚合物的二元混合物对其进行验证。服务于总体目标的研究活动被分成四个具体的任务：(I)根据线性-线性、星形-线性和星形-星形混合物的所有可用实验数据来验证EcoSLM，(Ii)使用经过验证的EcoSLM来绘制设计空间，以帮助识别管模型失败的区域，(Iii)开发扩展管模型，它使用EcoSLM来引导管模型到正确的物理/参数，以及(Iv)使用这些扩展管模型进行反向建模。该项目探索了平均场理论与随机模拟模型ecoSLM的增强，以便将速度和精度结合起来。如果成功，它可以为材料科学中寻求将平均场理论与简单模拟模型相结合的其他问题提供一个模板。使用贝叶斯框架进行逆向建模，这可能对材料表征产生深远影响，甚至在聚合物行业之外。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

NONTECHNICAL SUMMARYThis award supports computational research and education into the development of fast and accurate models of flow behavior in polymers. Properties of synthetic polymers can be tuned by engineering their molecular structure. This allows polymers to be used in a variety of applications ranging from packaging to automobiles, and from fabrics to surgical sutures. Polymers are typically processed in the liquid or melt state, where the molecular structure strongly influences flow properties. This relationship between molecular structure and flow properties is often studied using a theory called the tube model. Failures of the standard tube model can be traced to the manner in which it simplifies "multibody interactions" between polymer molecules. These interactions are more accurately described by a simulation model called the slip link model, which is unfortunately computationally costly. This project seeks to combine the strength of tube model (speed) with that of the slip link model (accuracy) through the development of augmented tube models. If successful, this research will provide a template for blending theories and simulations, which is expected have far reaching consequences for materials characterization, even outside the polymer industry.The award also supports the education of minority students at Florida State University by providing modern training in materials modeling, Markov chain Monte Carlo, and statistical analysis. The software and datasets developed by the research team will be made publicly available.TECHNICAL SUMMARYThis award supports computational research and education into the development of fast and accurate models of flow behavior in polymers. The tube model is a popular mean-field theory for entangled polymer melts. Unfortunately, it fails consistently even for relatively simple systems like binary blends of monodisperse polymers with widely separated relaxation times. On the other hand, different slip link models, all of which explicitly treat entanglements between chains as slip links, are remarkably successful for such systems. This is primarily due to the different treatment of a relaxation mechanism called constraint release, in the two models. In slip link models, the description of constraint release is "natural", while it can be complicated, and still inaccurate, in the tube model. A drawback of slip link models is their computational cost. This project seeks to combine the strength of tube models (speed) with that of the slip link models (natural and accurate representation of constraint release). The overarching research objectives are (i) to explore the development of augmented tube models, which combine these strengths, and (ii) to test these augmented models on the computationally intensive inverse problem of inferring the composition of a polymer mixture from experimental measurements of rheology. The award supports the development of a fast slip link model called ecoSLM, and its validation, using binary blends of star and linear polymers. Research activities to service the overall objectives are broken down into four specific tasks: (i) validating the ecoSLM against all available experimental data on linear-linear, star-linear, and star-star blends, (ii) using the validated ecoSLM to map out the design space to help identify regions where the tube model fails, (iii) developing augmented tube model, which uses ecoSLM to guide the tube model to the correct physics/parameters, and (iv) to use these augmented tube models for inverse modeling.The project explores the augmentation of a mean-field theory with a stochastic simulation model, ecoSLM, in order to combine speed with accuracy. If successful, it can provide a template for other problems in materials science that seek to combine a mean-field theory with simple simulation models. Using a Bayesian framework for inverse modeling, this can have far reaching implications for materials characterization, even outside the polymer industry.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.