Kinetic Multiscale Modeling and Simulation of Cluster Formation Processes in Supersonic Flows

团簇形成过程包罗万象，在对社会重要的无数应用中发现，例如医学研究、创造新的和新颖的材料以及开发低成本的基于空间的全球电信。最近的研究表明，沉积特定尺寸的团簇开辟了一类新的表面纳米结构制造材料，具有特定尺寸的催化性能。此外，为了减少在脉冲激光烧蚀沉积薄膜中团簇的形成，必须控制靶表面附近团簇的尺寸。这些只是许多例子中的两个，它们证明了开发一种理论和模拟工具的重要性，这种工具能够模拟团簇形成和演化的非平衡空间分布，以推进新的制造技术。 预测超音速反应控制喷流中的团簇大小分布和速度，将使太阳能电池阵列等敏感表面的智能放置成为可能，从而降低天基电信的成本。用于医疗外科手术的新材料正在通过簇形成过程开发，以在战略位置创建药物的纯微观植入物。显然，上述应用具有许多重要的全球性社会影响。 在这些工程应用中的气相中的团簇冷凝的预测建模目前提出了一个艰巨的任务，由于在一些非平衡物理化学过程的基本速率的不确定性，涉及团簇的演变和流动长度尺度的大的可变性。因此，一个合乎逻辑的步骤，在模拟工具的发展是考虑更简单的现象，超音速射流膨胀中的集群的形成和增长的过程中，只耦合非平衡气体动力学。超音速射流中团簇行为的成功建模将为未来涉及团簇演化的更复杂技术的建模提供原型。 成功的非均匀两相流系统的建模也将使我们能够建立一个相对重要的个人基本cross-sections.本研究的主题是定量表征的集群formedby均匀成核在超音速射流中，在非平衡空间分布的集群大小和内部和动能。本文提出了一个多尺度计算模型来模拟超音速射流中团簇的形成和演化过程，该模型基于一种适用于稀薄过渡流态的粒子模拟方法--直接模拟蒙特卡罗（DSMC）。该模型是多尺度的，因为除了DSMC，连续流体动力学（CFD）/Navier-Stokes（NS）和分子动力学（MD）配方也将被使用。计算流体力学的Navier-Stokes方法将被用来模拟稠密气体的初始膨胀阶段，在此期间冷凝可以忽略。分子动力学将主要用于建立DSMC的团簇反应截面数据库，并进一步扩展DSMC的模拟能力，以解决非二元碰撞凝聚的相关问题，发展一种统计有效的方法来估计反应截面和碰撞后结果。这项工作的主要重点涉及系统描述的pair-wisepotential，因为多尺度建模和模拟的可信度必须首先建立众所周知的，简单的气体。瑞利散射数据集不仅提供平均集群数量，但集群分布以及，将用于模拟验证。这些数据一直未得到充分利用，是建模社区的重要资源。该项目将促进发现和理解，同时促进教学。将开设一门关于稀薄流体计算模型及其在材料加工中的应用的研究生课程。本课程将涉及MD-DSMC方法的学习和使用，以在拟议的研究中发展。更具体地说，在学习本课程后，航空航天和机械工程以及工程科学的学生将能够模拟具有已知相互作用势的化学系统的耦合稀薄冷凝流。这门课是美国为数不多的要求学生进行DSMC模拟稀薄气体流对航天器影响的课堂项目的课程之一。本科生和研究生将参加拟议的学术活动，通过宾夕法尼亚州立大学的各种项目和同事的直接邀请，与少数学生接触。拟议的研究结合了稀薄气体动力学，物理化学，大规模并行计算，材料科学和统计学的元素，因此，是多学科的。

Cluster formation processes are all encompassing and found in countless applications important to societymedical research, creation of new and novel materials, and developing low cost space-based world widetelecommunications, to name a few. It has been shown recently that deposition of size-specific clustersopens up a new class of surface nanostructure fabrication of materials that have size-specific catalyticproperties. Additionally, to reduce cluster formation in pulsed laser ablation deposition of thin films thesize of clusters in the vicinity of a target surface must be controlled. These are just two of many examplesthat demonstrate the importance of developing a theory and simulation tool capable of modeling nonequilibrium spatial distributions of cluster formation and evolution to advance new fabrication technologies. Predicting cluster size distributions and velocities in supersonic reaction control jets will reduce the cost of space-based telecommunications by allowing intelligent placement of sensitive surfaces such as solar cell arrays. Novel materials for medical surgical procedures are being developed by cluster formation processes to create pure microscopic implants of pharmaceuticals in strategic locations. Clearly the above applications have many important global, societal implications. The predictive modeling of cluster condensation in the gas phase in these engineering applications currently presents a formidable task due to the uncertainties in some of the fundamental rates for nonequilibrium physical chemical processes involving the evolution of clusters and the large variability of flow length scales. Therefore, a logical step in the development of simulation tools is to consider the simpler phenomenon of supersonic jet expansion in which the process of cluster formation and growth is coupled only with non-equilibrium gas dynamics. Successful modeling of cluster behavior in supersonic jets will serve as a prototype for future modeling of more complex technologies involving the evolution of clusters. Successful modeling of a non-homogenous two-phase flow system will also allow us to establish the relative importance of individual fundamental cross sections.The subject of this proposed research is the quantitative characterization of clusters formedby homogeneous nucleation in a supersonic jet, in terms of non-equilibrium spatial distributionsof cluster size and internal and kinetic energies. We propose to model the processof cluster formation and evolution in supersonic jets by a multiscale computational model.The proposed model is based on a kinetic particle simulation method, the direct simulation Monte Carlo(DSMC), which is applicable in the transitional to rarefied flow regime. The proposed model is multiscalebecause, in addition to DSMC, continuum fluid dynamics (CFD)/Navier-Stokes (NS) and molecular dynamics(MD) formulations will also be used. The CFD Navier-Stokes approach will be used to simulate theinitial expansion stage of the dense gas, during which the condensation can be neglected. Molecular dynamicswill be used primarily to develop a cluster reaction cross section data base for DSMC and to extend theDSMC simulation capabilities further to address issues related to condensation due to non-binary collisions.A statistically efficient approach will be developed for estimation of both the reaction cross sections andpost-collisional outcomes. The primary emphasis of this work involves systems described by a pair-wisepotential, since multi-scale modeling and simulation credibility must first be established for well known,simple gases. Rayleigh scattering data sets providing not only average cluster quantities, but cluster distributions as well, will be used for simulation validation. These data have been under-utilized and represent an important resource for the modeling community.The project will advance discovery and understanding while promoting teaching and learning. A newgraduate course on computational modeling of rarefied flows with applications to materials processing willbe developed. The course will involve the learning and usage of the MD-DSMC methodology to be developedin the proposed research. More specifically, after taking this course Aerospace and Mechanical Engineeringand Engineering Science students will be able to simulate coupled rarefied-condensation flows for chemicalsystems with known interaction potentials. This course is one of the very few in the nation where studentsperform a class project requiring the use of DSMC for modeling the effects of rarefied gas flows on spacecraft.Undergraduate and graduate students will participate in the proposed schololarly activities with outreach tominority students through various programs at Penn State and direct solicitation of colleagues. The proposedresearch combines elements of rarefied gas dynamics, physical chemistry, massively parallel computing,materials science, and statistics and is, therefore, multidisciplinary.