ITR: A New Data Model and Extensible Software for Predictive Chemical Kinetics

ITR：用于预测化学动力学的新数据模型和可扩展软件

• 批准号: 0312359
• 负责人: William Green
• 金额: $ 42.92万
• 依托单位: Massachusetts Institute of Technology
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2003
• 资助国家: 美国
• 起止时间: 2003-09-01 至 2006-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0312359&HistoricalAwards=false
• 关键词: ITR; New; Data; Model; Extensible

Project SummaryThis project develops a completely new data model for chemical kinetics and corresponding extensible software, including graphical user interfaces and appropriate examples, that make it feasible to predict the rates and products of complex chemical processes. Many of the difficult tasks now performed manually by kineticists are automated, including the determination of which intermediates and reactions should be included in a simulation (based on numerical significance). In the new data model, parameters are required only for each type of functional group, rather than for each chemical species, and the xml format encourages users to document information on the provenance and uncertainty of each parameter. Because there are many fewer types of functional groups than chemical species, the new data model is compact enough that a human can check all the parameters individually. Furthermore, the parameters are classified in a scheme that encourages comparisons between related functional groups and the identification of outliers. To jump-start adoption of the new data model and software, a large, well-documented set of rate and thermochemical parameters is provided, as well as software appropriate for modeling a wide range of thermal and oxidative chemistry of hydrocarbons; new software tools for manipulating, comparing, and analyzing large-scale chemical kinetic simulations; and many examples.Chemical kinetic simulations of real-world processes are both extremely valuable, and extremely data intensive. For example, a model for gasoline combustion chemistry widely used to help design low-emission engines involves over 20,000 numerical parameters as inputs, and even this large set is certainly incomplete. The existing data model for chemical kinetics, created more than 20 years ago, was never designed to handle this level of complexity. It is adequate for conventional small a posteriori kinetic models, but not appropriate for predicting the chemical kinetics of technologically important systems, especially those involving complex mixtures. Accurate, large-scale simulation of reacting mixtures requires a whole new paradigm for how engineers, chemists, and environmental scientists interact with complex chemistry modeling software, and how they document the many assumptions and uncertainties that underlie the simulations. By reducing the size of the required input, and making the connection between the transferable fundamental chemistry and simulation results more transparent, this new approach will make a priori chemical kinetic simulation accessible to a much broader range of scientists and engineers. This will revolutionize the field of kinetics, open up significant new possibilities in the design of chemical products and processes, and provide a firmer basis for business and regulatory decision-making. A variety of outreach and educational activities are included. To encourage current practitioners of kinetics to take advantage of the new IT technology, the new methodology is demonstrated at conferences and promoted using existing Web collaboratories. It is also necessary to change the way kinetics is taught to encourage new engineers and scientists to embrace the new paradigm; to this end, new teaching modules are developed to introduce chemical kinetics to undergraduate and graduate students in a variety of disciplines including chemistry, chemical engineering, mechanical engineering, and environmental science.

该项目开发了一个全新的化学动力学数据模型和相应的可扩展软件，包括图形用户界面和适当的示例，使其能够预测复杂化学过程的速率和产物。 许多现在由动力学家手动完成的困难任务都是自动化的，包括确定哪些中间体和反应应该包括在模拟中（基于数值意义）。在新的数据模型中，参数只需要用于每种类型的官能团，而不是每种化学物质，xml格式鼓励用户记录每个参数的来源和不确定性信息。 由于官能团的类型比化学物种少得多，新的数据模型足够紧凑，人类可以单独检查所有参数。 此外，参数分类的计划，鼓励相关功能组之间的比较和识别离群值。 为了迅速采用新的数据模型和软件，提供了大量的、有据可查的速率和热化学参数，以及适用于模拟烃的各种热化学和氧化化学的软件;用于操纵、比较和分析大规模化学动力学模拟的新软件工具;实际过程的化学动力学模拟是非常有价值的，并且是非常数据密集的。 例如，广泛用于帮助设计低排放发动机的汽油燃烧化学模型涉及超过20，000个数值参数作为输入，即使是这么大的一组也肯定是不完整的。 20多年前创建的现有化学动力学数据模型从未设计用于处理这种复杂程度。 它是足够的传统的小后验动力学模型，但不适合预测的化学动力学的技术重要的系统，特别是那些涉及复杂的混合物。 精确、大规模的反应混合物模拟需要一个全新的范式，来指导工程师、化学家和环境科学家如何与复杂的化学建模软件进行交互，以及他们如何记录模拟背后的许多假设和不确定性。通过减少所需输入的大小，并使可转移的基础化学和模拟结果之间的联系更加透明，这种新方法将使更广泛的科学家和工程师可以使用先验化学动力学模拟。 这将彻底改变动力学领域，为化学产品和工艺的设计开辟重要的新可能性，并为商业和监管决策提供更坚实的基础。 包括各种外联和教育活动。 为了鼓励目前的从业者的动力学，以利用新的IT技术，新的方法是在会议上展示，并促进使用现有的Web协作。 也有必要改变动力学的教学方式，以鼓励新的工程师和科学家接受新的范式;为此，开发了新的教学模块，向包括化学，化学工程，机械工程和环境科学在内的各种学科的本科生和研究生介绍化学动力学。