Digital multidisciplinary analysis and design optimization platform for aeroderivative gas turbines

由于燃气轮机固有的系统复杂性，其设计是一项具有挑战性的任务。用于航空衍生气体 涡轮机(AGT)设计过程非常零散，严重依赖从以下项目获得的历史数据 以前的引擎。西门子加拿大AGT，领先的工业燃气轮机原始设备制造商 (OEM)，已确定当前划分的设计活动导致次优和不健壮 设计，而设计工程师将相当多的时间花在非附加值任务上。因此， 它制定了一项旨在支持高度集成的并行设计过程的技术战略 以提高其产品的性能和健壮性，从而最大化其竞争力，同时 将其财务风险降至最低。 该项目的目标是建立一个数字平台的模型和数据管理与集成 一套分析、设计和优化工具。提议的研究的新颖性之一是调整工具 集成来自软件工程的最佳实践和更改传播技术，以构建 坚实的计算环境基础，集成了功能数据管理、模型版本 控制、分析工具和优化算法可根据特定的工作流程进行定制 使用机器学习技术的不同工程团队。分析、设计和优化模块 数字平台的设计将包括关键静态和旋转部件的参数设计模型， 与生命周期相关的量化技术、多学科设计优化(MDO)方法和健壮 设计工具。

Gas turbine design is a challenging task because of inherent system complexity. For aero-derivative gas turbines (AGT) the design process is quite fragmented and relies strongly on historical data obtained from previous engines. Siemens Canada AGT, a leading industrial gas turbine original equipment manufacturer (OEM), has determined that currently compartmentalized design activities lead to sub-optimal and non-robust designs while design engineers spend a considerable amount of their time on non-added value tasks. Therefore, it has developed a technology strategy that aims at supporting a highly integrated concurrent design process in order to improve performance and robustness of its products, maximizing thus its competitiveness while minimizing its financial risks. The objective of this project is to build a digital platform for models and data management with an integrated suite of analysis, design, and optimization tools. One of the novelties of the proposed research is to adapt tool integration best practices and change propagation techniques from software engineering in order to build the solid foundation of a computational environment that integrates functional data management, model version control, analysis tools and optimization algorithms in a way that is customizable to the specific workflows of different engineering teams using machine learning techniques. The analysis, design and optimization modules of the digital platform will include parametric design models for critical static and rotary components, lifecycle-related quantification techniques, multidisciplinary design optimization (MDO) methods, and robust design tools.