参数化U形弯管流动数据集

本数据集由卡塞尔大学智能嵌入式系统部门创建，包含10,000个U形弯管的流体流动和热传递模拟数据。每个样本由28个设计参数描述，通过计算流体动力学方法处理。数据集内容丰富，包括设计参数和目标组合、五种不同分辨率的2D图像以及数值模拟的解变量，适用于设计优化领域的各种问题和方法研究。数据集的应用领域包括深度学习和设计优化，旨在通过多模态特征、转移学习等方法解决流体动力学中的优化问题。

This dataset was created by the Department of Intelligent Embedded Systems at the University of Kassel, and includes 10,000 sets of fluid flow and heat transfer simulation data for U-bends. Each sample is characterized by 28 design parameters, with all simulations conducted via Computational Fluid Dynamics (CFD) methods. The dataset is comprehensive, encompassing design parameters and their corresponding objective pairs, 2D images at five distinct resolutions, as well as solution variables derived from numerical simulations, making it suitable for research on various problems and methodological approaches in the design optimization domain. Its application areas cover deep learning and design optimization, with the goal of addressing optimization problems in fluid dynamics through techniques such as multimodal feature learning and transfer learning.