We developed a Python-based open-source software for scientific visualization of fluid dynamics datasets. This software is enabled by the VTK graphic rendering library and equipped with an interactive user interface built upon PyQt5. The concept of “scene module” is proposed to construct the rendered scene for graphic analysis. In the context of fluid dynamics, scene module is defined by associating a manifested geometry, such as a cut plane, an iso-surface, a set of streamlines and a vector field, with a set of visual attributes, including rendering scalar, colormap, and opacity. In the proposed framework, each scene module is processed by one single VTK pipeline, and a number of modules can be rendered simultaneously for multivariate graphic analysis. Beyond the VTK built-in functionality, additional implementation efforts are taken to enhance the annotating capability, improve the display of coordinate axes, and load input data of various formats. Two case studies are presented to demonstrate the capability of PyVT, especially the visualization of multivariate fluid dynamic datasets that typically possess complex and interdependent flow physics.

我们开发了一个基于 Python 的开源软件，用于流体动力学数据集的科学可视化。该软件由 VTK 图形渲染库启用，并配备了基于 PyQt5 构建的交互式用户界面。提出了“场景模块”的概念来构建渲染场景进行图形分析。在流体动力学的上下文中，场景模块的定义是通过将显示的几何图形（例如剖切面、等值面、一组流线和向量场）与一组视觉属性（包括渲染标量、颜色图、和不透明度。在所提出的框架中，每个场景模块都由一个单独的 VTK 管道处理，并且可以同时渲染多个模块以进行多变量图形分析。除了 VTK 内置功能，还采取了额外的实施措施来增强注释能力，改进坐标轴的显示，并加载各种格式的输入数据。两个案例研究展示了 PyVT 的能力，尤其是多元流体动力学数据集的可视化，这些数据集通常具有复杂且相互依赖的流动物理。