Steady State 2D Regularized Lid-Driven Cavity Low-resolution Spatiotemporal Snapshots - Part 2

This data was generated using the pseudo-spectral DNS method outlined in Lee, Dowell, and Balajewicz CNSNS (2017).

Filenames denote Reynolds number. Each file contains a velocity snapshot matrix where each column is sampled at a different timestep. Original simulations had a factor of 100 higher temporal resolution than do these matrices; the snapshots provided here were spaced 0.1 nondimensional time units apart. Each column comprises both horizontal and vertical velocity data with the former stacked atop the latter.

A provided convertCavitySnapshot.m function converts a single cavity snapshot into non-stacked organization. cheb.m builds the x-vector (y is identical) and differentiation matrix D where dx(square_snapshot) = matmul(-square_snapshot,D.transpose) and dy(square_snapshot) = matmul(D,square_snapshot). plotSnapshot.m likewise plots a snapshot for user viewing.

All data in this repository is discretized as 130 points in both the x and the y directions, viz. the 'res' variable passed to convertCavitySnapshot.m would be 130.

本数据集采用Lee、Dowell与Balajewicz于2017年发表于《CNSNS》的伪谱直接数值模拟（pseudo-spectral DNS）方法生成。

文件名代表雷诺数（Reynolds number）。每个文件均包含一个速度快照矩阵，矩阵的每一列对应一个不同时间步的采样结果。原始模拟的时间分辨率较本数据集的矩阵高出100倍；本次提供的快照采样间隔为0.1无量纲时间单位。每一列同时包含水平与竖直方向的速度数据，其中水平速度数据堆叠于竖直速度数据之上。

配套提供的convertCavitySnapshot.m函数可将单个腔体快照转换为非堆叠的组织形式。cheb.m用于生成x向量（y向量与之完全一致）以及微分矩阵D，满足dx(方形快照) = 矩阵乘法(-方形快照, D的转置)，dy(方形快照) = 矩阵乘法(D, 方形快照)。plotSnapshot.m函数同样可用于绘制快照以供用户查看。

本仓库中的所有数据均在x与y方向上离散为130个采样点，即传入convertCavitySnapshot.m函数的'res'变量值应为130。