|
s********5 发帖数: 73 | 2 我正在准备NIW的申请,有四篇被引用文章找不到,还请有此资源的大侠帮忙下载.多谢!
1. A New Numerical Approach for Predicting the Two-Phase Flow of
Refrigerants during Evaporation and Condensation, (I Fayssal, F Moukalled -
Numerical Heat Transfer Part B, 2012)
2. Eulerian-Eulerian modelling and computational fluid dynamics simulation
of wire mesh demisters in MSF plants, (H Al-Fulaij, A Cipollina, G Micale…
- Engineering Computations, 2014)
3. Performance Increase of Steam Turbine Condensers by CFD Analysis, (S
Hecker, A Auge, T Ellse... 阅读全帖 |
|
s********5 发帖数: 73 | 3 文章信息如下:
paper title: "Pressure drop across wire mesh demister in desalination plants
using Eulerian–Eulerian modeling and computational fluid dynamics
simulation"
Authors: D Al-Rabiah, H Al-Fulaij, H Ettouney,
Journal: Desalination and Water Treatment, 2015, Taylor & Francis
Paper link: http://www.tandfonline.com/doi/abs/10.1080/19443994.2015.1102774
若有朋友能找到下载请发到邮箱: [email protected]
/* */
多谢! |
|
w***n 发帖数: 1084 | 4 Eulerian or Lagrangian?
Eulerian can run on GPU. Lagrangian will be more difficult.
Memory is usually not the bottleneck. Multithreading is very useful. |
|
k*********g 发帖数: 791 | 5 in classical rigid body newtonian mechanics, and in solid mechanics, people
use: lagrangian framework for governing equations;
in fluid mechanics, popele use: eulerian framework for governing equations;
however, it turns out both are subsets & can be united;
the superset is: arbitrary lagrangian eulerian framewrok (ALE);
so, up to here, ALE has nothing to do with moving boundaries & free-surface
flows;
ALE is about how to formulate your equations;
level set is about how to capture moving interfa |
|
x********i 发帖数: 905 | 6 http://iccm.mcm.ac.cn/dct/page/1
Invited Lectures
Group 1
Fan Qin: Cluster algebras and monoidal categorification
Fang Li: Positivity of acyclic sign-skew-symmetric cluster algebras via
unfolding method and some related topics
Cheng-Chiang Tsai: An attempt for affine Springer theory
Li Cai: The Gross-Zagier formula: arithmetic applications
Ming-Hsuan Kang: Geometric zeta functions on reductive groups over non-
archimedean local fields
Huanchen Bao: Canonical bases arising... 阅读全帖 |
|
s***h 发帖数: 592 | 7 your idea of eulerian strain is wrong
strain is related to time?
that is the strain rate definition.
define the displacement gradient Ui,j and Uj,i
lagrange strain is Eij=(Ui,j+Uj,i+Ui,k*Uk,j)/2
eulerian strain is eij=(Ui,j+Uj,i-Ui,k*Uk,j)/2
so if we neglect high order component Ui,k*Uk,j
those two are same at small strain case
at FINITE strain case, Ui,k*Uk,j cannot be neglected |
|
L****8 发帖数: 3938 | 8 【 以下文字转载自 Programming 讨论区 】
发信人: Liber8 (Space-Time continuum), 信区: Programming
标 题: 100伪币悬赏:CNN这个东西本质上处理不了形变
发信站: BBS 未名空间站 (Wed Nov 22 00:14:51 2017, 美东)
物体大形变 只能通过 data augmentation 进行题海战术 记住所有变化
物体形变 本身是连续变化 Lagrangian view
但是变成图像 用像素描述物体 就是 Eulerian view,所有的基本距离度量,在像素
空间内统统失效。
举个例子: 在MNIST数据集上
很多基于deformable model的算法 可以获得非常高的准确度 用很少的训练样本 完全
不用data augmentation
如果用DNN,那就非得data augmentation 才能获胜
100伪币悬赏 ---------------------------------------------
图像是一个三角形的瀑布从上而下流
两个任务
1)请设计一个多入单出的DNN regre... 阅读全帖 |
|
f*****e 发帖数: 2992 | 9 找一个eulerian path。balanced就guarantee了。 |
|
f*****e 发帖数: 2992 | 10 Eulerian traversal有现成的非常有趣的算法,那本combinatorial generation书里有
,虽然写的乱七八糟,不过还是有部分价值。不过这题好像是经过每个node一次,不是
经过每条edge一次。 |
|
l***i 发帖数: 1309 | 11 you want hamitonian path, not eulerian path, shortest superstring is NP-hard |
|
j*********g 发帖数: 36 | 12 This is not gray code. No need to formulate it as TSP. Also, debrujin
graph only gives you a problem formulation.
You can solve by finding an Eulerian tour. Each node is a 3-digit number eg
100, and two nodes 100 and 001 are adjacent if they can be combined into
1001. Now every node has equal in-degree and out-degree, thus an Eulerien
tour exists and can be found using the standard algorithm. |
|
j*********g 发帖数: 36 | 13 Eulerian tour没有走多。看了我写的方法?关键是让每个edge是四位数,而不是node |
|
j*********g 发帖数: 36 | 14
对的,你的backtrack work就是因为Eulerian tour存在 |
|
c******w 发帖数: 1108 | 15 定义1000个node的有向图,每个node代表三位数字。每个node有10条outgoing edge,
对应0-9。每个node加上它的一条outgoing edge代表一组四位密码。
题目的最优解就是能遍历该有向图所有*edge*的最短的path。实际上就是要解这个图上的
chinese postman problem。
https://simple.m.wikipedia.org/wiki/Chinese_postman_problem
因为该有向图里每个node的in-degree和out-degree都等于10,其最优解为Eulerian
circuit :visits every edge exactly once。可以在O(|E|)内找到。 |
|
p*****g 发帖数: 604 | 16 请问lagrangian strain E和eulerian strain e之间如何转换?
在离散的情况下,是不是有以下的关系?
E=(L-L_0)/L_0
e=(L_0-L)/L
假如我用的是ln的形式,
E=ln(L/L_0),e=ln(L_0/L),他们是不是就只差一个正负号?
对于2维的E和e,如何转换呢? |
|
L****8 发帖数: 3938 | 17 物体大形变 只能通过 data augmentation 进行题海战术 记住所有变化
物体形变 本身是连续变化 Lagrangian view
但是变成图像 用像素描述物体 就是 Eulerian view,所有的基本距离度量,在像素
空间内统统失效。
举个例子: 在MNIST数据集上
很多基于deformable model的算法 可以获得非常高的准确度 用很少的训练样本 完全
不用data augmentation
如果用DNN,那就非得data augmentation 才能获胜
100伪币悬赏 ---------------------------------------------
图像是一个三角形的瀑布从上而下流
两个任务
1)请设计一个多入单出的DNN regressor,从图片计算瀑布下尖位置(决定整个形状)
2)请设计一个generative DNN, 根据下尖位置生成瀑布图片
请用前64张图训练 后64张图测试
不用data augmentation 不用transfer learning
能做出来的 我出100伪币 估计google会给你100w美元 |
|
k*********g 发帖数: 791 | 18 透露一个小道消息:
目前湍流的计算方法有:
1 DNS, direct numerical simulation;
2 LES, large eddy simulation;
3 RANS, reynolds average navier stokes;
4 PDF, probability density function;
以上是eulerian approachs;
还有:
5 lagrangian approach
于是,有人以为整个story is complete了,人生完整了;
thats wrongk; |
|
j****x 发帖数: 943 | 19 初学CFD, 正在研究 Free Surface Flow, 哪位高人给介绍一下有关 Arbitrary
Lagrangian-Eulerian 和 Level-Set 方面的资料? 谢了先!! |
|
p*****g 发帖数: 604 | 20 请问lagrangian strain E和eulerian strain e之间如何转换?
在离散的情况下,是不是有以下的关系?
E=(L-L_0)/L_0
e=(L_0-L)/L
假如我用的是ln的形式,
E=ln(L/L_0),e=ln(L_0/L),他们是不是就只差一个正负号?
对于2维的E和e,如何转换呢? |
|
a******n 发帖数: 98 | 21 linear lagrangian strain: Lij=(Ui,j+Uj,i)/2
eulerian strain is related to the time change, can not remember the
definition
对于2维的E和e,如何转换呢?tensor operation
find a bood of Continuum Mechanics and it should have the definition and
relation |
|
j****x 发帖数: 943 | 22 初学CFD, 正在研究 Free Surface Flow, 哪位高人给介绍一下有关 Arbitrary
Lagrangian-Eulerian 和 Level-Set 方面的资料? 谢了先!! |
|