l***o
发帖数: 47 | 1 工科PhD两年已过，现在的方向属于solid mechanics，课题主要是倒腾FEM simulation
，FEM不
算大牛，但加上有点在大型open source program上的coding经历，也算某种程度的精
通。
以前本科时候是学偏自动化的，后来上研究生了就随了做structural dynamics的导师
，到美国来纯
粹是mechanics，两年也几乎把所有mechanics的课程都学过了。
但发现自己对研究方向提不起兴趣，主要是研究对象跟工业界的实际应用不是很靠得上
，感觉研究没有
多大意义，对实际应用没有多大的impact。另外，所在系很不好找工作。
所以，我现在有点想换专业和方向，主要是三个思路：1.看能不能转向偏电的研究方向
（喜欢电是因为
以前本科毕业设计用西门子控制器来控制加工中心感觉做的很满足）2.或者，转向实验
和模拟并重，具
有实际工业应用价值的方向。3，脱离纯粹的mechanics，希望能研究其他物理量（
thermal，
fluid）
想听听这里当faculty的大牛们的意见 |
|
j******l
发帖数: 114 | 2 本人三流烂校POSTDOC， 现急着FILE NIW 已经E-FILE 140了，7天内得把SUPPORT DOC
送走，继续再找一封推荐信，有没有大牛能帮个忙帮写（我自己草拟）封推荐信。
我是CIVIL的但很多东西跟复合材料，FEM相关，如果您的背景是STRUCTURAL, concrete
materials,MECHANICS,MATERIAL SCIENCE,COMPOSITES,FEM,INFRASTRUCTURE甚至是分
子动力学相关的, computational chemistry都可
如果您在百忙之中能抽出点时间帮个忙，我非常感激！！！ |
|
p*******g
发帖数: 92 | 3 谢谢各位师兄了，最近找工作找的实在迷茫了，把自己的背景说说，看看大家有啥建议。
背景就是top 20的ee program里面一个很好的组里做piezo devices，有一定的mems
fab经验，项目做的比较杂，从自己的device的FEM simulation到fabrication到用自己
的device做的portable system的analog/MCU based control circuit电路设计都做，
除了piezo device，fem的acoustic simulation，还有自己的这个device相关的MCU
based 这个system也都投过文章。
找了3个月工作了，之前有个很不错的硬件公司的research lab给了个onsite，本来觉
得没啥问题，基本就算定了，老板也觉得他们之前都合作过，应该没啥问题，结果拖了
两个月把我拒了。我老板也很帮忙，没有说啥，给推荐了一个top 5的ee program的
postdoc，说不行就去这家吧，如果我想继续找，可以不忙着毕业。前两天老板含蓄的
跟我问了下找工作的情况，问我是怎么打算的，因为他有个新的pr... 阅读全帖 |
|
A***D
发帖数: 64 | 4 The AcuSolve group is seeking talents for multiphase flow solver development
. Please see the following:
Location: Mountain View, CA
Job Description:
We are looking for an expert in the development of simulation software for
multiphase flows. We are a team of highly-motivated CFD professionals. This
is an opportunity to join AcuSolve's core development team in Silicon Valley
California and play a significant role in the development of Altair's next-
generation multiphase technology.
Altair Acu... 阅读全帖 |
|
p*******g
发帖数: 92 | 5 谢谢各位师兄了，最近找工作找的实在迷茫了，把自己的背景说说，看看大家有啥建议。
背景就是top 20的ee program里面一个很好的组里做piezo devices，有一定的mems
fab经验，项目做的比较杂，从自己的device的FEM simulation到fabrication到用自己
的device做的portable system的analog/MCU based control circuit电路设计都做，
除了piezo device，fem的acoustic simulation，还有自己的这个device相关的MCU
based 这个system也都投过文章。
找了3个月工作了，之前有个很不错的硬件公司的research lab给了个onsite，本来觉
得没啥问题，基本就算定了，老板也觉得他们之前都合作过，应该没啥问题，结果拖了
两个月把我拒了。我老板也很帮忙，没有说啥，给推荐了一个top 5的ee program的
postdoc，说不行就去这家吧，如果我想继续找，可以不忙着毕业。前两天老板含蓄的
跟我问了下找工作的情况，问我是怎么打算的，因为他有个新的pr... 阅读全帖 |
|
|
S*******t
发帖数: 6 | 7 Department of Civil and Architectural Engineering
City University of Hong Kong
Kowloon, Hong Kong
Applications are invited for PhD scholarships or Graduate Research
Assistantships in the area of Geotechnical Engineering in the Department of
Civil and Architectural Engineering. The formal PhD program should start
from the fall semester of the academic calendar year (e.g., September 2012)
or occasionally from the spring semester (e.g., March 2012). Online
applications are accepted anytime of the y... 阅读全帖 |
|
w**k
发帖数: 6722 | 8 这里都是ETF
发包子吧
AADR AAIT AAXJ ACIM ACWI ACWV ACWX ADRA ADRD ADRE ADRU
ADZ AFK AGA AGF AGG AGLS AGND AGOL AGQ AGZ AGZD AIA
AIRR ALD ALFA ALTS AMJ AMLP AMPS AMU AMZA AND ANGL
AOA AOK AOM AOR ARGT ARKG ARKK ARKQ ARKW ASEA ASHR
ASHS AUNZ AUSE AXJL AXJS AXJV AYT AZIA BAB BABS BAL
BAR BBC BBH BBRC BCHP BCM BDCL BDCS BDD BFOR BIB BICK
BIK BIL BIS BIV BIZD BJK B... 阅读全帖 |
|
a**********e
发帖数: 147 | 9 其实EE/CS之类那里有您说的那样好? 很多都是钱多的，都是编程密集型的，整天至少8
小时面对屏幕，不好玩的。
1 估计您的孩子也是在美国生得，有美国护照啦，而且是男孩，这样也好分析。
2 其实很多传统工业都有不少机会，很多的盈利不比那些所谓的新兴行业差，例如传统
的石油工业对比网络业。
3 男孩子，爱动手，这个年纪喜欢捣鼓自己的自行车，摩托等机械，就可以准备让他念
个本科的mechanical engineering,重点放在画图(catia)+ 机械加工 + FEM分析(估计
本科不会涉及太深,而且孩子也要有能力念哦)。作为准备，最好上大学前替他联系当地
的修车铺头，做个培训，能拿个技师执照更好啦。这样做的目的就是
1) 让孩子可以在学习修车技能的时候，接触社会，同时掌握实际的机械维修技术，任
何科技都是源于实际应用。学会修车后，也就对一般的机械工具有了解，方便以后工作
，生活，例如自己盖房等了
2) mechanical engineering做画图，设计需要知道一定的机械加工，因为在目前的CAD
设计中，几乎所有的东西都可划出来，但是在实际加工中，很多东西太复杂了，则只有
5-... 阅读全帖 |
|
d********e
发帖数: 8 | 10 I doubt.
Check this out at http://la.wiktionary.org/wiki/alius
sing. masc. fem. neut. plur. masc. fem. neut.
nom. alius alia aliud nom. aliī aliae alia
acc. alium aliam aliud acc. aliōs aliās
alia
gen. alīus alīus alīus gen. aliōrum aliārum
aliōrum
dat. aliī aliī aliī dat. aliīs aliīs
aliīs
abl. aliō aliā aliō abl. aliī |
|
w********r
发帖数: 94 | 11
FEM软件恐怕只有windows才有。我个人觉得用笔记本做这些事情的话，主要工作就是做
Geometry，然后将Geometry导入到工作站上运行。一定要在laptop上跑FEM，FDTD或者
MoM之类的东西没有意义啊。
MAC OS+Parallel Desktop+Win7 挺方便的，这是满足你这种需求最好的办法了。 |
|
d******n
发帖数: 42 | 12 I write scientific computing programs (FEM) and I care about speed and
memory usage. So I chose C++.
I am not aware of any FEM program written in Java. |
|
B**T
发帖数: 294 | 13 1993 Ratings1 Faculty
Students2 Doctorate Recipients3
Institution 93Q 93E 93C Tot % % % Pub/
Gini Cite/ Gini Tot % Rpt % % %
% % MYD
Fac2 Full2 Supp4 Pub5 Fac5 Pub5
Fac5 Cite5 Stu Fem PhDs Fem Min US
RA TA
|
|
t******d
发帖数: 376 | 14 oh, just want to clarify that with you since I got confused.
my point is some companys (mainly in structural analysis, geotechnical-
structrual interaction and ground water flow areas) use FEM.
FEM", |
|
t******o
发帖数: 467 | 15 It depends. For residential or low rise building design, which is wood
structure with concrete/masonry wall/foundation, we nearly never use any FEM
analysis.
We use program SAP2000 and Risa for most steel and aluminum design. Most of
them were based on elastic ASD design or LRFD. For prestressed concrete
design, we use ADAPT with FEM analysis included in the program. |
|
a*****7
发帖数: 18 | 16 【 以下文字转载自 ME 讨论区 】
发信人: abc2007 (胭脂), 信区: ME
标 题: 谁用过FEMAP？
发信站: BBS 未名空间站 (Thu Mar 20 16:23:48 2008)
来说说
这个软件有什么针对性吗？譬如专门给做结构的。
可以解三维结构的问题吗？可以解断裂力学里裂纹尖端的受力吗？
我不懂FEM和FEM软件。麻烦解释地浅显些。谢谢。 |
|
H**E
发帖数: 620 | 17 FEM has nothing to do with the material/dimension.
what really matters is mathematically, what kind of the governing equation
you are dealing with.
I think you can use FEM as long as it's PDE field problem. |
|
S*******t
发帖数: 6 | 18 Department of Civil and Architectural Engineering
City University of Hong Kong
Kowloon, Hong Kong
Applications are invited for PhD scholarships or Graduate Research
Assistantships in the area of Geotechnical Engineering in the Department of
Civil and Architectural Engineering. The formal PhD program should start
from the fall semester of the academic calendar year (e.g., September 2012)
or occasionally from the spring semester (e.g., March 2012). Online
applications are accepted anytime of the y... 阅读全帖 |
|
r*******g
发帖数: 453 | 19 sc. Thanks for your information.
I have another question though ^_^
After you get the surface (of a bunch trangles), how to generate 3-D mesh for
FEM? What kind of FEA package did you use? And with it, how to incorporate the
surface information? Thanks.
specimen
FEM
sacnned
to |
|
f****r
发帖数: 27 | 20 来自主题: Computation版 - 一个困惑？
^^^^^^^^^^^^^^^^^^^^^^^^^^
I don't know any detail about your simulation. However, if your problem is
majorly about some very complex geometries, FEM certainly has its merits.
As for those issues about parallelization and computation, you sure can solve
them with some extra work.
ON the other hand, if your domain is usually organized, like some simple
rectangular or circular domains, ditch the FEM approach. It's too expensive
and slow.
Just my 2 cents. |
|
l****y
发帖数: 92 | 21 来自主题: Computation版 - 一个困惑？ Thanks for your reply.
这个牛人是这么评价FEM "Despite their geometrical flexbility, unstructured
girds are rarely used in plasma simulation because of their high programming
and computational overhead and because of the difficulty to parallelize such
codes."
btw, could you recommend some books about FEM for beginners?
限元
不适
有人
solve
不是 |
|
l****y
发帖数: 92 | 22 来自主题: Computation版 - 一个困惑？ 正是因为fem不是我们这个领域模拟的主要手段，大部分人（包括牛人）都是做有限差分
，所以我才觉得困惑啊。但是导师对他自己的code似乎很满意，但是目前还没有出什么文
章，我估计还是有不少问题要解决的
我们系没有关于fem的课，所以希望大家推荐一些经典的比较适合入门的书
新
够
programming
such |
|
f****r
发帖数: 27 | 23 来自主题: Computation版 - 一个困惑？ Well, I'm not a fan of FEM and don't use it in my research. Therefore,
I can't give you any good references.
There are many methods to solve the PDEs you are working on. Any methods out
of there, especially for those widely deployed ones, certainly have their
merits and defects. To figure out the best one, or more precisely,
the one that "universally" beats all, is like "mission impossible III".
For example, unstructured mesh with FEM is the standard method for
solid mechanics on arbitrarily com |
|
x*****u
发帖数: 3419 | 24 FDM,FEM,FVM,搞cfd的，有用fem的吗？有什么优缺点相比于fdm的？ |
|
c******m
发帖数: 599 | 25 inf-sup条件就是如果用NS方程计算的话
速度有限元空间和压力有限元空间姚满足某个条件,就是infsup条件
不是什么有限元空间都可以拿来用的
精度取决于格式吧,低阶的fem不会比高阶的fdm精确的
fem边界容易就是因为可以三角网格,而且自然边界条件无需处理
差分的话,边上的点总是麻烦的地方
而且如果区域古怪兮兮的,差分还要做个变换变成矩形区域 |
|
t*****o
发帖数: 74 | 26 you work on CFD?
I think the problem for fem is conservation. Most time the N-S need to write
in energy conservation form which is difficult for fem, like in the case of
shock. Maybe not true..... |
|
j****x
发帖数: 943 | 27 1. fluid mech和stat mech难道不是机械系的？
2. "finite element感觉并不实用，现在那么多现成fem软件可用", 不实用，为什么有
这么多现成
的fem软件？逻辑，逻辑。。。 |
|
L*******g
发帖数: 913 | 28 咬文嚼字，我说过机械系不能上这个课？
我说单纯学fem不实用，因为很多现成软件已经提供这个功能。就算没有fem知识，也能
很快上手。不知道你要争个什么？ |
|
j****x
发帖数: 943 | 29 我想说的是开这些课的主要是是机械系， 而不是物理系。如果物理系开这些课， then
they are taught with a very different perspective.
如果对fem不了解的话，对于fem软件的使用会有非常大的问题。如果你连natural
boundary condition 和 essential boundary condition 都搞不清楚，如何很快上手
呢，更不用说很多边界条件和计算方法的选择。 |
|
A*******n
发帖数: 14 | 30 以前都用matlab写程序。读博士做FEM，入门状态。哪位大牛肯给我段用c++写的关于
fem或者计算电磁的例子，让我感受一下啊！谢谢大牛！ |
|
H***F
发帖数: 2501 | 31 rf FEM还是有机会的。频率上不去，只能走CA，mimo路线。带来的增长是FEM更加复杂
高频段的硬伤是path loss啊，真的要做热点覆盖不知道要布置多少站点才能覆盖。从
4G以前的经验来看，新一个世代技术都是走先覆盖在热点容量提升的路线。毫米波做宏
覆盖恐怕没有人会相信，如果是微站甚至微微蜂窝，那么4G时代的典型技术small cell
失败是一个明例。
RF以及HW在未来成长性看得见是比较弱，相关的上下游包括电磁仿真，天线设计，RFIC
，HW等等成长空间都会压缩，大家放低期望值为好了。 |
|
c***s
发帖数: 1292 | 32 FEM should be broader ?
i am not expert in FEM
returned all those stuff to my teacher in China le, hehe |
|
c***s
发帖数: 1292 | 33 yeah, i know FEM is very useful
could you recommnd me a book about FEM?
have much interest in that
of course, i am not mechanics guy, just a general book is fine |
|
c***s
发帖数: 1292 | 34 apply apply
elementary la
i said i am not expert in FEM just want to have a rough idea
regret that I didn't learn FEM well in college |
|
t*****o
发帖数: 74 | 35 the job opportunity for FEM is much better than CFD ......
many position avaible for FEM |
|
k*****a
发帖数: 1518 | 36 nowadays FEM is way tooo easy to perfrom, even a college freshiers trained for
several months may get get most FMM work done pretty decently.
so, a Ph.D. is not always needed for a FEM analyst position.
get |
|
Z**X
发帖数: 366 | 37 How soon can one learn to use it?
suppose no pre-experience with commercial FEM software and only FEm basic
knowledge.
Thanks. |
|
r********o
发帖数: 1423 | 38 learn FEM, FEM is now 10K Gold Oil |
|
N**O
发帖数: 1106 | 39 ME里面做FEM的应该不错吧，我觉得做thermal和fem的应该工作好找一些
理论实践可以马上转换嘛，hoho
那还是manufacturing 好找咯？
好象我认识的ME里做control的都该行了，EE的control貌似也不好找？ |
|
a*****7
发帖数: 18 | 40 来说说
这个软件有什么针对性吗？譬如专门给做结构的。
可以解三维结构的问题吗？可以解断裂力学里裂纹尖端的受力吗？
我不懂FEM和FEM软件。麻烦解释地浅显些。谢谢。 |
|
j****x
发帖数: 943 | 41 FEM is just one category of methods. CFD uses FEM as well. If the problem is nonlinear enough, FEA has convergence problem too.
FEA involves more math? Try turbulence. |
|
e*l
发帖数: 37 | 42 很鄙视你这种做CFD发牢骚的！
求解CFD的数值方法有FVM, FDM和FEM，差分方法是最简单的，FEM是最难的，怎么
用FEA的会是最幸福的？
CFD中数学用得最多是数值分析，PDE等都有现成的数值解法，又不是让你重新搞
一个新的格式。而且CFD基本都是围绕N-S方程，再加几种典型的湍流模型，都是
针对具体的东西研究。跟人家数学系本科生都差得远，还根研究生比，以为人家
数学系的都干的跟你一样的活啊？
流体力学是很经典的学科了，看个边界层理论就能把你看晕，凭你这水平学其它
学科也好不到哪去。流体力学里最复杂的还是湍流，到现在都还没研究清楚机理。
涉及到的计算机知识其实很简单。C/Fortran在数值计算中使用其实很简单，新手
看个一个星期基本就能写代码了；python就更不用说了，看一个下午足以；C++语
法复杂，但是其实你一个搞计算的能用到它的多少功能；shell编程，比如bash，
更简单；matlab是学工科的基本工具，傻瓜都会用；linux你也只是用用而已，能
有多难，难道你搞系统开发；并行计算就那点概念，MPI、openMP能有多难，人家
现成的库给你用，还嫌用起来麻烦？有 |
|
s**********a
发帖数: 3 | 43 Hi,
I am a new beginner to FEM and just started to use FEM to simulate 3D fluid
flow to solve N-S equation. We have our own mesh format which is different
from FLUENT, so I need to change mesh format. Is there anybody could tell me
what's the format of FLUENT mesh or send one copy of FLUENT mesh? I search
online, but cannot find it.
Another question is I have already tried COMSOL to simulate, but it seems
COMSOL is bad at 3D. I have no idea whether FLUENT will be good? Is there
anyone have good |
|
i****a
发帖数: 88 | 44 MOR usually works on linear system. Assume you have n = 10k nodes in FEM
model, MOR projects the system
matrices into a subspace with order, say ~ 10. So instead of solving a 10k x
10k dimension ODEs in FEM
software, a system of equations with 10 x 10 dimension matrices would be
much faster to solve. Once you get
the solution on these 30 nodes in the subspace, you can project the results
back to the 10k nodes to get the
actual results.
we |
|
i****a
发帖数: 88 | 45 Hehe, even minutes is so slow in my view as I am not a big fan of simulation
. MOR is a hot topic now - very
promising for fast simulation and real time prediction purposes. As a
comparison, simulation taking minutes
in FEM will only spend ~1s using MOR. Isn't it attractive? The accuracy
depends the order you choose to
approximate the original model in FEM. It tries to match the first several
moments of the system in the
frequency-domain. For typical thermal problem which is a low pass filter ... 阅读全帖 |
|
i****a
发帖数: 88 | 46 MOR usually works on linear system. Assume you have n = 10k nodes in FEM
model, MOR projects the system
matrices into a subspace with order, say ~ 10. So instead of solving a 10k x
10k dimension ODEs in FEM
software, a system of equations with 10 x 10 dimension matrices would be
much faster to solve. Once you get
the solution on these 30 nodes in the subspace, you can project the results
back to the 10k nodes to get the
actual results.
we |
|
i****a
发帖数: 88 | 47 Hehe, even minutes is so slow in my view as I am not a big fan of simulation
. MOR is a hot topic now - very
promising for fast simulation and real time prediction purposes. As a
comparison, simulation taking minutes
in FEM will only spend ~1s using MOR. Isn't it attractive? The accuracy
depends the order you choose to
approximate the original model in FEM. It tries to match the first several
moments of the system in the
frequency-domain. For typical thermal problem which is a low pass filter ... 阅读全帖 |
|
j******l
发帖数: 114 | 48 本人三流烂校POSTDOC， 现急着FILE NIW 已经E-FILE 140了，7天内得把SUPPORT DOC
送走，继续再找一封推荐信，有没有大牛能帮个忙帮写（我自己草拟）封推荐信。
我是CIVIL的但很多东西跟复合材料，FEM相关，如果您的背景是STRUCTURAL,
MECHANICS,MATERIAL SCIENCE,COMPOSITES,FEM,INFRASTRUCTURE甚至是分子动力学相关
的都可
如果您在百忙之中能抽出点时间帮个忙，我非常感激！！！ |
|
A***D
发帖数: 64 | 49 The AcuSolve group is seeking talents for multiphase flow solver development
. Please see the following:
Position: Senior CFD Software Developer for Multiphase Technology
Location: Mountain View, CA
Job Description:
We are looking for an expert in the development of simulation software for
multiphase flows. We are a team of highly-motivated CFD professionals. This
is an opportunity to join AcuSolve's core development team in Silicon Valley
California and play a significant role in the devel... 阅读全帖 |
|
g*********n
发帖数: 808 | 50 【讨论】扫描电子显微镜的设计思想和工作原理
扫描电子显微镜的设计思想和工作原理，早在1935年便已被提出来了。1942年，英国首
先制成一台实验室用的扫描电镜，但由于成像的分辨率很差，照相时间太长，所以实用
价值不大。经过各国科学工作者的努力，尤其是随着电子工业技术水平的不断发展，到
1956年开始生产商品扫描电镜。近数十年来，扫描电镜已广泛地应用在生物学、医学、
冶金学等学科的领域中，促进了各有关学科的发展。
一．扫描电镜的特点
和光学显微镜及透射电镜相比，扫描电镜具有以下特点：
(一) 能够直接观察样品表面的结构，样品的尺寸可大至120mm×80mm×50mm。
(二) 样品制备过程简单，不用切成薄片。
(三) 样品可以在样品室中作三度空间的平移和旋转，因此，可以从各种角度对样品进
行观察。
(四) 景深大，图象富有立体感。扫描电镜的景深较光学显微镜大几百倍，比透射电镜
大几十倍。
(五) 图象的放大范围广，分辨率也比较高。可放大十几倍到几十万倍，它基本上包括
了从放大镜、光学显微镜直到透射电镜的放大范围。分辨率介于光学显微镜与透射电镜
之间，可达3nm。
(六) 电子束对样品的损... 阅读全帖 |
|
