刚到海南岛玩了一趟,巨爽8D
    尤其是三亚的亚龙湾，“天下第一湾”果然名不虚传，当然也看到了三把膜结构大伞，远看尚可，近看伞顶有皱纹。
   伞形膜结构建筑小品恐怕是现在国内膜结构运用最多的一种形式，但我所看到的伞形膜结构的伞顶处几乎都有皱纹，这恐怕也算是质量通病了吧，随便再举几例：上海浦东世纪大道上有两把双伞，宝山牡丹江路白玉兰广场的一把双伞，莘庄地铁出口的三联伞。。。。。。上海的坛友可以去看看。
   产生伞顶皱纹的原因在下认为是安装时伞箍或伞柱的扭转造成，这在设计时是无法考虑的，另外伞顶位置安装误差也很难控制。在安装调整过程中还会出现当时调整无皱了，但过了几天又会有皱的现象。
   哎，伞膜除皱恐怕跟女人除皱一样难:~)

伞形膜结构的伞顶容易出现皱褶, 这是肯定的了。我认为主要有以下几个原因。
1， 从设计的角度来看，伞顶是圆环并为固定节点，设计不合理的话与计算找形有出入。这个很重要，其实需要注意的都是一些小问题，但被忽略后就会到来问题。
2， 伞形膜结构一般来说热合缝都是经向布置，到伞顶热合缝过于密集，有加工误差和热合造成的膜变形。热合时这个位置的加工难度比较大。
3， 伞顶其实也是边界，膜中间部分的误差往往会在边界体现出来。
4，安装时伞顶的定位点偏差。膜设计点要与钢架设计点完全重合。膜及钢架都要有明显标记，忽略这个过程会造成施工误差使膜受扭。
  
从我个人经验来说，设计是否合理是保证工程质量的前提，有了正确的设计这并不是很大的问题，可以解决。

另外，你去海南岛玩去博鳌了吗？ 那个亚洲论坛的的主会场你看到了吗？
这是我的得意之作。上过新闻联播的。哈哈。
  
^_^.......

Wrinkling is a common phenomenon.   
  
Actually, wrinkling is a buckling problem of membrane structures( I mean structural wrinkles not material wrinkles here). Since the bending stiffness of fabric is very small, when the minor principal stress is zero or negative, i.e. compressive stress, the wrinkle occurs.
  
   There are many factors can induce wrinkles not  installation only. There are a few methods for analysis wrinkling problem.

hhux2002
  
你说的非常有道理。
的确，膜褶皱的产生是因为单元丧失预张力处于松弛状态。
formfinding，analysis，design，patternning，fabrication and contruction这里面哪一个环节有问题都会造成皱褶。
有的时候，皱褶对整个结构的受力也是极为有害的。我见到深圳有一个工程，皱褶特别大，钢索也没有施加上预应力。感觉很危险。
  
你对膜结构研究的很深，有机会多多向你学习。

补充。
膜在双向预应力都能达到设计要求时，没有皱褶。有时只是单向有预应力，会产生竖向波纹，但结构稳定还能保证。最怕的就是感觉双向全没有力，我说的那个深圳的工程看上去就象这样。

zhaoyu xiong ke qi le, xiang hu xue xi.
  
If the major principal stress is also non-positive, i.e. both two directions without tension, as you said, the membrane is not wrinkled but slack. A slack membrane or a slack region of membrane can not carry any load. In this case, the membrane is just a cover but not a structural member. Such a structure, in view of mine, is not a membrane structure. Membrane structures are tension structures, how can we call a membrane without tension as a structure? it's just a fabric, a piece of cloth only!
  
Lets concentrate my topic on wrinkling problem here. Wrinkles are the out-of-plane deformation. There are two types of wrinkles, i.e., structural wrinkles & material wrinkles. Structural wrinkles due to localised buckling/compressive stress. Material wrinkles due to material, fabrication and folded packaging. The wrinkles that you mentioned in No.2 are material wrinkles, the others and rockmad mentioned are structural wrinkles.  Material wrinkles are permanent deformations. Structural wrinkles are not permanent, they may be disappear when change the prestress or boundary conditions or applied load, in other words, they may be eliminated through careful design and installation.  However, the analysis model and design software need further study and development.
  
To make the wrinkles disappear is tough, but find a girl without wrinkles is possible. rockmad xiong,  form-finding. ;)
  
qing fu zheng.

The drawbacks of the wrinkled membrane
  
The previous posts in this topic are mainly about the wrinkling criterion and the types of wrinkles.
  
It seems no more response. I put one more post here and welcome everyone join to discuss. I am sorry for my English input again, I can read but can not type Chinese by my computer.
  
The drawbacks of the wrinkled membrane are the wrinkles not only affected the look but also affected the structural behaviour. Once the wrinkles occur, that means the minor principal stress of the membrane was zero or negative; in the regions nearby the wrinkles, the prestress also did not reach the design value; the force in the surface of the membrane was re-distributed. Hence, the stress in somewhere was decreased, while in otherwhere increased.
  
In the regions that the stress decreased, the stiffness of the structure also lower, the membrane will vibrate under the wind easily. In the regions or direction that the stress increased, the overloaded stress will induce large strain and large time-depandent deformation, moreover, it may make the membrane torn.
  
The reasons of wrinkling have mentioned in previous posts but not in particular. Which friend can analysis in detail?.

半谷裕彦有做过这样的研究。你去找找他的文章，可能对你有些启发！
：－）

Wrinkling Analysis
  
Thanks to sillyboy and zhaoyu.  I talk a little more about the wrinkling problem.
  
For analysis wrinkling problem of membrane structures, there are three types of theory, i.e.
      1. Tension Field Theory;
      2. Wrinkling Theory; and
      3. Critical Compressive Stress Theory.
No compressive stress to be allowed for the membrane in the tension field theory and the wrinkling theory, but in the critical compressive stress theory, a small compressive strss is allowed.
  
For finite element analysis of wrinkling problem, there are two types of element available, i.e.
      1. base on tension field theory and wrinkling theory, use membrane element with no bending stiffness;
      2. base on critical compressive stress theory and bifurcation buckling theory, use shell element, treat the wrinkling problem as a buckling prolem.
  
Why I have interesting in this topic?
  
When perform load analysis, a few researchers use a factor for reduce the stress of wrinkled element, but in view of mine, wrinkling means the stiffness of the membrane is not adequate, or in other words, the initial geometric surface is unresonable, in such case, the analysis should be return to the formfinding and modify the initial curved surface.  
The formfinding, load state analysis(include wrinkling analysis) and cutting pattern generation should be a whole-process.  
  
Current software, such as Surface and Patterner, do not include the load analysis module; others such as Easy, Forten32, WinFabric etc. can perform formfinding, load analysis and cutting pattern generation, but can not perform wrinkling analysis.
  

What you have written is reasonable. But it is similar to doing  some optimise if it returns to the step of formfinding. I am not sure whether I have got it or not. However, it is inconvenient unless the software is intelligent. Is there an alternative? How about to modify the stiffness of wrinkled membrane element?

Which one should be modified?
  
Maybe  we can call the modification as a shape optimization. There is a paper about the shape optimization in this forum, but I can not open it ( I can not open CAJ file since there is no Chinese Word in my computer). I met the author when he was a MEng candidate. I have interesting in knowing that whether the shape optimization is based on the results of load state analysis.
  
I am not very clear the meaning of  " to modify the stiffness of the wrinkled membrane element".  As a numerical method, we can modify the stiffness of  wrinkled membrane element ( to reduce or  ignore the stiffness ), but what about the result? Maybe we can say that we have consider the affect of the wrinkled membrane element in our analysis ( as I read in a few papers ), but if we do not modify the structure ( to modify the B.C or prestress ),  the structural stiffness does not changed.
  
Which one should be modified, the stiffness matrix or   the structural stiffness ?

我接触膜结构已有几年了，根据我的观察及经验，膜结构的褶皱主要是由以下几个方面造成的：
1.设计：包括膜边界的确定和膜的设计。
2.裁剪：即确定合理的裁剪方案及加工工艺。
3.安装：安装必须严格按照安装方案进行。

我还以为是那个“钢圈”做小了或者是做“扁”了，原来还有这么多原因，看来做膜结构真不容易！

我认为最主要的原因是膜的加工工艺问题。
当然设计与安装也是其中的一些原因，但不应该是主要原因。
至于钢结构的加工问题应该不大，国内的小公司基本上都可以做到按图加工，误差不会很大，更不用说稍微大一点的公司了！

有几家公司能够在张拉时对张拉力进行控制（索测力计）？

经常可以看到在伞形膜结构的顶部圆环处有褶皱出现，如图所示。在实际工程中我发现在这个部位索和膜是分开的，但在设计中这里是按照索膜连在一起设计的，这是不是发生褶皱的原因之一呢？不知哪位高手有解决办法？

看了hhux2002对膜皱折问题的论述很受启发，也有了比较清晰的认识。
很少有论文讨论过膜的皱折问题，偶尔有几篇论文也是简单地认为膜中出现压应力即认为其退出工作。
hhux2002提到了三种理论解决这个问题，那么具体做法怎么样？还望您进一步赐教。
  
  

回复请见： 褶皱分析的理论和方法  

首先感谢hhux2002的系列谈,让一个门外汉也知道了一些索膜钢结构的力学问题。
有一个肤浅的问题:在大伙的讨论中,为什么没提及热胀冷缩问题,热胀冷缩可以不计吗?
「在安装调整过程中还会出现当时调整无皱了，但过了几天又会有皱的现象。」是不是与热胀冷缩也有关。热胀后,张力降低,从而产生皱纹。

谢谢 sjtu79  的鼓励。
  
说明一下，系列谈不是我个人的，许多朋友都有参与，skyland  和 mibao  等已经为系列谈带来了一些专题帖子，其他帖子及其他朋友答应为系列谈撰写的帖子也将陆续贴出。
  
许多朋友包括 okok、zhaoyu、成钢教授等对系列谈或提出建议、或参与讨论、或表示鼓励和关怀。
  
今天是 Vesak Day 补假，我借用别人的电脑；平时发在系列谈中的帖子都是由 gussds  朋友帮忙录入中文的。
  
希望有更多朋友尤其是从事设计和施工的朋友加盟系列谈，谢谢。
  
热胀冷缩问题或者说环境温度及其变化应该是膜结构设计时要考虑的因素之一，不知各位有何高见？

This is an interesting topic; I have these additional issues to be considered.
  
a.  More often than not the structure is different from the computer model. One example of this is typically we have cable and fabric connected together in the model but separated in reality. In order to re-distribute the imbalance force from the cable (cable now wants to have constant force in the entire length), the stress in the membrane can no longer maintain uniform. It is not a problem for an open fabric area, but it will cause wrinkles at corners and edges.
b.  When there is not enough definition in the patterning model to represent the curvature in the membrane, wrinkles will happen. The mast top is a good candidate.
c.  The sequence of stressing the membrane system can influence the outcome of the look of the membrane as well. Extra care is warranted when you have two catenaries cables merged at a narrow angle to a connection point. The friction between cable and fabric can cause wrinkles.
  
This is a joined effort among architect, engineer, fabricator and installer. Do not take this issue lightly because first impression is so crucial for our business.

对于框架支承膜结构而言，膜面的张力是直接由支承构件承受的。支承构件在张力作用下的变形，将导致膜面张力损失；且由于跨中变形较大、支座或构件相交处变形较小，因而有可能导致膜面出现褶皱。
  
在进行框架支承膜结构的构件设计时，有必要采用比普通钢结构设计中更严格的变形控制标准。

如果膜的找形和荷载分析的程序有了以后，
要用非线形编得刚结构的模块不难
在一定意义上的协同计算可以实现。
easy中有beam的模块，
可惜没有具体的看过
sigh!:)

不要试图在设计上根本解决这个wrinkle，因为这是现在结构研究的一个主要方向；
  
其中有些学者认为这和稳定问题有关－－是的，受拉结构也有稳定问题。
  
网上有很多相关的论文，自己看：
  
key words: membrance structure, wrinkle。。。

伞顶皱纹可以基本上在施工中解决（设计和膜体制作没有太大的偏差！）。安装此类膜结构时，可先把伞顶膜体拉高，并临时固定（注意不要让膜体滑落！），然后固定周边并张拉到位。最后根据具体情况调整伞顶皱纹，直到达到最佳视觉效果为止。（调顶时可松开周边，但必需作好顶部缺陷记录，以便安顶时调整）

的确，你所说的原因是确实是主要的原因，另外上图中部分褶皱可能跟安装不到位也有关系。因为从图中看来，顶部的褶皱也不全是一样的。针对你所说的那部分原因的褶皱，目前国内最常采用的方法，主要还是靠剪裁的经验来控制，而这个经验的得来一方面是靠工程的积累，另一方面你可以采取一些别的方法来获得，比较做些小样等。