These are priceless pictures. This is not a criticism to Goodidea, I only wish there are more.
Picture 1 shows 2 sails properly connected to the supporting elements. From Pictures 2 & 3, we may see that fabric had slid down the catenary cables from the top connection to the building center tower. Picture 2 also indicated that one of the posts (on the far left but not shown in the picture) was leaning to the right and the post right to it leaning to the right as well. It is also evident that fabric was in contact with the roof top railing, the railing cut the fabric while fabric/cable dented the railing.
Though it is not clear where does this broken clevis belong to and not knowing what type of membrane plate design for all the corners, I can see there is a broken cable in Picture 6, looks like a Guy cable, I come up with this scenario of the event.
1. A Guy cable attached to the left post in Picture 6 broke.
2. This post lost equilibrium and began to tilt to the right.
3. Catenary cables attached to this post lost tension.
4. Fabric began to slide down the Catenary cables.
5. Loose fabric caught a lot of wind, the rest is history.
May be there are lessons learned here:
1. Any structural system should have certain redundancy build in. If a structure like this one is counting on the fabric to maintain stability, other security measurement should be put in place such as overhead tie cables.
2. Quality control of all essential items such as cable clevis. A qualified clevis should be able to survive at least 200,000 times cyclic dynamic loading test.
3. The cause of the broken clevis seems to be fatigue. One problem would be the quality of the clevis; the other would be the tension in the system was not enough. When pre-tension is low, too much vibration will damage the clevis.
4. Corner fabric was not held in position with clamping bar. Without the restraining device, fabric could slide down the cables creating the bloom.
Luckily no one was hurt. We need to pay attention to small details even for small structures. Just like a tailor, it requires the same process to make a shirt for a giant or a new born. The difference is only in quantity.作者:
wangshen1977 时间: 2005-3-10 13:24 标题: 回复： 千载难逢，亲眼目击某景观膜结构破坏全过程！
A few things I would like to point out for discussion:
a. It appears that the only tension still remaining in the fabric comes from self-weight gravity force. It requires re-tensioning in both WARP & FILL directions immediately. Other than Fiberglass material fabric, it is essential to have re-tensioning mechanism designed into structural elements. Without proper tension in the system, this structure will be easily damaged by either wind or snow.
b. Some adjustability in cables can re-tension the system. Instead of cables, steel bars with threaded ends can easily achieve what we need. Turnbuckles may be expensive for this application.
c. I hope the fabric corner is reinforced with bias patches.
d. I also hope the cables are galvanized.
e. Design of cable anchor is also not without questions. 1. Why anchor bolts are not parallel to the cable similar to the post anchor? 2. Why the line of action is not through the center of the anchor? 3. Why anchor bolts spaced so far apart? 4. Do we know the base plate is thick enough for bending?
f. The concrete work for the anchors is suspect. If the whole anchor is only the concrete block we see in the picture, the weight of it is simply not enough to resist the load from the fabric. I do not believe this is the case. But the other possibility is also structurally wrong. It appears that the top block is cast separately from the underground portion. This means the top block can move independently and bend the anchor bolts. I don’t believe these bolts are designed for bending.
This is alarming. When designers don’t pay attention to seemingly innocent details, we are in for trouble.作者:
zyqzzz 时间: 2005-3-23 12:59 标题: 回复： 千载难逢，亲眼目击某景观膜结构破坏全过程！
1、应该说，在风荷载不大的情况下。两片膜之前已经丧失了预应力，并且边界点（膜角）已经没有刚度了（可能膜布已从膜角板处脱开，随着边索滑动，或支撑结构-落地拉索已经破坏）。如图1。
2、在风荷载继续作用下，膜布逐渐从膜角向两边滑落（slide down） ，逐渐被撕开。同时远端的支撑杆倾斜，一根拉索破坏，柱脚耳板弯曲屈服，丧失刚度。如图4。
3、davidchen分析得非常有道理！When pre-tension is low, too much vibration will damage the clevis（预应力的不足，结构振颤会摧毁整个结构）。从受损的配件看到，某些销轴截面已经减小了近1/3！
4、破坏的直接原因很难讲得清楚。很不幸，因为它有很多缺陷！作者:
zyqzzz 时间: 2005-3-23 13:01 标题: 回复： 千载难逢，亲眼目击某景观膜结构破坏全过程！
Could you please be more specific? Your opinion and suggestion can help everyone for the future.
1、 膜的张拉不到位: Can you tell us why? May be there was no mechanical means to do that? Or it requires re-tensioning?
2、 该结构在找形中有先天缺陷: Do you mean it needs more curvature? Or there were flat areas?
3、 工人安装水平一般: Are you implying bad workmanship caused the failure?
4、 我觉得结构是有问题的, 在安装时就能感觉到: This is a very broad accusation, can you enlighten us more? Are you referring to member size, fabrication quality, foundation work or fabric?
I thank you in advance for your help in this matter.作者:
zhlinlong 时间: 2005-4-8 11:40 标题: 回复： 千载难逢，亲眼目击某景观膜结构破坏全过程！
