Adaptable membrane by Yü Chen

MEMBRANE SPACES studio course AHO fall 2008

adaptable membrane membrane aho studio professor michael additional birger dafne s t u miray liam yü

group individual

course in

AHO

The oslo school of architecture and design membrane spaces fall`08

spaces fall 08 charge: hensel staff: sevaldson sunguroğlu n t s oktem rote-hahre chen

project/ project

adaptable membrane

double layered membransystem-controlable permeability

yü chen // professor in charge: michael hensel // additional staff: birger sevaldson and defne sunguroglu

MEMBRANE SPACES studio course AHO fall 2008 DEVELOPING FROM FORMER PROJECTS

content

STIFF MEMBRANE

PHYSICAL MODEL

DIGITAL MODEL

DIGITAL MODEL PROBLEM

ELASTIC MEMBRANE

1. Basic p04

2. LINKED MEMBRANES p05

3. PROPERTIES p.08

4. SYSTEM OF MEMBRANES p.09

5. CONTROLLING FALLING OBJECTS p.09

2 tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne Sunguroğlu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YÜ CHEN

6. BASIC EXPERIMENT p.10

7. FORMFINDING p.16

MEMBRANE SPACES studio course AHO fall 2008

visual penetration

moveable surface

geometry

1:? motion

cutting pattern

research

scale

shadow analysies

two layer

8.individual project approach p.20

wind flow

9.Shape p.22

10.resarch p.30

11.performance p.32

12.spatial exploration p.39

MEMBRANE SPACES studio course AHO fall 2008 Basic knowledges about Membranes

Advantages of Membranes:

A membrane is in general a thin, synthetic or natural,material and it`s structer belongs to the group of Tension Structures. Def: Tension Structures are only loaded in tension, with no requierments to take compression or bending forces.

- light - slim - structural efficiency - good performances in case of fire Kg

a DISAdvantages of Membranes: The biggest problem of a membran structure is it`s low resistance against aerodynamic excitation (wind fluttering). Solving this Problem leads to the basic shapes of a membran structure. b a. adding mass b. double curvature c. application of tension d. convex and concarve surface

4 methods to prevend membrane structures of fluttering: aI application of mass bI double, in transverse dirctons curved surfaces cI application of tension dI forming a convex and a concarve surface and holding them in tension against each other

behavior of slim/thin elements in different force conditions aI tension bI compression cI bending

d 4

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne SunguroÄ&#x;lu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YĂ&#x153; CHEN

MEMBRANE SPACES studio course AHO fall 2008 The two basic Forms The saddleshape and the coneshape are the most commonly used double curved surfaces to make a construction stiff. They are also minimalsurfaces, which means that the structure, completly loaded in tension, is working on it utmost efficiency.

e a. saddle shape b. cone shape c. membaren with 5 control points d. enclosed membrane e.moving of a membrane system

b To understand the characteristics of membranes we created a basic saddle and a basic cone shape (a). After this started to manipulate it to find the limits of the material, how much tension may be applied and in which directions, before it wrinkles.

In the next task, we were supposed to create three interconnected membranes to understand how it changes the tension in all the system and effects all the membranes in the system when one of the control points in the system is changed. (c) During this experiment, we understood that manipulating a membrane by changing even a single control point changes all the system. So it is hard to change the place of a control point without getting any wrinkles 5

MEMBRANE SPACES studio course AHO fall 2008

way of grapes in different size through a membrane sytem. Frames of a videofilm.

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne Sunguroğlu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YÜ CHEN

top view

MEMBRANE SPACES studio course AHO fall 2008

Bigger grapes skiped the hole and jumped only twice between the membranes

Medium Grapes where hitting more edges and jumped more rendomly

After understanding basicly how membranes works, we decided on two ways to investigate more on membran properties. One was to look closer on the shape (see p.8), and the other was to learn more about its reaction to other object. This membrane system is developed in order to catch a dropping object (in this case different grapes) and reflect them to another membrane and in this , influencing the fall of the objects. We suspended the membranes according to falling direction of the grapes from the former membrane. Conclusion: The bouncing angle depends on the the size(mass of the object. Discovering that different sized grapes (or with different mass) would go different ways we thought about a system to asort objects. Since we dind*t had enough knowledge about the parameters which influences the way of an object through oure system, we decided to do deeper investigation on the bouncing properties of a simple membrane. This leeds to the experimentseries on p.XX.

Small grapes jumped through the hole onto the next membrane

MEMBRANE SPACES studio course AHO fall 2008 Using stiff Minimalsurfaces for a compression structure.

PE and hardener. To learn more about membranes, we decided to work with them in a different way.By producing stiff membranes using a polyester and fiberglass, we created an anticlastic non-elastic object with a minimal surface. We can preserve some properties of the membranes even if it does not work as a membrane anymore. We painted different shaped membranes with polyester and we also put fiberglass to some of them to make it stronger. We tried this method on saddles, cones, barrels and even membrane systems with more than one membrane and minimal holes.

membrane painted with polyester

The membranes all kept their shapes very well. The one coated in fiberglass where extremly strong and rigid. Discovering this, we thought that it could be a way for finding material efficianed compressing structures, similar to buildings like Sagrada Familia (Gaudi).

the catenary is the theoretical shape of a hanging flexible chain or cable when supported at its ends and acted upon by a uniform gravitational force (its own weight) and in equilibrium. It is mathematicaly seen also a minimalsurface. Turned upside down it shows the ideal shape for an arch in almost pure compression while suppoting it*s own weight without bending moments in the material.

Since a membrane in a full loaded tension shows similar properties, (minimalsurface, equibilirium of forces) We wondered if a membrane shape used in the other way around could also show this ideal performance loaded with the same amount of compresson forces as it`s counterpart in tension forces. Also we were perfectly aware of the fact, that if it as we thought than certainly someone on the world would already using this, We liked the idea of a membrane minimal surface as a rigid object used as an compression structure. Thinking of saddleshaped stiff object we made two design proposals.

membrane coated with fiberglassmath and polyester 8

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne SunguroÄ&#x;lu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YĂ&#x153; CHEN

MEMBRANE SPACES studio course AHO fall 2008

The task was to create 12 or similar membrane patches that are connected ot each other with a system and harmony. Before starting this task, we wanted to choose ourselves an aim, in order to explore membranes from performance point of view. We wanted work on a membrane system which works as a reflection path and reflects objects in a certain way that we would be able to direct different objects in different paths according to their size, weight or speed. We talked over ways of doing that over sketches and then started building a membrane which would collect falling objects but not direct them. We thought of this as a way of understanding how membranes reacts when objects ofdifferent sizes fall on them. We created 2 groups of 6 membranes that are interconnected and another half group and assumed that our system is endless.

As we progreesed in the membrane-bounce research, we replicated a former model. We created a completely rendom layered cluster as an intend of prolonging the balls flight. After many experiments that conclude the cluster being very uncontrollable, it evolves into two other clusters. One enclosed and one vortex.

The enclosed cluster is an interconnected set of equal size membranes. When we drop the ball from above as we had done in all the models, it bounces horizontally as well as vertically.

We could easily see that the ball would bounce more randomly in this pre-selected and systematic cluster than in the randonly built one. In the vortex, there was another factor which became very clear; speed. When the ball entered the vortex shaped cluster, it accelerated quickly and had random exits.

We tried to visualize the reflection path of a ball in ne of our membrane systems. We took a series of photos while we dropped the ball to see if we can record the path but it was not helpful, because of the coulers in the environment it was too hard to see only the path of the ball.(d) So we decided to use a fluoresant ball and record it with a video camera under blacklight.(c) Through testing and recording results with video and photo, we chose to concentrate on one membrane at a time. 9

MEMBRANE SPACES studio course AHO fall 2008

After trying to understand the reflection path created by different membrane systems, we decided to go back to basics to be able to understand them on a more primitive level, so that we would be able to build knowledge about them from the foundation. We decided to work with just one membrane and record all the different reactions that membrane gives to objects with a different weight, angle or speed. We recorded the reflection path of different situations to understand the basic principles of falling object-membrane relationships. The pictures above are reflection path illustrations of a ball dropped on a flat membrane. In each photo a certain property of the ball has been changed to understand the relationship between these properties and the reflection paths. We worked with two different size of balls, 3 different speeds (we used the distance the ball went on the dispenser before it hits the membrane to create different speeds) and the angle between the membrane and the ball.By recording this data visually, we created a table of information that allows us to compare different situations and retrieve information on how we should use membranes to create the reflection path we want to create.

Procedure

three different angles (30,60, 90) three different starting points in each angle

two different sizes of balls

two membranes of different tensions

Setting up the experiment

running the experiment 5 times each

capturing with a videocam

take several frames of each experiment

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne SunguroÄ&#x;lu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YĂ&#x153; CHEN

merge the frames to one pic.

trace the path in a CAD programm and analyse the angles

compare the information gained from the experiments

MEMBRANE SPACES studio course AHO fall 2008

Almost the same table as on the former page but with the information of reflecting angles in different conditions.

Size/ tightness

Loose

90 Fast I

Big A

Slow III

Fast I

Slow III

Tight

MEMBRANE SPACES studio course AHO fall 2008 Heavy - Light Heavy object bounces higher than the lighter one. d

Loose - Tight In a loose membrane, a dropping object bounces higher and with a bigger angle. However when the membrane is tighter, the ball bounces less, with a steeper angle which sometimes creates multiple bounces on the membrane. The only exception to that is the situation where the falling angle of the object is very low and the membrane is loose, then it does not bounce on the membrane but rolls over it instead

Angle Falling angle of the object creates different bouncing patterns. When the angle is 30, it tends to jump out of the membrane directly due to the direction of force applied on the object. So it is hard to say if the angle has an effect on the ??? When the object is sent to the membrane with an angle of 90, it bounces higher as the power is applied in that particular direction but when the angle is lower, it tends to keep going in that angle.

Slow - Fast When the ball has a slower speed, it bounces higher but nearer. This causes the ball to stay on the membrane longer than the faster one.

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne SunguroÄ&#x;lu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YĂ&#x153; CHEN

MEMBRANE SPACES studio course AHO fall 2008

suspender fixed

three different falling objects

Ball 1 - small

two membranes of the same properties suspended in an angle of 60 to each other

Ball 2 - big Man dummy

Then we decided to narrow down the parameters and do another experiment to see the difference in reflection paths of a small ball, a bigger and heavier ball and a man dummy we created. We wanted to see if having more than one piece connected with joints would effect the fall of an object. The graph on the left shows the reflection paths of these 3 different objects

small ball

big ball

dummy for a human( ball with jointed limbs)

MEMBRANE SPACES studio course AHO fall 2008

Digital Modelling For digital modelling we used rhino with a relaxation script for surfaces and meshes. The amount of squads define the smoothness of the resulting object. But it takes also calculating resources. parameters we can set is, geometrical links and tension factor for each point. The modelling of a barrel with interconected points was a real task for us.To have to points linked but not fixed in the space we have to weld them together and relaxe at the same time. To weld the right point to each other takes a long time. Also the shapes turned out not to be the way we expected it. We produced a lot of useless shapes, but learned a lot about digital modelling.

tutors I PROF.MCHAELl U. HENSEL I Prof. Dr. Birger Sevaldson I Defne SunguroÄ&#x;lu // students I MIRAY OKTEM I LIAM ROTE-HAHRE I YĂ&#x153; CHEN

MEMBRANE SPACES studio course AHO fall 2008

Digital Modelling We also tried to simulate a droping ball on a surface. For this we used Autodesk Maya. The first problem was to import a surface into maya wihtout keeping the bounding box. in our first tries the ball alway drops of the bounding box, wihtout touching the membrane. The second task was to make the membrane a textile. We used the nCloth function. The problem was that the n Cloth scrpit can make a surface reacting like a textile butt not as a elastic textile. The mebrane always starts to flutter when we added gravity to it. Also the ball sometime just ignores the membrane and dropes through it.