Design of fabric membrane roofs

I have just started studying the design of a fabric membrane roof with membrane and cable prestressing. It's not clear to me that:

1) From an architectural drawing of the final shape of a membrane structure, how to get an undeformed (stress-free or cutting) shape of the membrane? I'm understanding that
the specified final shape is under the action of membrane selfweight, membrane prestress, and cable prestress.

From the calculation of a fabricator, it seems to me that they starts from the projected shape of the structure. Then, using their inhouse geometry-analysis software, a geometry is obtained under uniform
tension that matches a set of specified geometric constraints. All the subsequent stress analyses are on the obtained geometry. Is this the usual practice? If it is, then
2) What is the theory behind the form-finding to get the uniform tension geometry that matches the specified constraints?
3) Varying the constraints will produce different stress-free shapes, which then under selfweight and prestressing will give different final shapes. How to obtain the constraints that give the stress-free shape which deforms to the required configuration under selfweight and prestressing?

 

[Ingenuity]

muttum,

I do not think i can answer all your questions but the following may be helpful...

The analysis of tensioned fabric structures is a non-linear problem. Actually, the materials used are also highly non-linear, but fortunately the materials are approximately linear for the allowable stress range of most structural membranes.

Scaled models using stretched fabric or soap films are often used to investigate the initial behavior, and the initial shape or form.

Load analysis is done from a starting shape that is in equilibrium with the initial prestress. Form finding is the process of determining this initial equilibrium shape, and finding this unique shape compatible with the given prestress usually requires a nonlinear computer analysis.

Most companies that design and constructed these specialized structures (eg Birdair) have in-house software that somewhat automates the form finding process. Such programs use techniques such as non-linear stiffness analysis, dynamic relaxation, or force density to converge on the equilibrium shape.

It is a specialized subject area, and you are probably best to do a literature search on the topic. Guru's like Horst Berger, Frei Otto and Peter Rice have written substantial works on this topic, including topics covering the art and architecture of these structures.

Also the ASCE has a publication called "A Practical Introduction to Tensioned Fabric Structures" that may prove useful to you.

HART.

 

Thank you very much for your response. I got some docs on membrane structures but they don't give much info about the design process. I'm studying a steel structure being attached with fabric membranes and stainless cables and trying to verify the reaction from membrane fittings and cables given by a fabricator. My problem is about the starting shape to do nonlinear FE analysis as you mentioned. Will try to get more docs as suggested.

 

[ishvaaag]

2 references

Fabdes on CDROM- A Design and Educational Resource for Tensioned
Fabric Structures
Dr Peter Kneen, Consultant
Engineering Education Development Project
Jakarta, Indonesia

(a downloadable pdf, search in google)

http://planeta.terra.com.br/educacao/varcaro/membrane/

 

[Ingenuity]

Dr Peter Kneen is the author of FABDES and was a professor at the University of New South Wales, Sydney, Australia.

FABDES does have form finding capabilities, but it is written in FORTRAN and the input and output is a bit cumbersome - not graphical input etc. His notes that are available on the web (as per ishvaaag post) are good and worth reading but Peter no longer works on supporting FABDES, so you may get some good info, but you probably will not be able to get any additional support.

I recently recevied a sample copy of STRAND7 FEA software(

http://www.strand7.com)

where i asked for a sample of its fabric membrane modelling capabilities and they provided me with a example that is able to model the fabric structure, with prestress, and form finding capabilities.

HTH

 

[chandr]

Thank you very much for all inputs. Ingenuity. Could you mail me the Strand7 example? I had few experiences with Strand7 in the past and would like to take a look how it looks like.

tomtumx@yahoo.com

 

Ingenuity.

I've contacted STRAND7's dealer and got a pdf file showing how to model staged construction and a membrane example but doesn't show any form finding capability. The model I got contains only 2 fabric mebranes stretched among 6 circular steel pipe columns with cable along membranes' edge. Could you mail me the example you got or tell me the title of the model file?