Computational & Technology Resources
an online resource for computational,
engineering & technology publications |
|
Civil-Comp Proceedings
ISSN 1759-3433 CCP: 75
PROCEEDINGS OF THE SIXTH INTERNATIONAL CONFERENCE ON COMPUTATIONAL STRUCTURES TECHNOLOGY Edited by: B.H.V. Topping and Z. Bittnar
Paper 91
Optimal Design of Tensile Fabric Structures T. Nouri-Baranger+ and P. Trompette*
+Universite Claude Bernard-Lyon 1, Villeurbanne, France
T. Nouri-Baranger, P. Trompette, "Optimal Design of Tensile Fabric Structures", in B.H.V. Topping, Z. Bittnar, (Editors), "Proceedings of the Sixth International Conference on Computational Structures Technology", Civil-Comp Press, Stirlingshire, UK, Paper 91, 2002. doi:10.4203/ccp.75.91
Keywords: fabric, cable, tension, form finding, sensitivity analysis, optimisation.
Summary
The final design of a tensile membrane is achieved when the stress
states generated by different and normalised climatic (snow,
wind…) loadings have been deemed as acceptable by the official
control bodies. For this reason, and since in tensile structures
the shape and the stress states are strongly related, the whole
design may result from a repeated iterative process between the
designers, i.e. the firm of architects and the research
consultants. This motivates the presentation in this paper of an
optimisation tool devoted to fabric structures which could be used
to ease the design process in decreasing the number of
back-and-forth interactions between all the different
participants. First a specific and new shape form finding
procedure is proposed; taking into account the usual geometrical
constraints or data requirements, but also a desired biaxial (warp
and weft) non uniform stress state, an (optimal must be
understood in the sense: nearest as possible) initial shape is
found. Then the climatic loads are applied and the current loaded
positions calculated. A sensitivity analysis of these equilibrium
positions depending on different parameters such as the positions
of the anchorage points and cable tensions, allows one to modify
with efficiency the shape and/or the stress state in order to
better satisfy the requirements. This optimisation loop is
included in a larger one that takes into account possible
modifications of the cutting patterns, which is an other way of
optimising the stress states. The efficiency of this procedure is
demonstrated on an actual example.
In the design of a tensile fabric structure the first step is called shape finding. The usual data given, by the architect are the coordinate list of several fixed points : the mast position(s) and the anchorage points. One other main data set is the tension values of all the boarder or intermediate cables and the warp and weft tensile optimal stresses in the fabric. From these data an equilibrium position is found by solving the non linear matrix equations. is the geometric stiffness matrix, the nodal displacement vector and the optimal stress state which is defined as follows :
Any initial geometry of the membrane may be chosen but a definition coupled with a finite element surface generator of the whole structure is a very convenient starting point (a simpler one is a flat mesh). The shape is easily varied in modifying the values. Constraints on maximum and minimum values of are taken into account. As soon as the initial shape is determined a non linear finite element analysis of the structure is performed using cable and orthotropic membrane elements to evaluate the stresses and the shape of the loaded structure. From this analysis, it may be necessary to modify the initial shape if over or under stressed parts which generate wrinkles or pockets are found. In order to ease the designer work, a sensitivity analysis of this equilibrium position may be performed; the design variables are: the tensions in the cables and the positions of the anchorage points. Two objective functions have been defined: one is concerned with the nodal displacements to minimise their total norm and the second with the stresses to maximise the minimum stress areas and to minimise the maximum stress areas. The application of this optimisation procedure yields a definition of an optimal structure shape as close as possible to the architect design. It also can be used to take into account of the influence of uncertainties on the implementation in situ of the structure When the three dimensional shape is defined there remains to define the cutting patterns. The flattening process uses a specific method. To take into account the desired stress state in the three dimensional fabric i.e. to correct the patterns obtained, an iterative stress calculation process between the and the shapes is performed. The example presented in this paper is a swimming pool roofing. The covered surface has the dimensions: / / . The roof is made up of rows of inclined masts ( hoops) and the fabric is divided in pieces. The finite element mesh is made of triangular linear membrane elements and linear cable elements. References
purchase the full-text of this paper (price £20)
go to the previous paper |
|