Both cable and fabric have significant relaxation and creep problems require extra attention and care. Due to construction method of products, these materials have trouble maintaining their original fabricated dimension.
Total stretch of a cable under load is the result of two components: (1) the construction stretch caused by lengthening of the strand lay and subsequent adjustment of the wires in a strand into a denser cross section; and (2) the elastic stretch by a given load. It is recommended to pre-stretching cable before fabrication if no adjustment capability.
Woven fabric consists of yarns in two orthogonal directions, FILL & WARP. WARP yarn usually straight while FILL yarn threads up and down the WARP yarn. Similar to cable, there are stretches caused by construction and elastic. The trouble maker for some fabric material is Creep.
Other than fiberglass, fabric materials are subjected to short term and long term creep phenomenon which will reduce the pre-stress in the fabric and change the whole structural system from design state. To ensure proper pre-stress in the system, adjustability has to be part of the design process. Usually adjustment is required 3 months, 6 months and 12 months after completion.
Higher pre-stress in the fabric is not recommended. The reasons are:
(1) Much higher capacity jack is needed to handle much higher pre-stress.
(2) Longer jacking distance to pull the fabric.
(3) Much higher friction between fabric and cable will occur.
(4) Longer installation period and longer exposure to unknown weather condition.
(5) Larger and more critical deformation in the fabric, especially at corners.
(6) Require additional analyses to simulate higher pre-stress situation.
(7) Above all, the risk is very high.
Large fabric structures in the world such as 广州新白云机场 and Cable Domes usually go with PTFE Fiberglass material to stay away from re-tensioning the system.
Bi-Axial test can help predicting creep behavior if proper stress cycles are prescribed by the engineer. Design engineer needs to implement devises which can adjust fabric stress in the WARP or FILL direction independently.
This is an issue with no simple solution; I suggest if the geometry is controlled as designed and fabric stress is distributed properly, cable force shouldn’t be a problem.
Density of a fabric mesh has to do with the purpose of the model. To quickly evaluate a structural system for member force and reaction, a coarse model is sufficient; to understand local behavior and stress, a fine grid model is desired; for patterning, use grid size close to the actual fabric panel width would be adequate. We have to connect the analytical model to the real structure, otherwise the system only “works on paper”.