Characterising the fatigue response of corrugated roof cladding

Lovisa, Amy (2015) Characterising the fatigue response of corrugated roof cladding. PhD thesis, James Cook University.

[img]
Preview
PDF (Thesis)
Download (9MB) | Preview
View at Publisher Website: https://doi.org/10.25903/541j-6a80
 
290


Abstract

Current fatigue design of roof cladding is entirely reliant on complex and repetitive prototype testing. Standardised cladding fatigue design has successfully minimized the occurrence of fatigue failure during severe wind storms such as cyclones, but the prototype tests required in fatigue design are costly and time consuming. All variations of the cladding assembly, including the various profiles, manufacturers, fasteners and fastener arrangements, must be tested multiple times. A greater understanding of the mechanism underlying fatigue of roof cladding in combination with a numerical model capable of simulating cladding fatigue could reduce the total reliance on prototype testing in fatigue design. Such a numerical model would employ fracture mechanics to simulate crack initiation in the cladding and could to some extent replace aspects of the prototype test or enable a semi-empirical approach to fatigue design.

This thesis investigates the mechanism of fatigue crack initiation in corrugated roof cladding in a bid to develop a model capable of simulating this phenomenon numerically. Characterising fatigue crack initiation required the development of a specialized testing arrangement which isolated a single cladding-fastener connection. The test, known as the isolated fastener connection (IFC) test, restrained all four edges of a 720 mm × 150 mm section of corrugated cladding with a fastener located at the centre displacing to induce load. The IFC test successfully recreated the response of the cladding that was observed in full scale testing when subject to both monotonic and cyclic loads. In particular, the IFC test reproduced the two common crack formations observed in cladding fatigue, the 'star' type crack and 'H' type crack, and produced the cracks after a comparable number of cycles.

A unique and inexpensive photogrammetric method was developed to measure the displacement of the cladding sample within the IFC test when subject to monotonic loads. The photogrammetric method involved texturizing the cladding surface and taking 70 photographs of the cladding at distinct angles. These photographs were passed through commercial photogrammetry software to produce a 3D digital reconstruction of the cladding surface. The displacement of the cladding was then calculated by comparing the digital reconstruction of the deformed cladding with a digital reconstruction of the cladding under no applied load. The digital surface models produced using the photogrammetric method was also used to extend numerical model validation studies. The digital reconstructions were compared directly with the deformed mesh of a numerical model to provide a full-field comparison of the displacement of the cladding sheet.

The IFC test was employed to study the mechanism underlying crack initiation. Corrugated cladding samples were cycled under constant amplitude loading to form a 'star' type crack. The fastener hole edge was monitored by both digital microscopes and strain gauge rosettes. The longitudinal arm of the 'star' type crack opened in the underside of the cladding in regions of tensile strain. The crack proceeded to propagate through the thickness of the cladding along the plane of maximum shear. Once the thickness of the cladding had been entirely bisected, the crack then rapidly opened into the plane of the cladding. A similar mechanism was observed for the transverse leg of the crack. Crack initiation was then defined as the point in which the crack had entirely bisected the thickness of the cladding and had begun to open into the plane of the cladding. Under a number of load amplitudes this point of crack initiation occurred at a crack length of 0.9 mm and 0.5 mm for the longitudinal and transverse cracks respectively. Based on this definition of crack initiation, the strain field at the crack tip at crack initiation was monitored to produce a strain based crack initiation criterion. A crack initiation criterion based on the principal surface strains and distinct to the longitudinal and transverse directions was found.

Attempts were also made to identify the kinematic material properties of the cladding steel. Anti-buckling fixtures were developed to enable tension-compression testing of the thin steel sheet but the cladding softened too much and continued to buckle under compressive loads. Preliminary results suggest the cladding is prone to significant cyclic softening and the nonlinear combined kinematic and isotropic hardening model is required to accurately describe the kinematic material properties of the steel in a numerical simulation.

A numerical model of the cladding subject to monotonic loads in an IFC test was also developed. The model accurately predicted the fastener reaction, strain surrounding the fastener hole and the displacement of the cladding sheet. Under a maximum applied load the cladding predicted the displacement of the cladding to within 89% of that documented experimentally. Development of the model involved a mesh convergence study, in depth analysis of the boundary conditions and compression studies of the EPDM (ethylene propylene diene monomer) seal. The boundary conditions of the model were simplified to reduce the simulation run time making the model more appropriate for a fatigue study. These simplifications significantly reduced the computation cost of the simulation but performing a fatigue simulation was still too costly. Furthermore, a fatigue model requires the kinematic material properties of the steel which remain unknown.

This study identified the mechanism underlying crack initiation in corrugated cladding and characterised the phenomenon using a strain based crack initiation criterion. This investigation also successfully developed a numerical model capable of simulating the response of the cladding subject to monotonic loads. This crack initiation criterion and validated numerical model provide a strong foundation for future development of a numerical model capable of simulating cladding fatigue. Development of such a numerical model is likely to be an iterative process combining both experiments and numerical simulations. A model capable of simulating cladding fatigue could directly reduce the total reliance on prototype testing in cladding fatigue design by either replacing components of the prototype test or enabling a fracture mechanics based analysis of cladding fatigue.

Item ID: 44647
Item Type: Thesis (PhD)
Keywords: bonded metal roofing; corrugated roof cladding; corrugated roofing; crack growth; crack initiation; cracking (material); fatigue (material); fatigue analysis; fatigue studies; fatigue testing
Related URLs:
Additional Information:

Publications arising from this thesis are available from the Related URLs field. The publications are:

Lovisa, A.C., Henderson, D.J., and Ginger, J.D. (2015) An inexpensive method for measuring deformation of corrugated cladding using close range photogrammetry. Experimental Mechanics, 55 (3). pp. 599-609.

Lovisa, Amy C., Wang, Vincent Z., Henderson, David J., and Ginger, John D. (2013) Development and validation of a numerical model for steel roof cladding subject to static uplift loads. Wind & Structures, 17 (5). pp. 495-513.

Lovisa, A.C, Henderson, D.J., Ginger, J.D., and Walker, G. (2016) Characterising fatigue macrocrack initiation in profiled steel roof cladding. Engineering Structures, 125. pp. 364-373, 125. pp. 364–373.

Date Deposited: 11 Aug 2016 04:45
FoR Codes: 09 ENGINEERING > 0905 Civil Engineering > 090506 Structural Engineering @ 100%
SEO Codes: 87 CONSTRUCTION > 8702 Construction Design > 870201 Civil Construction Design @ 50%
87 CONSTRUCTION > 8703 Construction Materials Performance and Processes > 870302 Metals (e.g. Composites, Coatings, Bonding) @ 50%
Downloads: Total: 290
Last 12 Months: 13
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page