Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.426592
Title: Modelling aspects of the influence of edge effects on expansion anchors
Author: Watson, David Stewart.
Awarding Body: University of Glasgow
Current Institution: University of Glasgow
Date of Award: 2006
Availability of Full Text:
Access through EThOS:
Abstract:
The principal aim of this work was to investigate and develop modelling techniques capable of accurately and robustly analysing expansion anchor bolts in concrete under tensile loading. Of special interest was the influence of low edge distances on such devices. Since the 20th century increased demands for flexibility, safety and cost have lead to significant development of new anchor products. Modern design methods for new products follow a scientific approach but still rely on substantial and expensive programs of experimental testing. Current design methods for structural designers using anchors are based on semi-empirical approaches derived from extensive experimental testing. It is proposed that much of this experimental work can be replaced with numerical modelling. A number of suitable finite element constitutive models are considered. Initially a Multisurface Plasticity Model and a Traditional Crack Model using a Multiple Fixed Crack (MFC) formulation are considered. Both are shown to give satisfactory results when used to analyse a common, plane-stress benchmark problem. However, although the Plasticity Model gave a better post peak response a 3D implementation was not available within the chosen FE framework. Spurious stress accumulation was identified as the cause of the problems with the MFC Model and its various causes are investigated in detail. A Total Strain Based Rotating Crack Model was chosen as an alternative constitutive model and together with suitable modelling parameters was able to reduce these spurious stress accumulation effects to an acceptable level. 3D modelling of a non-expanding, fully bonded anchor at various distances to the free edge accurately predicted the expected reduction in strength and compared well with reduction factors supplied by anchor manufacturers. The study was extended to include the effect of two free edges and results allowed the strength reduction to be calculated for any arbitrary position rather than for just the single edge approach given in the anchor design guides. Modelling of anchor expansion was tackled on two fronts. Firstly anchor-concrete interfacial behaviour was considered. A Coulomb Friction Model applied to zero thickness structural interface elements to simulate the pressure dependant frictional bond. The role of FE model geometry and material properties in producing a realistic interfacial stress profile was studied in detail. For the kinematics of the expansion modelling of the expander mechanism as a contact problem was found to be the most accurate approach. However, limitations of the modelling framework required that the contact analysis be performed separately and resulting contact stress profile be applied to the existing, noncontact problem. This approach, although somewhat inflexible, provided a useful insight into the important factors pertaining to both the geometric and constitutive models. Results showed realistic crack patterns and demonstrated the effect of varying expansion pressures on the structural response of the anchor bolt. The modelling approach used in this study was highly complex in terms of the multiple non-linear material models and the associated solution process. This resulted in problems with robustness and stability. As an alternative and inherently stable modelling framework a Sequentially Linear (SL) Model was developed. In its isotropic form it proved fast, accurate and reliable for plane-stress anchor problems. Orthotropic fracturing and 3D analysis capabilities were introduced to the model and a number of rules for crack initiation and orientation were tested. Although limitations in the possible crack orientations produced significant mesh bias to the crack pattern, the model was able to capture the changes in anchor behaviour associated with reduced edge distance. The overall assessment is that that SL Model has great potential especially for highly nonlinear problems where stability and robustness are issues
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.426592  DOI: Not available
Share: