Pull-out behavior of galvanized steel strip in foam concrete
© Ramezani et al.; licensee Springer. 2013
Received: 23 June 2013
Accepted: 30 August 2013
Published: 27 September 2013
Skip to main content
© Ramezani et al.; licensee Springer. 2013
Received: 23 June 2013
Accepted: 30 August 2013
Published: 27 September 2013
The interface and bond between concrete and reinforcing steel are the most fundamental problems of reinforced concrete structures. In this paper, the pull-out strength of galvanized steel strips with different geometries and hole patterns in foam concrete blocks are investigated experimentally and numerically. Foam concrete mixtures of 1,200-kg/m3 density were obtained by mixing cement and water in a mortar mixer together with ultrafoam as the foaming agent and Quick-Gel as the viscosifier. A theoretical model is developed to predict the bond-slip relationship between the strip and the concrete. This model is further implemented in a finite element simulation of the pull-out tests through the ABAQUS user subroutine. The results show good correlation between experimental, theoretical, and finite element simulation analyses. The influence of the steel strip geometries on the maximum pull-out force is also studied, and it was found that the strips with the greatest hole area, hole diameter, and circumference areas have higher pull-out forces and the increase is nearly linear.
In the industrialized world, new and modern techniques and materials, such as composites, have become common in today's construction industry. These new materials are lightweight, energy efficient, aesthetically attractive, and efficiently handled and erected. Composite structural assemblies (CSAs) are products with performance superior to existing building elements, based on combinations of materials such as embedded light gauge steel components, settable fillers, and coatings of sheet materials to provide a variety of finishes (Akin et al. 2005). The materials are combined or assembled in various ways to provide improved performance in strength and stiffness and associated properties of acoustic filtering, thermal energy conservation, vibration resistance, moisture barriers or absorption, and fire behavior. The structural strength of these products is determined by the grade and thickness of the steel, the configuration of internal panel components, and by the features of infill material (Balendran et al. 2002).
The current paper focuses on novel materials for application in CSAs and the bonding performance between steel and infill materials. The bond between concrete and steel elements is one of the most important properties contributing to the successful functioning of a composite panel (Kayali 2004). The main contribution to bond strength comes from the chemical adhesion and the friction resistance occurring between the steel and the concrete as a result of the surface effects. Composite structures made of various materials can fail in several ways depending on material properties of components, design methods, and loading cases, but the most important is the cooperation of all elements (Alterman et al. 2011). The lack of compatibility between elements leads to bond failure, sliding of reinforcement bars and/or steel strips, local deformations, and finally cracking. Destructive measurements of shear strength through pull-out and push-in tests are commonly used methods to assess the quality of a connection between steel elements and fillers, e.g., concrete.
There are only few studies available in the literature about the structural bond properties between lightweight concrete and steel bars with pull-out test (see e.g., Cao and Chung 2001; De Lorenzis et al. 2002; Chang 2003; Al-mahmoud et al. 2007). Ben Romdhane and Ulm (2002) investigated the steel-concrete interface with computational mechanics. They modelled micro-to-macro mechanisms of the interface and studied the numerical implementation of this model in combination with a discrete crack model for concrete. Few finite element (FE) simulations of pull-out tests are also available in the literature (e.g., Wu et al. 2009; Wei et al. 2012), yet without accurate modelling of the interface and bond behavior. Casanova et al. (2012) presented a new finite element approach to model the steel-concrete bond effects. Their model proposed to relate steel, represented by truss elements, with the surrounding concrete in the case where the two meshes are not necessarily coincident. Most of the previous investigations about bonding behavior focused mainly on experimental and theoretical studies of shear stress between reinforcing bars and concrete specimens (e.g., Bouazaoui and Li, 2008; Fang et al. 2006; Khandaker and Hossain 2008), and plastic bars and concrete samples (e.g. Won et al. 2008), but not directly on steel strips, which are applicable to this project. This study investigates bonding performance and analyzes it through pull-out tests to provide data for the theoretical and FE analysis of CSAs. Because bonding is a general problem for wall panel design and other uses of concrete that employ steel to concrete contact zone, this research undertakes an analysis of the influence of various steel strip designs to understand bonding behavior. With the obtained results, the research enables understanding on the main features of metal face materials commonly used in wall panels.
The force in the strip at x = L corresponds to the pull-out load P, and therefore, we have s′(L) = q(L) = β P. Now, the initial value problem (Equation 9) can be easily solved in an iterative process for any given T(s) and δ, with Runge-Kutta numerical integration procedure, and the normalized pull-out force can be determined for a given displacement d, such that the boundary condition at the bottom of the strip s′(0) = 0 is satisfied.
Foam concrete mixtures were prepared at 1,200-kg/m3 density by using stable foam concrete compositions, which were designed based on the investigation of Kearsley and Mostert (2005). Properties of the foam concrete were analyzed for density and compressive strength based on the standard testing methods for physical and mechanical properties (ASTM C495 and ASTM C39). Experiments showed that by using the theoretical design method of Kearsley and Mostert (2005), it is possible to predict the compressive strength and density for the concrete mixtures. Ordinary Portland Cement was used for the foam concrete mixtures. To prepare the foam, ultrafoam was used as the foaming agent and Quick-Gel (Baroid, Houston, TX, USA) as the viscosifier. These were mixed with water in the foam generator until the foam bubble size was uniform and stable (usually 2 min). Then, the foam was added to the cement matrix while stirring, with continual mixing for 1 to 2 min. When the mixture of foam and cement matrix was uniform, the prepared foam concrete was poured into the test mold with slight vibrations to fill up the mold completely. Specimens were demolded after 24 h and then moist cured in a standard curing room for a further 28 days in order to test the samples with a standard testing machine.
Standard cylinders of 100-mm diameter were used for the compressive strength tests, which were carried out in a testing machine of 100-kN capacity at a loading displacement rate of 0.1 mm/s. Three cylinders were tested for compressive strength after 28 days according to the ASTM C39 standard test method for compressive strength of cylindrical concrete specimens, and the mean value was calculated. The average compressive strength of the foam concrete was 8.8 MPa.
The pull-out test method was used to evaluate the shear bonding strength between the steel and the foam concrete. All the pull-out tests followed the ASTM C900 standard. The ram velocity for all the tests was set at 0.2 mm/s to allow slow deformation of the concrete and the steel strips. The pull-out test samples include a foam concrete cube (100 mm × 100 mm × 100 mm) with galvanized steel sheet (50 mm × 150 mm) standing in the middle of the concrete cube. A hot-dip galvanized steel sheet, 0.75-mm thick, grade G250 from New Zealand Steel (Private Bag, New Zealand) was used. Steel sheets were pre-cleaned and degreased by using soapy water without damaging the coating before getting it in contact with the fresh concrete.
Relationships of hole patterns
Number of holes
Radius of holes (mm)
Area of holes (mm2)
The pull-out test was further studied by finite element simulations. A three-dimensional finite element model was developed using ABAQUS release 6.12 (Simulia, Providence, RI, USA) for this purpose. This is an implicit analysis code suitable for modelling both material and geometric nonlinearities and has good contact modelling. The FE analysis comprised an elastic-plastic simulation of the pull-out process and the determination of the load-displacement curve in the steel strip. The FE simulations were conducted for all the hole patterns depicted in Figure 3, and the FE results were compared with experimental and theoretical results. The material data for FE simulations were obtained from compression test of foam concrete and tensile test of steel strip. The galvanized steel sheet had the elastic modulus of 200 GPa and Poisson's ratio of 0.3 with the ultimate yield stress of 357 GPa. The foam concrete had a density of 1,200 kg/m3, Poisson's ratio of 0.2, and compressive strength of 8.8 MPa. The elastic modulus of the foam concrete was 5,535 MPa.
In the FE simulations, the bond stress-slip (τ - s) relations obtained from experiments and theoretical modelling were used as the input data for the FE model. Bond stress-slip relations are considered as the constitutive relations for the behavior of the contact interface. The pull-out tests were then simulated by applying the shear constitutive behavior to the FE model by a user-defined subroutine. UINTER user-defined subroutine was used to define the interfacial constitutive behavior which proposes a new numerical model for simulating the bond-slip behavior of the composite structure. User subroutine UINTER is called at points on the slave surface of a contact pair with a user-defined constitutive model defining the interaction between the surfaces. The subroutine can be used to define the mechanical (normal and shear) interactions between surfaces. In this way, any special or proprietary interfacial constitutive behavior could be defined through the user subroutine UINTER written in FORTRAN in ABAQUS/Standard, which may include both the normal and tangential behaviors. This feature provides a strong tool for simulating the bond-slip behavior. User subroutine UINTER was called for each contact constraint location of affected contact pairs in every increment of the ABAQUS/Standard analysis. The input to this user subroutine includes the current relative position of a particular constraint point on the slave surface with respect to the corresponding closest point on the master surface, as well as the incremental relative motion between these two points. The normal direction constitutive behavior was modelled by using the penalty approach once the interaction surfaces are in contact; otherwise the normal strength is zero.
The concrete damaged plasticity model was assumed for the concrete. The concrete damaged plasticity model provides a general capability for modelling concrete and other quasi-brittle materials in all types of structures. This model uses concepts of isotropic damaged elasticity in combination with isotropic tensile and compressive plasticity to represent the inelastic behavior of concrete. The concrete damaged plasticity model is based on the assumption of scalar (isotropic) damage and is designed for applications in which the concrete is subjected to arbitrary loading conditions. The model takes into consideration the degradation of elastic stiffness induced by plastic straining both in tension and compression. It also accounts for stiffness recovery effects under cyclic loading.
The node-to-surface contact between the strip and the concrete was modelled through the master-slave algorithm. General contact automatically assigns master and slave roles for contact interactions, and ABAQUS calculates an overclosure tolerance based on the size of the underlying element facets on a slave surface. Slave surfaces (or nodes) in a particular interaction are repositioned onto the associated master surface (or nodes) if the two surfaces are initially overclosed by a distance smaller than the calculated tolerance. Initial gaps between surfaces remain unchanged by default adjustments. If a portion of a slave surface is initially overclosed by a distance greater than the calculated tolerance, ABAQUS automatically generates a contact exclusion for this surface portion and its associated master surface. Therefore, general contact does not create interactions between surfaces (or portions of surfaces) that are severely overclosed in the initial configuration of the model, and these surfaces can freely penetrate each other throughout the analysis.
The results of the pull-out tests between the steel strips and the aerated concrete, and theoretical and FE analyses are presented and discussed here. The main purpose of the pull-out experiments was to find relationships between ten-hole patterns and mechanical properties of aerated concrete in order to evaluate and predict the bonding between steel and lightweight concrete through a numerical simulation.
Three specimens for each single set (three identical samples) of strip design were prepared and tested to determine the bonding strength during pull-out tests. The reference hole area was 100.6 mm2 with a 11.32-mm diameter. The reference area was increased gradually in order to verify the effect of the anchorage of the concrete embedded into holes in various designs.
The goal of this paper was to describe bonding processes between steel and concrete specimens experimentally and by applying analytical and FE methods. The results of investigations on the bonding behavior between lightweight concrete and steel strips enable a better understanding on the use of lightweight concrete in wall panels. The results show that the experimental, theoretical, and FE simulation results match very well. Three geometrical parameters affecting the pull-out force were also investigated in this paper, namely the total area of the holes, the diameter of the holes, and the circumference area of the holes. The diameter of the holes or circumference area of the holes was found to give stronger correlation with the pull-out forces; however, weaker correlations exist between the pull-out forces and the total area of the holes. The strips with the greatest hole area, hole diameter, and circumference areas had higher pull-out forces, and the increase is nearly linear.
The authors like to thank the Foundation of Research, Science and Technology (FRST) of New Zealand for the generous financial support received during the investigation.
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.