Numerical modeling of pull out test for composite slabs using steel deck

Melchor López Ávila
Puerto Padre, Cuba
melchor@uclv.edu.cu, melchor_bia@yahoo.com

Rafael Larrúa Quevedo
Carlos A. Recarey Morfa


ABSTRACT

Numerical modeling is used to reproduce pull out tests that are carry out at composite slabs in small scale to determine shear longitudinal forces between metallic sheet and concrete slab. Is been considered an elastic- plastic model for the steel and a plastic damage model for the concrete. Is used ABAQUS/Implicit 6.6 Commercial Code to simulate numerically using FEA the pull out test. Is conceptually defined the modeling and the numerical model calibration taking as basis experimental studies. Are established the parametric study basis and an analytic method to estimate longitudinal shear.


1. INTRODUCTION

Composite slabs are very useful structural forms, where are combined steel deck and concrete block for working together. Usually, structural fail occur when is broken the longitudinal shear resistance between concrete block and steel deck, and slip is producing, losing composite work of these materials. Chemical bond, friction and mechanical interaction are the main factors that make up longitudinal shear resistance between steel deck and concrete slab.

To characterize composite slabs usually are carry out experimental tests, loading singles spam slabs to quantify flexion fail (EN 1994-1-1, 2004, NRMC 082, 2004, ASTM E8-00b, 2001, CSSBI S3-2002, 2003), or pull out small specimens of cyclic rib width to estimate longitudinal shear resistance between reinforced concrete slab and metallic sheet (Daniels, 1988, Guex, 2002, Edder, 2003).

Numerical modeling of composite slabs had been analyzed by many researchers (Widjaja, 1997, Veljkovic, 1998, Schuurman, 2000, Edder, 2003, Abdullah, 2004, Ferrer, 2006, Mistakidis, 2007). Using different forms, but always try to reproduce longitudinal shear effect. Are outstanding the Abdullah and Ferrer studies results. Although Mistakidis is more recent research, he focused in bending behavior of there composite structures.

Abdullah use ABAQUS/Explicit commercial code, and a configuration of sheet without embossments. He represented numerically a steel deck rib, along the complete length, using shell finite elements for steel deck, and three-dimensional solid FE for concrete slab (Abdullah, 2004). Interaction between these surfaces and embossments work is simulated using connectors with lineal rigidity. These simplifications avoid realize studies of embossments geometry contribution in the mechanical interaction, as to as local distribution of tensions in embossments wall and closeness. With these assumptions in the modeling of the problem, is not possible estimate with trustworthiness the chemical bond, friction and mechanical interaction phenomena.

Ferrer use ANSYS commercial code and he assumed Coulomb rigid friction model without initial adherence, dynamic effects or top tangential tension (Ferrer, 2006). For represent the steel that conform the sheet, is employed a linear elastic plastic model and the concrete of the slab is simulated considering it as an infinity rigid surface, that restrict evaluate failure mechanisms in the concrete, as breach that is noted during experimental studies. The author takes advantage of geometrical symmetry cooperatively with loads application form in the test, as to as the cyclic embossments patron along the specimen length. He shapes the steel deck using finite elements of shell with reduced integration family, considering the sheet modeling by your axis. The load is represented by means of a longitudinal displacement imposed over all nodes of each of boundaries transversal at sliding direction (Ferrer, 2006). This simplification feigns avoid evaluate completely local bending that can appear in edge sheets due to loads application respect real test.

In this work is developed a numerical model for ABAQUS/Implicit commercial code that take advantage of Ferrer’s symmetry proposals, but with the novelty of employ damage constitutive model for the concrete material. To modeling loads, must be simulated the pull effect y the self- weight effect taking into account the reference position in which is practice the test (Figure 1). The traction force is represented across uniform load (in area) over side of concrete block transversal at displacement direction. To simulate self- weight loads are applied laterally al concrete block (Figure 1). As steel deck many as concrete block are modeling geometrically using three- dimensional finite elements.