Finger-joints with Adherends of Dissimilar Stiness

Since wood is a material showing large variability in its material properties, it is very likely that the two halves of a nger-joint are dissimilar in stiness. Therefore analyses were performed in which the Young's modulus of one half of the nger-jointed lamination was varied. Since the reference value for the Young's modulus is rather high,Ex= 16:8 GPa, this study was performed by letting one half of the nger-jointed lamination consist of a material with lower Young's modulus. The ratios between the orthotropic stiness parameters were kept unchanged so that: Ex=Ey = 30,Ex=Gxy= 16,xy= 0:45.

A series of analyses involving reducedExin the left and in the right half of the model, respectively, were performed using the FE-model shown in Figure 4. The reduction in the elastic constants was 20, 40 and 60%, respectively, corresponding to a modulus of elasticity in the bre direction of 13.44, 10.08 and 6.72 GPa. The stress distribution in the outermost bondline is not exactly symmetric and the case of the left side having a lesser stiness is more strength reducing. This can be explained by looking at the right side of Figure 6: the left half of the bondline is more severely stressed than the right half. Now if the stiness of the left half of the lamination is lower than that of the right half, the stresses transferred by the bondline are concentrated to the left half even more, resulting in the load-bearing capacity of the nger-joint being lowered. It should be noted that the simulations were made assuming the failure of the nger-joints to take place in the bondline and not in the wood itself. Since the strength of wood can be assumed to be correlated to its stiness, it is possible, that large stiness variations tend to lead to wood failure. The results of the parameter study are summarized in Table 2.

The results show that the strength of a nger-joint can be just as much inuenced by only one of the pieces being of low stiness as if the mean value of the stinesses is low.

Table 2: Inuence on the strength, ft, of nger-jointing pieces of dissimilar stiness.

Numerical predictions by FEM. Reference value of Young's modulus isE0= 16:8 GPa.

Eleft=E0 Eright=E0 ftMPa] !ft%]

1.0 1.0 79.9 0

0.8 1.0 69.3 -13

0.6 1.0 57.5 -28

0.4 1.0 46.7 -42

1.0 0.8 75.4 -6

1.0 0.6 61.9 -23

1.0 0.4 47.2 -41

0.8 0.8 73.4 -8

0.6 0.6 65.3 -18

0.4 0.4 54.6 -32

14

6 Conclusions

The following conclusions can be drawn from the present study.

 The strain softening behaviour of a resorcinol-phenol (RP) adhesive bondline cut out from a nger joint was recorded in stable tests for dierent modes of loading.

 A nonlinear fracture softening model was implemented in a nite element software and used to predict the behaviour of a wooden nger-joint.

 The strength of the nger-joint showed to be strongly inuenced by the behaviour of the bondline after peak stress.

 For the examined RP-adhesive there is a large potential of increasing the nger-joint strength by increasing the bondline fracture energy.

 Parameter studies performed showed the nger-joint strength to be strongly de-pendent on the presence and location of small defects (voids) in the bondline, especially in the outermost bondline.

 Parameter studies showed the nger-joint strength to be strongly inuenced by the stiness of the lamination and by dierence in stiness of the two nger-joint halves.

References

1] Colling, F., Tragfahigkeit von Biegetragern aus Brettschichtholz in Abhangigkeit von den festigkeitsrelevanten Einugroen. Berichte der Versuchsanstalt fur Stahl, Holz und Steine der Universitat Fridericiana, Karlsruhe, Germany, 1990.

2] Wernersson, H., Fracture Characterization of Wood Adhesive Joints. Report TVSM-1006, Lund University, Division of Structural Mechanics, Lund, Sweden, 1994.

3] Serrano, E., Finger-joints for Laminated Beams. Experimental and numerical studies of mechanical behaviour. Report TVSM-3021, Lund University, Division of Struc-tural Mechanics, Lund, Sweden, 1997.

4] Aicher, S., Klock, W., Spannungsberechnungen zur Optimierung von

Keilzinkenpro-len fur Brettschichtholz-Lamellen. Bauen mit Holz. Vol. 92(5), pp. 356-362, 1990.

5] Hibbitt, Karlsson & Sorensen, ABAQUS Users'manual, Version 5.5. Pawtucket, RI, USA, 1995.

6] Bla, H. J. et al. (ed.), Timber Engineering STEP 1. Basis of design, material prop-erties, structural components and joints. First edition, Centrum Hout, The Nether-lands, 1995.

7] Gustafsson, P. J., Fracture Mechanics Studies of Non-yielding Materials Like Con-crete: Modelling of Tensile Fracture and Applied Strength Analyses. Report TVBM-1007, Lund Institute of Technology, Division of Building Materials, Lund, Sweden, 1985.

Modeling of Finger-joint Failure in Glued-laminated Timber Beams

Erik Serrano, Per J. Gustafsson and Hans J. Larsen by

Modeling of Finger-joint Failure in Glued-laminated Timber Beams

Erik Serrano1, Per J. Gustafsson2 andHans J. Larsen3 Submitted for publication in ASCE { Journal of Structural Engineering

Abstract

This paper presents a novel approach to the modeling of failure of nger-joints in glued-laminated beams. A nonlinear strain-softening model with stochastic material parameters was used to characterize the failure zone of the nger-joint.

Monte Carlo simulations of the behavior of nger-jointed laminations and of lami-nated beams were performed using the nite element method. Various lamination thicknesses and beam depths were investigated, as well as the eect of varying the ductility of the nger-joint. Finally the use of simplied nger-joint material models was also investigated. The results show that the proposed approach can account for such phenomena as the size eect and the laminating eect. Another observation is that the nger-joint ductility has a major inuence on the lamina-tion and beam strength and that a special case of the present modeling approach can be made to coincide with the classic weakest link theory of Weibull.

Keywords: laminated wood, glued-laminated timber, nger-joint, laminating eect, size eect, tensile strength, bending strength, nite element analysis

1 Introduction

1.1 Background

The present study is part of a research project dealing with the properties of nger-joints and laminated wood products. The overall aim of this research project is to contribute knowledge and methods to the rational modeling, analysis and prediction of the strength of glued-laminated timber (glulam).

To predict the strength properties of glulam two methods are, in principle, possible:

testing of a number of glulam members, using a sample size su ciently large to determine the properties on a statistical basis

using a model, veried by testing, linking the properties of the glulam to the properties of the individual laminations.

1Division of Structural Mechanics, Lund University, P.O. Box 118, SE-221 00, Lund, Sweden. Tele-phone: +46 46 222 95 88. Fax: +46 46 222 44 20. E-mail: Erik.Serrano@byggmek.lth.se

2Assoc. Prof., Division of Structural Mechanics, Lund University, Sweden.

Prof., Division of Structural Mechanics, Lund University, Sweden.

Taking into account the many parameters inuencing the properties of glulam - e.g.

timber species and grade, lamination thickness, the properties of nger-joints, member depth - and the number of specimens needed, the former method is clearly too expensive as a general method. A model which makes it possible to estimate the properties of glulam from those of the laminations is therefore desirable. In addition, a rational model adds to the understanding of glulam behavior.