Comparative analyses between a nonlinear response composite structure and a linear response structure.

Petrescu, Horia ; Hadar, Anton ; Vlasceanu, Daniel 等

1. INTRODUCTION

The engineering importance of hybrid structures made of metal-laminated composite is obvious. One speaks about problems concerning the methodologies, modeling, analysis, processing, interpretation and capitalization of the obtained results, with practical conclusions for researchers and engineers.

One considers that the composite material is manufactured by specialized companies, with safe technologies that guarantee the mechanical, physical, chemical and elastic properties of the composite, ensuring thus an optimum use by the beneficiary. For modeling and analysis of such structures, many consider taking into account only the characteristics given to us by the manufacturers of by specific laboratory tests. Why not consider the composite materials as an individual substructure and create a model of our material down to its components ... the fiber and the resin?(Gheorghiu et al., 1999)

2. INTERFACES

Considering what had been said in the introduction, the problem for analysis would appear on the touching faces of the resin and the fiber. Another problem is possible to appear at the interface composite-metal. For this, why not create a model and analyses of the glue used between the two substructures (Sorohan et al., 2003)

3. MODELING AND ANALYSIS

In the case of hybrid metal-laminated composite structures, a different situation appears: experimental determinations of the mechanical strength of the junction and the adhesive layer between composite and metal are necessary Gheorghiu et al., 1998.

The idea of modeling the adhesive layer using the finite element method in order to determine its behavior is questionable and inefficient due to the following reasons:

1. The model should be elaborated at the scale of the thickness of the adhesive layer and the height of roughness of the contact surfaces, namely hundredths of millimeters.

2. The configuration of the contact surface and the thickness of the adhesive are strongly irregular and random, thus involving a very laborious modeling

3. The obtained information may not be capitalized since the mechanical behavior of the adhesive at such a scale is not known (its characteristics are global, at the scale of the whole structure).

Considering this why not transform our two substructure system--composite-material (C-M), into a three substructure system composite-adhesive-material (C-A-M), even more into a four substructure system resin-glass-fiber-adhesive-material (R-GF-A-M).

4. RESULTS

In order to detail the concepts and ideas presented above, there were two structures that had been modeled and analyzed using the finite element method, one of them a metal junction and one a hybrid R-GF-A-M structure. One had in view that these variants are comparable from the point of view of dimensions, materials, loads, constraints, mesh, etc. Thus, the results may be comparable also in order to draw conclusions with a certain level of generality and useful for the design engineers (Jiga, 2003). In Fig. 1 and 2 the following are presented:

--The geometry of the model;

--The direction of detailed analysis--

[FIGURE 1 OMITTED]

[FIGURE 2 OMITTED]

The values of the elastic constants of the three materials of the studied structures are listed in Tab. 1.

[FIGURE 3 OMITTED]

In Figs. 4, 5, 6 and 7 are presented the variation of the stress according to the directions shown in figure 2:

--fig. 4 corresponds to direction A (fiber)

--fig. 5 corresponds to direction B (composite)

--fig. 6 corresponds to direction C (adhesive)

--fig. 7 corresponds to direction D (structure)

[FIGURE 4 OMITTED]

[FIGURE 5 OMITTED]

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

The stress map shown in figure 8 is the distribution of geometrical identical structure but made of aluminum. All of the directions defined before present only linear stress distribution. Figure 4 gives us a pretty good idea about the nonlinear stress distribution at the interface between glass-fiber and resin (Skuda, 1998).

[FIGURE 8 OMITTED]

5. CONCLUSIONS

1. For the considered junction, the most simple possible, a complex stress state was obtained with a clear nonlinear character for the stress <ix;

2. A strong stress concentration effect may be noticed from the obtained stress field, in the adhesive layer (see details and results in Fig. 6);

3. The aluminum structure shows no nonlinear problems (see figure 8) as can be seen in the composite layered junction (see figure 3)

4. A great volume of valuable information about the junction is obtained through finite element modeling and analysis. This is not always sufficient. As it was shown in this paper, supplementary experimental investigations may be necessary

6. ACKNOWLEDGEMENTS

"University Politehnica of Bucharest" financed through the P.O.S.D.R.U. program

7. REFERENCES

Gheorghiu H., Hadar A., Constantin N. (1998). Analysis of anizotrop and isotrop structures, Publisher Printech, Bucharest, ISBN 973-9402-23-2

Gheorghiu, H., Ivan, M., Hadar A. (1999). Mecanique des solides non deformables et deformables, Printech ISBN 973-9402-92-5

Jiga, G. (2003). Fundamentals in composite materials structures (in Romanian), Ed .Atlas Press, Bucharest, Romania

Sorohan, S., Constantinescu, I.N. (2003). The modelling and finite element analysis practice (in Romanian), Ed. Politehnica Press, Bucharest, Romania

Skuda, A.M. (1985). Micromechanics of failure of reinforced plastics, Handbook of Composites, Elsevier Science Publishers B. pp 1-69

Tab 1. Material characteristics

 Elastic modulus Shear modulus Poisson's
Material [E.sub.x], [MPa] [G.sub.xy] [MPa] ratio [v.sub.xy]

Aluminum 70000 25000 0.35
Glass-Fiber 244 3800 0.183
Resin 2440 1200 0.46
Adhesive 244 120 0.4