JEC COMPOSITES MAGAZINE - Issue #48 - May 2009 - 33

Design of a fewer angle-ply definition in a laminate reduces the minimum gauge thickness, thus increasing the versatility of the laminate. The data set approach Another design practice commonly used in the industry is generating allowable data using notched specimens. Tests performed in this approach are both material and geometry dependent. This necessitates a significant number of tests to characterize the behaviour of notched specimens, thus resulting in a significant expenditure to set up the specimens and conduct the tests. Exacerbating this situation, data obtained from one set of specimens cannot be superimposed to those from another set of specimens. For example, data generated using an open-hole specimen under unidirectional tension and compression cannot be used to predict the behaviour of such specimens under biaxial loading. Figure 5 shows the ultimate failure envelopes for a smooth and a notched (open-hole) quasi-isotropic specimen generated using the quadratic failure criteria and finite element simulation. From these failure envelopes, it can be construed that the biaxial strength cannot be estimated by those obtained from uniaxial load tests. In addition to this, the safety factor value changes from one load condition to another. This raises a flag against a common practice in the industry of using uniaxial tension and compression data to estimate biaxial data and of using a knockdown factor to obtain data for one specimen from the test results for another specimen. It is evident from Figure 5(b) that if the safety factors from uniaxial tests, shown by anchor points for tension (OHT) and compression (OHC), are used for other loading conditions, the approach might be either over-conservative or underconservative. In fact, the data generated using notched specimens are specific to the particular loading condition, laminate definition and geometry of the specimen and thus are not exchangeable. Therefore, the best practice for characterizing composites would be to generate a data set from smooth specimens and then use numerical simulation coupled with the data set to obtain results for the notched specimens. This not only reduces the cost of material characterization but also eliminates the above-mentioned erroneous practices. This underlines the significance of simulation, which also provides a cost-effective and general approach. For example, if a panel with an oblong opening needs to be optimized for biaxial loading, the efficient route would be first to generate data from the smooth specimen of the material to be used. This would be followed by numerical simulations to obtain the optimal layup. The optimal design can then be validated by experiments. This approach circumvents the need for sacrificing numerous test specimens and test hours during the optimization phase, thus significantly reducing the expenditure. Hence, to maximize the benefit from the use of composites, the envelope has to be pushed beyond the paradigm of black aluminium, and the conservative design rules should be challenged. Some of the current design and data generation practices existing in the industry are expensive and indeterminate. This can be avoided by utilizing the numerical Fig. 2: 10% rule (for [π/3] laminates) reduces the number of layup combinations [1]. Fig. 3: The moduli of three-angle-ply laminates compared to that of aluminium (or black-aluminium, in the case of general CFRPs) [1]. Fig. 4: A three-angle-ply laminate can compete with a four-angle-ply laminate in terms of strength (shown here are the failure envelopes for FPF-first ply failure and LPF-last ply failure) [1]. No48 May 2009 / jec composites magazine 33

JEC COMPOSITES MAGAZINE - Issue #48 - May 2009

Table of Contents for the Digital Edition of JEC COMPOSITES MAGAZINE - Issue #48 - May 2009