In developing a neural network technique for a finite element model updating, researchers have been shown that the number of training samples and their quality, significantly affect the accuracy of the NN predication. In this study, based on the genetic algorithm (GA) method, we reduce the number of analyses required to develop the training pairs and reduce the amount of time for training the NN. In the other words, a uniform distribution of design points inside the design space will be obtained by means of this approach. To validate the efficiency of GA sample selection, random generation (RG) method is used for comparison. Two comparisons are made based on a numerical and experimental example. One is updated from 10 degrees of freedom lumped parameters system and the other is updated from a free-free beam using test data. The results indicate that the GA sample selection can reduce the number of training samples without affecting the accuracy of the NN predication. In our present study, also, the advantages of using frequency response function (FRF) data as input to the NN are kept and the drawback of having too many frequency points is overcome by the application of principal component analyses (PCA).

The maximizing of sound transmission loss (TL) across a functionally graded material (FGM) cylindrical shell has been conducted using a genetic algorithm (GA). To prevent the softening effect from occurring due to optimization, the objective function is modified based on the first resonant frequency. Optimization is performed over the frequency range 1000–4000 Hz, where the ear is the most sensitive. The weighting constants are chosen here to correspond to an A-weighting scale. Since the weight of the shell structure is an important concern in most applications, the weight of the optimized structure is constrained. Several traditional materials are used and the result shows that optimized shells with aluminum-nickel and aluminum-steel FGM are the most effective at maximizing TL at both stiffness and mass control region, while they have minimum weight.

Purpose – The purpose of this paper is to quantify uncertainty design parameters of long service usage material and including their effects explicitly in analysis of fracture toughness for evaluation of reliability indexes. Design/methodology/approach – In this paper structural reliability algorithm was incorporated into comprehensive finite element software. In this algorithm, the limit state function and relative gradient of uncertainty parameters has been determined using finite element model. Findings – Results from this paper show that the effect of material degradation due to long service usage on the Young’s modulus for corroded samples and non-corroded samples are approximately 5 and 2.5 percent below the handbook value, respectively. The uncertainty in Young’s modulus is small, the randomness is unsymmetrical, and the dispersion is more above the mean than below the mean. The failure probability increases with the increasing uncertainty of Young’s modulus, and can be much larger than the probabilities calculated for a deterministic elasticity. Originality/value – Reliable structural analysis of aging aircraft is crucial of safe operation.

The purpose of this work is to describe the effects of the countersink depth on the residual hoop stress in a flash riveted single lap joint. In this research instead of three dimensional finite elements, a force-controlled two-dimensional axisymmetric finite element analysis has been carried out to simulate the rivet installation. Results from this analysis show that with decrease in countersunk portion of the outer sheet, the rivet expansion is larger in the upper skin, leading to an increase in the compressive residual hoop stress near the hole edge. Furthermore the countersink depth must not exceed 60% of the skin thickness and anything beyond that will cause the skin to become knife edged. Using press countersinking instead of machine countersinking is highly recommended for sheet thickness less than 0.032 inch