Uncertain analysis for the prediction of residual elastic modulus of damaged composite

To overcome the limitation of traditional probabilistic statistical method, two non-probabilistic methods to research the residual elastic modulus of the damaged composite with uncertainties were presented. In the non-probabilistic methods, uncertain variables were described as interval numbers or convex sets, and then the interval range of residual elastic modulus could be obtained by means of Taylor expansion and interval calculation. The main advantages of non-probabilistic methods are that the needed uncertain information is decreased; the methods are simple, easily applicable and the results are highly precise. Through two cases of a numerical example, the residual elastic modulus of damaged composite with uncertainties were calculated using the presented non-probabilistic methods. The results indicate that when the information of uncertain parameters is less, using the presented methods can obtain satisfying results for the uncertain problem.