Are probabilistic methods a way to get rid of fudge factors?

There are specific reasons that make the dimensioning of adhesively bonded joint sparticularly arduous; the sharp stress peaks at the end of the overlaps are certainly the most critical one. Over the last decades, several techniques have been developed to overcome associated problems, which are particularly critical in conjunction with material brittleness. However, there is no denying that most of them rely in one way or another on ad hocquantities or fudge factors introduced into the calculation, formula or model in order to make their results fit experimental observations or analytical predictions. In contrast to these models, that at one level or another introduce the above mentioned parameters, the still generally discarded probabilistic approach allows overcoming a series of limitations regarding the dimensioning of adhesively bonded joints. This article aims at shedding some light on the concepts behind probabilistic methods, presents how they can be implemented in practical design procedures for adhesively bonded joints, and summarizes a series of examples of joint capacity predictions covering a wide range of relevant engineering materials [1-2].