Lesson learnt from static pulling tests on trees: an experimental study on toppling behaviour of complex foundations

Standard procedures for stability assessment of unstable trees are based, among other, on the interpretation of on-site, non-destructive static pulling tests. To this goal, a simple phenomenological equation is usually adopted in professional agronomic practice, and an estimation of the ultimate toppling resistance is extrapolated by fitting the test data, without taking root geometrical parameters and soil mechanical properties into account. From a geotechnical point of view, however, the root plate of a tree plays the role of a “living foundation”, and its behaviour under toppling actions (like those produced by intense wind gusts) conceptually corresponds to the mechanical response of shallow foundations under rocking loads. In the paper, several static pulling tests on real-scale trees (some of them have been run until the complete collapse, after some unloading–reloading cycles) and some tests taken from the literature are considered in order to investigate the toppling behaviour. A possible new interpretative equation is proposed and critically compared with the existing one against experimental results. The new equation allows for a mechanically meaningful description of the toppling curve of the tree and accounts for strength and deformability issues. It allows to introduce innovative “performance-based” approaches, which are commonly neglected by practitioners and professional agronomists in this field. Nevertheless, the experimental results show that tree toppling is a complex phenomenon, and capturing its failure condition requires more advanced multi-mechanism models and second-order effects to be accounted for. From a practical point of view, the proposed equation, employed within the same standard interpretative procedure currently adopted in practice for pulling tests, seems to provide conservative estimations of “operational” values of the ultimate toppling resistance, and in perspective, it could be used to significantly optimize—when needed—the design of structural stabilizing interventions on potentially unstable trees.