Pile driving in glacial soil
Stochastic modelling of boulder-induced premature refusal
Time: Fri 2025-06-13 13.00
Location: B3, Brinellvägen 23, Stockholm
Video link: https://kth-se.zoom.us/j/63212508471
Language: English
Subject area: Civil and Architectural Engineering, Soil and Rock Mechanics
Doctoral student: Chiara Cannizzaro , Jord- och bergmekanik
Opponent: Dr Leena Korkiala-Tanttu, Aalto University, Aalto, Finland
Supervisor: Docent Johan Spross, Jord- och bergmekanik
QC 20250521
Abstract
Pile foundations provide support to structure by distributing external loads to deeper and firmer soil layers. The most efficient installation method consists in driving piles into the soil using impact hammers. However, the presence of obstructions inside the soil can compromise an efficient pile driving operation. A natural type of obstructions in glacial soils are boulders, rock fragments resulting from glacial erosion. The random presence of boulders within the soil can severely damage the piles, compromise their structural integrity and leading to structural failure, referred to as premature pile refusal.
Large uncertainties related to the spatial distribution of boulders, their geometric properties and mechanical resistance, as well as the variability of the soil layers and the dynamic interaction between pile and boulders in firm soil, aggravate the complexity of the problem. Failing to properly quantify the risk of premature refusal can cause large additional costs and delays, as well as inefficient use of environmental resources, for a given construction project.
The current approaches available to practicing engineers are ambiguous and mostly deterministic, relying mainly on practical experience. The construction industry especially in Sweden, where a large part of the landscape is covered in a glacial soil known as till, could greatly benefit from a risk-based approach that accurately and systematically identify and model the significant sources of uncertainties of the problem, while remaining intuitive, practical and simple to use. Such an approach is especially crucial in the design stage, to select the most suitable pile material and installation method for specific soil conditions, towards more economical and sustainable pile design solutions.
This licentiate thesis presents a systematic method to analyse the risk of premature refusal during pile driving in glacial till. This includes the development of a one-dimensional homogeneous Poisson process to probabilistically model the presence of boulders and the variability of the soil layers thickness along the pile driving direction. This thesis also presents a cost model to identify and quantify the key economic implications of premature refusal, to achieve a better balance in pile design between cost, construction time and environmental efficiency.