Theoretical and experimental study of fully grouted rockbolts and cablebolts under axial loads

Towards a new constitutive law for the bolt-grout interface

Applications in coal mining in North america

Fully grouted rockbolts and cablebolts have been broadly used since the 1960s not only in mining but also in civil engineering applications. Indeed, rockbolting is the most effective and the most economical means of supporting excavations in rock. At present, hundreds of millions of bolts are installed each year worldwide: only roof bolting applications in coal mining in North America represent annually about 108 bolts. The main advantages of rock and cablebolting include an easy and fast installation, the capability to reinforce large underground openings, the minimization of the obstruction in the excavated cross-section and their interesting price as compared to other reinforcement technologies. 

In order to gain more insight into the load transfer mechanism between the reinforcement element and the surrounding ground, the behaviour of the bolt-grout and the grout-ground interfaces needs to be addressed. In this thesis, attention has been focused on the bolt-grout interface because experience proves that it is often weaker than the grout-ground interface.


View of the laboratory facility 

 

Pull-out tests and analytical tools

To study the bolt-grout interface, a relative slip between the bar and the grout must be created, so that the two materials are no longer coupled. This is normally achieved by means of pull-out tests, in which an axial tensile load is applied to one extremity of the bolt, while the other end is left free. Since the interface is not directly accessible to measurements, the proposed alternative in this thesis is to combine experimental data issued from pull-out tests with suitable analytical tools that let convert the measured variables (taken in the surrounding materials) into interface variables. Therefore, not only the tests should be carried out under accurate conditions, but also the surrounding materials should be well characterized.

 
Cutaway section of the pull-out test bench (scale 1:10)

 

 

A large experimental campaign was conducted during the Ph.D. Most of the tests were conducted at laboratory-scale, using a new dedicated experimental bench, first-time used and tuned during the Ph.D. This bench was calibrated and improved progressively to study the bolt-grout interface as accurately as possible. In addition, the tests carrying out and the analysis of the results have proved that the study of the interface is very demanding and that the test set-up and operating conditions should be addressed carefully. The research has been conducted in static conditions, and only monotonic loadings have been considered.

Furthermore, the laboratory-scale study has been completed with an in situ investigation at Andra’s underground research laboratory in North-Eastern France, where fully grouted rockbolts are systematically integrated in the reinforcement pattern.

 

 
From the available data, the interface variables were derived using classical and new analytical tools developed during the thesis. A semi-empirical model for the shear stress and the radial opening has been proposed. The new interface model is based on a joint formation at very low values of shear slip and points out that friction is not the only component to the shear stress, especially before the peak strength. Moreover, the process of joint decoupling needs to be considered. An important aspect of the model is that it accounts for the effect of the bolt profile on the radial opening of the interface.

                             
                   Model 1: comparison between experimental and modelled shear stresses.
                   Model 2: comparison between experimental and modelled radial openings.

To conclude, the novelty of this thesis lies in the method developed to reach the interface itself and to define an interface behaviour model. Eventually, one of the most practical uses of this method would be the implementation of the interface model into a numerical code (finite element, finite differences, etc.) to predict the response of a bolted structure more accurately and to help define the most suitable reinforcing pattern. In this sense, the use of pull-out tests (i.e., the use a tensile load on the bolt) to study the bolt-grout interface is justified: the numerical code will superimpose the contribution of each loading (axial, bending, etc.) on the reinforcement element. The critical aspect is then the accurate modelling of each component, including the interface.

From Oviedo School of Mines to Berkeley Lab

Laura Blanco Martin started her undergraduate studies in October 2003 at the Oviedo School of Mines, Spain. In September 2006, she joined the second year at the “cycle ingénieur civil” at MINES-ParisTech. She obtained her Spanish and French engineering degrees in 2008, then decided to start a Ph.D. at the Geosciences Department of MINES-ParisTech, where she had conducted her masters final internship. Her Ph.D. supervisors are Michel Tijani and Faouzi Hadj-Hassen.

Since August 2012, Laura is a postdoc fellow at the Earth Sciences Division of Lawrence Berkeley National Laboratory, California. She is working with Jonny Rutqvist’s team on the long-term evaluation of coupled thermal-hydraulic-mechanical processes in salt-based repositories for high-level nuclear waste. Laura Blanco Martin received the French Pierre Londe 2013's prize of thesis.