Keywords: ground anchorage, vibration response, dynamics, impulse, modelling, simulation.

In many civil engineering structures, anchorage systems are utilised for the stabilisation of the surrounding ground. This is particularly true in mining, tunnelling and dams where such anchorages are safety critical components designed to ensure safe function of the structure over its lifetime. A problem associated with such ground anchorage systems however is the assessment of the condition of the anchorage over its lifetime and possibly most importantly the level of load within the anchorage. Recently a new system has been devised which allows non- destructive testing of such anchorages. This new system, GRANIT® which utilises a small impulse into the system, can provide a reliable estimate of the load in an anchorage system. This paper discusses some aspects of the excitation applied to the system and shows, using results from a numerical model, how careful tuning of the impulse can ensure good estimates of the stress in the anchorage.

A typical anchorage comprises a fixed length of tendon bonded to the surrounding rock/ground, a free length of tendon, which may have some secondary protection, the anchorage head and a short protruding length of tendon. The impact device of the GRANIT® system is attached to the protruding length and applies a small impulse. This is transmitted through the anchorage head and into the body of the anchorage. The resulting vibration response is recorded, processed and used to diagnose the load in the anchorage. The main features, which affect the response, are the geometry of the anchorage, the stiffness characteristic of the anchorage head and the load in the anchorage. The stiffness function of the anchorage head has been shown to be critical in the response of the anchorage to an applied axial tension impulse load[1]. The impact device of the GRANIT® system is essentially a simple air hammer, which comprises three main elements, the main body, a piston and the clamp assembly. The main element of the clamping assembly is a split tapered collet component inside the central tube of the device. The collet is slid over the tendon or in the case of bolt screwed onto the tendon. A screw adjuster within the device is then tightened, forcing the collet into a taper and so tightening it onto the tendon, ensuring a good grip. Collets can be manufactured for any type of tendon up to 30mm in diameter. The air used to operate the device is controlled by a ruggedised computer, which also samples the response waveform.

An important aspect of the function of the device is to ensure that energy is input to those features of the response, which provide most information about the load level in the anchorage. During testing it has been found that to do this the shape/duration of the impulse must be shaped to provide energy in the correct frequency band. Currently this is undertaken by utilising a series of mechanical filters (inserts) which modify the properties of the impact surfaces. As an aid to understanding this procedure on a more fundamental level, a series of simulations has been undertaken using a recently developed lumped mass model of an anchorage system[1]. This model, which discretises the tendon, grout and surrounding rock mass into an assembly of spring/mass/damper units and also includes the mass and stiffness of the anchorage head and the impact unit, allows the effect of different impulses to be assessed for a particular anchorage. In addition the effect of position of the impact device can also be assessed.

Initially an eigenvalue analysis of the anchorage model can be used to determine the frequencies, which are potentially of interest for diagnosing the load of the anchorage. The mode shapes (eigenvectors) can also be determined at this stage to ascertain whether the desired modes can be readily excited by application of an impulse at the protruding length. A transient analysis can then be undertaken to show the effect of the impulse duration and shape. The transient vibration analysis of the model is solved numerically using MATLAB. Some work has also been undertaken on attempting to find cavities in the proximal end of the grouting of anchorage systems. To provide enough information to assess the condition of the grout, an impulse must be applied which can penetrate the grout and be reflected back with sufficient amplitude from any interfaces to allow detection. The model is also being used to attack such problems.

The paper describes the GRANIT system and the numerical model in some detail and presents results from a number of simulations showing the effect of impact duration and shape on the response of an anchorage. This is related to test results obtained from real anchorage systems. Good correlation is shown. The final aim of the work will be to indicate in advance the most suitable form of input to a particular anchorage system and some pointers are given to development of the system.

1

A. Ivanovic, R.D. Neilson, and A.A. Rodger "Numerical Modelling of Single Tendon Ground Anchorage Systems", Geotechnical Engineering 149 (2) pp. 103-113, 2001. doi:10.1680/geng.149.2.103.39325

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description
purchase this book (price £122 +P&P)