The first part of this article concluded that the common interpretation of convergence and geodetic deformation measurements remains at an empirical level, yielding no in-depth insight into the mechanics of a tunnel system. Nor does it provide any clear evidence on the load-bearing capacity of a shotcrete tunnel.
On the basis of such measurements alone, it would not be possible to specify any factor of safety against failure of the tunnel construction and its lining.
Efforts to correct this are being made by Rokahr and co-workers of the University of Hanover. These include extensive on-site testing at a number of Austrian and German tunnelling projects.
In realising that the stress-strain relationship of shotcrete is highly time-dependent, Rokahr claims that science has advanced to the stage where this relationship can be specified with a sufficient degree of accuracy and confidence. Employing numerical modelling procedures, he converts the measured strain of the shotcrete lining (as deduced from the displacement measurements) into stresses. On the basis of this information, it is then possible to specify the actual degree of loading, the load-bearing capacity and the actual factor of safety of the shotcrete lining.
Fig. 5 shows an example for a 925m-long shotcrete tunnel section. It indicates that, while subject to significant local fluctuations, the capacity of the shotcrete lining (i.e. the actual load divided by the load-bearing capacity) was maximally occupied to about a level of 50-80%. This is equivalent to a factor of safety against failure of the shotcrete lining of between 1.25 and 2.0.
The derivation of the load bearing capacity of the shotcrete lining is based on a number of steps as indicated in fig 6. Each step is intrinsically associated with assumptions and errors. It is therefore highly desirable to check the end result (the computed stresses acting in the lining) by direct measurement.
Recently this has been done by employing the slot relief and flat jack compensation method. This has a proven record of reliability in stress determination of concrete and similar materials at accessible surfaces.
The slot relief and flat jack compensation method is carried out in a sequence of three steps (fig 7). Measuring marks are set next to the intended cut (fig 7a). Then the slot is cut by a diamond saw, typically 450 mm, and measurements taken on the convergence of the measuring marks caused by slotting (fig 7b). Finally, a flat jack is inserted into the slot and hydraulically inflated until the point of full reversal of the measured convergence (fig 7c). This is termed the compensation point. Accordingly, the pressure acting in the jack at the compensation point is termed compensation pressure. This pressure is (nearly) equivalent to the stress in the shotcrete to be determined. Note that no knowledge of material parameters (such as the Young’s modulus E) is required for this test.
In the example of the tunnel project shown in fig 5, the result of the comparison between computed stresses (as deduced from geodetic displacement measurements) and measured stresses (employing the flat jack compensation method) is shown in fig 8.
The evaluation of the capacity of shotcrete linings is expected to become a widely applied standard in tunnel construction in the near future. The base requirement for this will be extensive geodetic deformation monitoring, accompanied with regular checking of the computed stresses by the slot cutting and flat jack compensation method.
Monitoring displacements and stresses
It has already been indicated that convergence and/or geodetic deformation measurements alone are insufficient for a full judgement of the mechanical behaviour of the tunnel system. Two reasons can be identified.
1) The ground surrounding the tunnel is not directly monitored. This, however, is necessary as the ground has a definite load-bearing capability and is one of the contributing factors in the overall stability of the tunnel system.
2) A mechanical description of a tunnel system remains incomplete if it is solely based on displacements and its derivatives. Knowledge of the forces (and stresses) are definitely required for completeness.
In a tunnel monitoring programme which is specifically set up for control of the tunnel design, at least all of the standard instrumentation should be installed to provide a sufficiently broad data base for comparisons to be made between measurements and predictions. This monitoring programme consists of:
1) Deformation measurements of the excavated tunnel surface, as already described.
Instrumentation: total stations and reflector targets (in special cases: convergence tape).
2) Deformation measurements of the ground surrounding the tunnel.
Instrumentation: borehole extensometer (especially the three-point extensometer)
Comment: This is hardly used in Europe now as installation interferes with tunnelling operations. The situation is different in near-surface tunnelling (see below).
3) Control of the ground support element ‘anchor’ or ‘rock bolt’.
Instrumentation: anchor load cell – monitoring of the forces at the head of the anchor; measuring anchor – strain monitoring over the length of the anchor. This yields information on the required length of the anchor.
4) Control of the ground support element shotcrete.
Instrumentation: total pressure cells (TPC) – passive hydraulic flat jacks for monitoring of the radial and tangential stresses.
Comment: In Europe, many tunnel engineers are disenchanted with TPCs for shotcrete stress monitoring. Clearly, the performance of TPC is critically dependent on a number of factors, among them the TPC design, local conditions and, in particular, the quality of the installation. Some engineers prefer concrete embedment strain gauges instead of TPCs. This requires an in-depth knowledge of the time-dependent stress-strain relationship of the shotcrete (see also Rokahr’s approach as above).
5) Control of the ground support element ‘steel arches’.
Instrumentation: strain gauges and total pressure cells.
Comment: Not commonly used in Europe.
Measurement of ground displacements
In near-surface tunnelling, the standard instrumentation will be modified accordingly. In particular, this applies to the measurement of the ground movements by borehole extensometers which are now installed from the ground surface and not, as previously, from within the tunnel excavation.
Extensometers which are installed from the ground surface offer the following advantages:
– no interference with tunnel construction operations (this is seen as the major advantage);
– Installation and measurements are not restricted to the post-excavation phase. All ground deformations can be monitored including those ahead of tunnelling which are of particular concern in inner-city tunnelling;
– problem-free installation and convenient measuring operations with high-definition probe extensometers.
The measuring example (fig 9) shows these advantages.
Borehole extensometers measure the particular component of ground displacements which is directed along the axis of the borehole. For monitoring the complete deformation state of the ground, instruments must be employed, in addition to extensometers, which measure the displacement components acting across the borehole axis. This is achieved by standard inclinometers (in vertical boreholes) or by horizontal inclinometers or deflectometers (in inclined or horizontal boreholes), either stationary as fixed borehole chains or as mobile borehole probes.
Fig10 shows a measuring example of a deep-seated tunnel in which the ground displacements were measured by the conjunctive use of mobile extensometer and inclinometer.
The measuring example of fig10 also indicates the intrinsic purpose of this type of monitoring which is to provide the basis for a comparison between measurements and predictions. The displacements, as predicted from numerical modelling studies, are indicated in fig 10 by arrows. It is for the geotechnical engineer to judge the degree of agreement achieved and to decide whether the geotechnical model will be acceptable or has to be refined.
In Europe, mobile borehole probes for high-definition ground displacement measurements around underground openings are in common use. The most popular extensometer probe is the sliding micrometer, manufactured by SolExperts with a wide distribution in countries such as Switzerland, Germany and eastern European countries. The INCREX probe, manufactured by Interfels, is popular in Italy, Austria and parts of Germany. The market for mobile inclinometers is highly contested between Glötzl, SisGeoin Italy and Slope Indicator in Switzerland.
The least common instrument is the mobile deflectometer. However, in recent years Interfels, with its newly designed deflectometer probe, has won a sizeable market share. Combined probes are manufactured by SolExperts. They are the TRIVEC extensometer/ inclinometer probe and the extensometer/deflectometer probe LADEX.
Measuring ground stresses
While a complete, well-proven and widely used set of instrumentation exists for monitoring deformations in the ground, this is not necessarily the case for stress measurements and stress monitoring. From both a conceptual and technical point of view, the measurement of ground stresses (and the change of stresses) is generally much more difficult than that of the displacements.
Some engineers make a virtue of this situation when arguing that they can do without any stress measurement and stress monitoring. However, as mentioned before, stresses are an intrinsic part of any geomechanical system and cannot be ignored if our considerations are to be complete.
Continuing problems with measuring and monitoring of ground stresses have lead a number of instrumentation manufacturers towards the development of improved or innovative stress measuring methods. Table 3 gives an overview of current testing. The table also gives an indication of the author’s evaluation of the future development potential of the various methods.
Fig 11 shows a stress measurement example. It indicates the distribution of the circumferential stresses in the sidewall rock of a tunnel, determined by borehole slotting. The high-definition measurements clearly delineate a plastic zone about 2m deep. This information is important for the tunnel design, it gives, for example, indication on the post-failure characteristics of the ground, confirmation of the rock loads and selection of proper lengths of the rock bolts.
Detailed and systematic investigations with objective comparisons between the various stress measuring and monitoring methods as well as comparisons between measured and predicted values are yet to be carried out in tunnel construction. Such investigations, however, are absolutely essential for improving our knowledge of tunnel systems, thereby permitting better design, safer and more efficient construction procedures.
Conclusions
In continental Europe, the following trends can be identified in performance monitoring of tunnels and other underground structures for design purposes.
– In terms of volume of work and turnover figures, geodetic deformation monitoring represents the real backbone of today’s tunnel performance monitoring work.
– Much emphasis is given to the integration of geodetic tunnelling surveying with traditional geotechnical monitoring methods. The key for this approach is the availability of a suitable integrated acquisition and evaluation software.
– The interpretation of the common convergence and geodetic deformation measurements remains at a rather empirical level. This is seen as a major deficiency in current practice. It can be expected that in the near future a more rigorous evaluation of the load-bearing capacity of shotcrete linings will become a widespread standard. The base requirements for such evaluation will be extensive geodetic deformation monitoring, in-depth knowledge of the material law of shotcrete and regular checking of the computed stresses by the slot cutting and flat jack compensation method.
– Measurements and monitoring of the ground stresses, principally desirable from an engineering point of view, are still not in widespread use. New high-definition stressmeters can delineate primary and secondary stresses around a tunnel at reasonable costs.
– Overall, it appears that tunnel engineers no longer have the strong interest in detailed instrumentation programs for checking and improving tunnel design they had some decades ago. This is in marked contrast to booming instrumentation demands for control of the construction procedures.
Related Files
Figure 5: Capacity factor of a shotcrete lining
Figure 10: Ground displacements
Figure 7: Principle of stress
Figure 11: High definitionn stress measurement
Figure 8: Tangenital stresses
Figure 6: Evaluation
Figure 9: Monitoring of the loosening of roof strata
