Methods for the evaluation and interpretation of d

The introduction of absolute displacement monitor-ing using geodetic methods has significantly im-proved the basic concept of the observational ap-proach for tunnel construction. Rabensteiner (1996) already described the scope of optical displacement monitoring and the procedure to determine reliable 3D coordinates of monitoring points installed at the circumference of the tunnel. The huge amount of displacement data which are commonly produced on a daily basis require appropriate evaluation methods in order not to get lost in data tables, as well as to extract as much information as possible. In the mid of the last decade, several authors have given an overview of useful evaluation methods to improve the understanding of geomechanical processes dur-ing the tunnel excavation (Schubert & Steindorfer 1996, Vavrovsky & Schubert 1994, 1995). They also have introduced new techniques for the prediction of geological conditions ahead of the tunnel face by in-terpreting the trend of the displacement vector orien-tation (Schubert & Budil 1995). During the last dec-ade, these evaluation methods have been applied on many sites primarily during tunnel projects in the Alps. The experience continuously gained led to fur-ther improvement in the interpretation reliability.

The following chapters will give a rough over-view on the basic evaluation procedure as well as on the different main evaluation methods and their ap-plicability with regard to specific questions. Finally, with the help of a case history from Austria the in-terpretation process will be shown using a combina-tion of different methods.

2EVALUATION PROCEDURE

Depending on the expected ground characteristics and boundary conditions different levels of sophisti-cation of measurement data evaluation will be ap-plied. In relatively good and homogeneous rock mass, where small displacements are expected, the routine evaluation will be in the form of time dis-placement plots. Those are used to check if the ac-tual behaviour is within the expected range.

In cases of complex or weak rock mass conditions, where higher displacements are expected, and/or re-strictions in displacements due to the presence of buildings or other infrastructure have to be ob-served, more advanced evaluation procedures will be required.

A typical daily evaluation procedure during the tunnel excavation consists of two main categories:

- The evaluation and assessment of the tunnel sta-bilization process, and

- the interpretation of the displacement results with regard to the prediction of the geological-geotechnical conditions ahead and outside the tunnel profile.

Methods for the evaluation and interpretation of displacement monitoring data in tunnelling

K. Grossauer & W. Schubert

Institute for Rock Mechanics and Tunnelling, Graz University of Technology, Austria

ABSTRACT: Absolute displacement monitoring has become common practice on many tunnel sites. Those data in combination with the information on geological and boundary conditions have a tremendous potential to enhance the insight into mechanical processes around the tunnel and the ground-support interaction. Dif-ferent methods of data evaluation are used for specific purposes. Displacement history plots can be used to evaluate the stabilization process or to predict the final displacements rather than to predict geotechnical con-ditions ahead of the face. For such a request the use of more advanced evaluation methods are required.

To describe the spatial and transient development of displacements during tunnel excavation, analytical func-tions are used. The daily evaluation process will be enormously improved by comparing the currently moni-tored displacements to the expected displacement characteristic, determined using the analytical function. Any deviation from the normal behaviour can be identified in time. Critical displacement trends can be auto-matically detected as well as typical trends used to predict the ground conditions ahead of the face.2.1Assessment of the stabilization process

The interpretation of the time-displacement curve for one displacement component, mainly the settle-ments for the crown and sidewall points as well as the transversal component of the sidewall points, is the most significant evaluation method for this pur-pose. The condition for a satisfying stabilization, re-spectively the stress redistribution is a steadily de-creasing displacement rate, in case of homogeneous rock mass conditions and continuous advance rate. When the rock mass is heterogeneous or the advance rate not constant, the time displacement plot only does not provide conclusive evidence of the stabili-zation process. Schubert et al. (2002) exemplarily il-lustrated the influence of varying excavation pro-gress rates on the development of the settlements versus time. For the examples the formula for the time dependent closure of tunnels, developed by Guenot et al. (1985) and Sulem et al. (1987) was used.

This method is based on analytical functions that describe displacements in a plane perpendicular to the tunnel axis as a function of advancing face and time. Barlow (1986) and Sellner (2000) modified this approach. The displacement behaviour of the rock mass and support basically is represented by four function parameters. Two parameters (T and m) are used to simulate the time dependency and an-other two parameters (X and C) the face advance ef-fect. These parameters can be back-calculated from case histories using curve fitting techniques.

The system of these analytical functions was im-plemented in the program package called GeoFit®(Sellner 2000, GeoFit 2001). It provides easy-to-use tools for back-calculating displacement monitoring data by curve fitting and for the prediction of dis-placements. Both monitored and predicted results are shown and can be compared at any time. This procedure allows predicting displacements for any time and point of the tunnel wall as well as the ground surface considering different construction stages and supports.

With some experience appropriate function pa-rameters (X, T, C and m) can be reliably determined. Grossauer et al. (2003 & 2005) described a method using the development of typical displacement trends and the function parameters along the tunnel axis to extrapolate a parameter set for newly in-stalled monitoring sections. Thus the expected “normal” displacement development in relation to time, progress, construction sequence and support can be predicted after only a few readings. The con-tinuous comparison of the actually measured and the predicted data allows an easy identification of any deviation of the actual from the predicted behaviour. With this information additional investigations can be performed to clarify the reasons for the deviation and/or appropriate countermeasures taken in time.

The following case histories from two tunnel pro-jects in Austria should exemplify the assessment of the stabilization process by means of time displace-ment plots. The geological settings as well as sup-plementary information for the case histories can be found in Sellner & Grossauer (2002).

The upper part of Figure 1 shows the develop-ment of the settlements versus time for the upper left sidewall monitoring point at a certain cross section in the top heading. Both, monitored displacements (zero reading and two subsequent readings - dis-played as black circles) and predicted displacements (grey lines) are displayed. The dark-grey dashed line represents the expected displacement development over time for the top heading excavation, while the light-grey dashed line corresponds to the predicted displacements considering an additional support in the form of a temporary top heading invert. The pre-dicted displacements have been established using the analytical functions described above (Sellner 2000). The lower part of this figure shows the asso-ciated construction phase, e.g. the progress of the top heading (dark-grey line) and the subsequent temporary top heading invert installation (light-grey line). The solid lines represent the already accom-plished part of the respective construction phases while the dashed lines the scheduled ones.

top heading excavation

.

temp. invert installation

1000

900

800

700

Time in [days]

D

i

s

p

l

a

c

e

m

e

n

t

i

n

[

m

m

]

C

h

a

i

n

a

g

e

i

n

[

m

]

Time in [days]

-40-200+20+40+60+80+100

Figure 1. Displacement history plot for settlements of the up-per left sidewall point; comparison of the actually monitored displacements (circles) and predicted displacements (dark-grey and light-grey dashed lines) – associated construction phase diagram.

The excavation has been stopped approximately two days after the installation of the measuring sec-tion for operational reasons. The settlement reached about 30mm one day after installation of the target, and about 50mm after two days, when the excava-tion was stopped. These two readings were used to predict the displacement development (dark-grey

dashed line). As the prediction showed rather high displacements after the planned restart of the top heading excavation 50 days after the stop (approxi-mately 140 mm total), a temporary shotcrete invert was installed in the top heading. The predicted dis-placement with the temporary invert is represented as light-grey dashed line in Figure 1. During the pe-riod of the bench excavation far behind the monitor-ing section no construction activities were planned in the vicinity of the top heading face. Therefore only a slight increase in the settlements for the ob-served point due to creep of rock mass and support was predicted. The additional displacements after restart of the top heading were predicted with ap-proximately 20mm, reducing the total displacements from 140mm to about 80 mm.

Figure 2 shows the comparison of both, the actually observed settlements and the predicted ones. The predicted displacement history coincides quite well with the finally observed curve even with the un-steady excavation rate including a construction stop over a period of approx. 50 days and the excavation restart and further progress.

top heading excavation

.

temp. invert installation 1000900800700Time in [days]

D i s p l a c e m e n t i n [m m ]

C h a i n a g e i n [m ]

Time in [days]

-40

-20

+20+40+60+80+100

Figure 2. Comparison of actually observed settlements (circles) and the predicted settlements (grey lines).

Figure 3 shows a similar situation as described above on a different tunnel site. Displayed is the de-velopment of both, the measured and predicted set-tlements versus time for the crown point. The pre-dicted displacement curves are shown for the influence of top heading excavation (dark-grey dashed line) as well as the effect of the temporary top heading invert (light-grey solid line). Analogue to the prior example, the top heading excavation was stopped 2 days after the monitoring section has been installed and restarted after a period of about 20 days. The installation of the temporary top heading invert was done close to the face and alternately to

the top heading excavation. The development of the settlements shows a similar characteristic as the prior example but lower displacement values. In the first two days the settlements reached about 20mm, followed by slight increase during the stop. The re-start of the top heading excavation caused a pro-nounced increase in the crown settlements up to approx. 40mm in total. With increasingly distance between the monitoring section and the face, a de-creasing displacement rate indicated a satisfying sta-bilization. Up to this point, no significant deviation of the monitored settlements from the predicted normal displacement development can be identified. A sudden increase of about 10mm (marked with the dashed ellipse) occurred approximately 32 days after the zero reading, clearly indicating an abnormal sys-tem behaviour calling for an investigation of this situation and, if necessary the implementation of countermeasures to guarantee the tunnel stability. The shotcrete had settled during the stop and lost its ductility, thus behaving relatively brittle also close to the face. As no visual damage of the lining of the crown could be identified, the reason for the increase of displacements could either be a failure of the temporary invert or a failure in the surrounding rock mass. To be able to judge the stability of the system, two scenarios were considered. In the first case - failure of temporary invert - it could be ex-pected, that the displacement development would follow the predicted one for the case with no tempo-rary invert installed (dark-grey dashed line in Figure 3). In the case of a failure in the rock mass it could be expected, that the displacements would exceed those of the system without temporary invert. This situation might be classified as critical and addi-tional remedial measures would have to be taken.

top heading excavation temp. invert installation

950900850800

Time in [days]

D i s p l a c e m e n t i n [m m ]

C h a i n a g e i n [m ]

Time in [days]

-20

-10

+10

+20

+30

+40

+50

+60

Figure 3. Displacement history plot for settlements of the crown point; comparison of the actually monitored displace-ments (circles) and predicted displacements (dark-grey dashed line and light-grey solid line) – associated construction phase diagram.

As shown in Figure 3, the measured displace-ments followed the predicted path for the case with-out temporary invert, indicating that the temporary invert had failed. In this case, the prediction of the displacement development with and without the consideration of the temporary top heading invert provided a valuable assistance in the judgment of the unexpected situation.

Both case histories have shown that the definition of the expected normal behaviour is one of the cru-cial issues assessing the tunnel stabilization process. 2.2 Interpretation of displacement monitoring data

in combination with the geological-geotechnical conditions A tunnel excavation in homogeneous and isotopic rock mass with uniform primary stress conditions will cause a more or less symmetric displacement pattern in the profile, as well as a certain and regular characteristic development in the longitudinal direc-tion. Any changes in the rock mass structure or in the primary stress state influence the stress distribu-tion around the tunnel and ahead of the face, which again result in a change in the displacement pattern. Figure 4 shows the monitored displacement char-acteristic for the top heading excavation as vector plot in a plane perpendicular and parallel to the tun-nel axis of a tunnel in Austria, which has been exca-vated in Flysch, consisting mainly of a series of clay- and siltstones. The relevant geological struc-ture in terms of displacements is the bedding, dip-ping steeply in excavation direction and striking al-most perpendicular to the tunnel axis. The displacement pattern can be described as symmetri-cal in the cross section with slightly increased dis-placements on the left side. The plot in longitudinal direction shows displacement vectors tending against the excavation direction with a curved path after the first few readings.

both perpendicular (left) and parallel to the tunnel axis; situa-tion characterized as Flysch with a main structure nearly dip-ping vertical and striking nearly perpendicular to the tunnel axis from right to the left in an angle of approx. 80°.

Numerical simulations using Flac 3D with a ubiquitous joint model have shown that the influ-ence of a vertically dipping and perpendicularly

striking foliation results in a displacement character-istic more or less identical to the situation described above. Figure 5 shows the related displacement vec-tors in a cross- and longitudinal section. Additional cases with various anisotropy orientation settings have been investigated and the results compared to observed displacement data on site. Considering the site specific and displacement relevant geological structures, the actually observed and the simulated displacement characteristics showed a clear correla-tion. Goricki et al. (2005) and Button et al. (2006) show a couple of displacement plots for sites con-structed in Austria with regard to the anisotropy ori-entation and compare the plots with different syn-thetic examples from numerical simulations.

investigations using a ubiquitous joint model; joint orientation dipping vertical and striking perpendicular to the tunnel axis.

Using the knowledge about the influence of the anisotropy orientation on the displacement devel-opment, the geological structure encountered on site can be used to predict the expected spatial displace-ment vector orientation. Any significant deviation from this normal behaviour will thus reflect an ab-normal behaviour. In case the reason for the abnor-mality does not originate from a failed support sys-tem, the type of deviation can be used to predict changes in the rock mass condition ahead of the tun-nel. The evaluation of data gained from the excava-tion of tunnels constructed in Austria showed, that the ratio between radial and longitudinal displace-ment varied in a wide range. Matching the observed phenomena with the geological documentation, it was found that deviations of the ratio appeared when zones of different deformability were approached with the excavation (Schubert 1993). To verify the hypothesis, numerical 3D simulations have been performed. The results showed that changing rock mass conditions ahead of the tunnel face clearly in-fluence the displacement vector orientation (Stein-dorfer 1998). To quantify the influence of weak zones on stresses and displacements, further re-search has been conducted by Grossauer (2001) and later on by Jeon (2005). The amount of the deviation depends on the stiffness contrast between the rock masses and also on the width of the fault zone. The deviation increases with increasing fault zone length

up to a certain critical length, above which no fur-ther increase of the vector orientation can be ob-served. This critical zone length is in between 2.5 and 4 tunnel diameters, as shown in Figure 6 for 3 different stiffness contrasts.

d e v i a t i o n f r o m “n o r m a l ” [°]

Fault zone length [D]

02468

Figure 6. Deviation of the displacement vector orientation from ‘normal’ at the transition from intact rock mass to the fault zone, depending on the stiffness contrast and the width of the fault zone.

Research activities performed during the past decade with regard to the determination of the influ-ence of specific geological features on the displace-ment behaviour of tunnels have significantly im-proved the displacement monitoring data evaluation capabilities on site.

3 EVALUATION METHODS

The following chapter shows a set of different evaluation methods chosen for the geotechnical in-terpretation with regard to the geological conditions.

3.1 Value of evaluation methods Before interpreting the particular displacement

curves, the applicability and information capability

of the single evaluation methods have to be consid-ered. Vavrovsky & Schubert (1994) have given an

overview of the geomechanical relevance of some

monitored parameters. Schubert et al. (2002) sum-marized the different displacement monitoring data

evaluation methods and rated their value with regard to specific questions. Table 1 shows a brief and pre-liminary listing of the applicability of the single evaluation and display methods and the appropriate information values. It has to be pointed out, that usually a combination of evaluation methods is re-quired to obtain a clear understanding of the geo-technical situation and the rock mass and tunnel be-haviour. The achievable accuracy using trigonometric measurements for the determination of the displacements of targets mounted on the tunnel wall under the consideration of an acceptable effort is in a range of about +/- 1mm. This takes into ac-count numerous possible sources of error and the ex-tremely dynamic environment during tunnelling. Several evaluation methods are based on the ratios of displacement components, e.g. the ratio between longitudinal displacements and settlements. For situations with monitored displacements of less than 1cm, an error of +/- 1mm will lead to a considerable variation in vector orientations and thus might lead to misinterpretations.

3.2 Combination of different evaluation methods To obtain a complete picture of the mechanisms oc-curring around a tunnel, in general different methods of monitoring data evaluation have to be used.

Figure 7 shows a set of evaluation methods in combination with the simplified geological situation for a 200m long section of the Inntal tunnel, Austria. The overall rock mass in this area can be described as a fault zone consisting of Quartz Phyllite, in-tensely sheared in different degrees. The hatching indicates the intensity of faulting. Narrow hatching signifies intense faulting. Polygons are used to indi-cate zones with block-in matrix-rocks.

The interpretation of the displacement character-istic of the 200m long area has been done using the following evaluation methods:

- Deflection curve diagram for the settlements of the crown point,

- trend line of the displacement vector orientation (H/S for the crown point, L/S for both, the crown and left sidewall point), and

- trend line of the displacement ratios calculated for the horizontal, as well as the vertical dis-placements.

H/S … ratio of horizontal displacements and settle-ments, expressed as an angle in degrees - re-flects the deviation of the displacement vector

from the vertical direction in a vertical plane perpendicular to the tunnel axis. L/S … ratio of longitudinal displacements and set-tlements, expressed as an angle in degrees - re-flects the deviation of the displacement vector from the vertical direction in a vertical plane parallel to the tunnel axis. All trend lines have been generated in a distance of

12m behind the top heading face. The displacement ratios have been evaluated for the horizontal dis-placements of the left and right sidewall point. The

last diagram shows the displacement ratio evaluated

for the settlements of the crown and left sidewall point. In the case of the horizontal displacements (diagram E), a ratio of 1 denotes equal horizontal displace-ment magnitudes on both measuring points. Positive

values higher than 1 signify that the horizontal dis-placements of point 2 (left sidewall point) are larger

tive values stand for the case of lager horizontal dis-placements of point 3 than point 2. In the case of the vertical displacements ratios (diagram F in Figure 7), positive values indicate larger crown settlements than sidewall settlements. The plotted values show the ratio of displacements between the two points.

The settlements evaluated for the crown point (deflection curves, diagram A) show magnitudes ranging from 20cm up to maximum 50cm. On the average the vertical displacement in the section shown are around 35cm. Starting the displacement interpretation from the left hand side, a pronounced deviation of the trend lines from a normal range can be observed in the displacement ratio diagrams E and F. The normal range is indicated in both dia-grams by the hatched area.

In diagram E, the horizontal displacement ratio of the left and right sidewall point shows relatively high displacements at the right sidewall at tunnel chainage 4040. A quasi identical trend development can be observed for the settlement ratio of the crown and left sidewall point (diagram F). Both displace-ments ratio trends reflect an asymmetric displace-ment pattern, with larger displacements at the right sidewall. This characteristic can also be observed in the trend development of the crown vector orienta-tion H/S (diagram B). From a more or less vertical orientation, the vector increasingly points to the left, reaching a value of about 5°. The vector orientation L/S of the left sidewall point (diagram C) also shows a clear deviation from the normal range. The vector orientation increasingly tends against the excavation direction, indicating weaker rock mass ahead (Schu-bert & Steindorfer 1996, Steindorfer 1998). The moderately dipping fault above the crown, which crosses the tunnel axis in an acute angle influences the stress distribution due to the tunnel excavation. wall and the crown of the tunnel, leading to in-creased displacements. The “abnormal” displace-ment characteristic in the section between 4020m to 4040m clearly indicates a weak zone outside the visible tunnel profile; the fault is visible in the tun-nel profile only at around station 4060m. The value of observing displacement ratio trends is obvious, as it shows also the effect of non-visible geological features, allowing a timely adjustment of excavation and support. It is interesting to note that although weaker ground is ahead of the face, the vector orien-tation trend L/S for the crown (diagram C) remains within the normal orientation of about 5° against the excavation direction, while usually a larger devia-tion would be expected (Golser & Steindorfer 2000, Grossauer 2001, Steindorfer 1998). In contrast to steeply dipping faults, smoothly inclined faults dip-ping in the direction of excavation do not result in a significant change in the displacement vector orien-tation in the crown. A slight trend deviation in the vector L/S (diagram C) can be observed at approx. 4070m. The reasons for this vector deviation tending against the excavation direction, is probably caused by the relative weak rock mass at chainage 4095, embedded between the 2 intensely faulted zones. Crown settlements monitored at this section show magnitudes of 50cm.

The next relevant deviations of trends from the normal range can be observed in diagram E at chain-age approx. 4090 as well as 4120. Both trends indi-cate larger horizontal displacements of the left side-wall than the right and are caused by weaker material on the left side. At the end of the section displayed here, similar trends as discussed in the be-ginning can be observed in diagrams E and F, which indicate another weak zone in a similar orientation as that one crossing the tunnel at 4060m.

Chainage

d i s p l a

c e m e n t [m m ]

D i s p l a c e m e n t R a t i o [-]Ratio 12

Figure 7. Development of different trend lines evaluated for a 200m section of the Inntal-tunnel - associated geological situation.

The most common way to assess the tunnel sta-bility is by evaluating time-displacement plots. However, the value of information of this evaluation is limited, as the face advance plays a major role in the displacement development. Tools for the predic-tion of displacements with respect to tunnel advance and time have been developed. Once a “normal” be-haviour is predicted, the observed displacements can be compared to the predicted ones, and deviations identified, and appropriate measures taken.

In particular in complex rock mass conditions, the evaluation of displacement histories is not suffi-cient to allow a timely reaction to the changing ground conditions. To be able to observe the influ-ence of geological features outside of the tunnel pro-file, trends of displacements and trends of ratios of different displacement components can be used. In this way, the geotechnical conditions ahead of the face and outside the visible area in the tunnel can be predicted.

The daily interpretation of a huge number of dia-grams requires extensive experience and profound understanding of rock mechanics. A research pro-gram is currently undertaken on the Institute for Rock Mechanics and Tunnelling at the Graz Univer-sity of Technology with the aim to establish an ex-pert system, which evaluates the different trends and provides an interpretation automatically. This tool combined with the monitoring data evaluation soft-ware will ease the day to day work of geotechnical engineers on site, and make tunnelling in complex geological conditions less risky.

5REFERENCES

Barlow, J.P. 1986. Interpretation of Tunnel Convergence Measurements. MSc Thesis, Dep. of Civil Engineering, University of Alberta, Canada

Button, E.A., Leitner, R., Pötsch, M. & Schubert, W. 2006.

Spatial relationships between discontinuity orientation and system behaviour in underground excavations. In David P.

Yale et al. (eds.), ARMA’s Golden Rocks 2006. Proc. 41st U.S. R. Mech. Symp. 2006, Golden, Colorado, US: CD GeoFit, 2001. www.geofit.3-g.at

Gloser, H. & Steindorfer, A. 2000. Displacement Vector Ori-entations in Tunnelling. In Felsbau 18 (2000), No. 2: 16-

21. Essen: VGE

Goricki, A., Button, E.A., Schubert, W., Pötsch, M. & Leitner, R. 2005. The Influence of Discontinuity Orientation on the Behaviour of Tunnels. In Felsbau 23 (2005) No.5: 12-18.

Essen: VGE

Grossauer, K. 2001. Tunnelling in Heterogeneous Ground – Numerical Investigation of Stresses and Displacements.

Master thesis at the Institute for Rock Mechanics and Tun-nelling, Graz University of Technology, Austria Grossauer, K., Schubert, W. & Kim, C.Y. 2003. Tunnelling in Heterogeneous Ground – Stresses and Displacements. In Technology Roadmap for Rock Mechanics; Proc. of the 10th Congress of the ISRM, Johannesburg, South Africa, 8-

12 September 2003: 437-440. Johannesburg: The South Af-

rican Institute of Mining and Metallurgy

Grossauer, K., Schubert, W. & Sellner, P. 2005. The Impor-tance of Displacement Prediction. In Erdem, Y., Solak, T.(eds.), Underground Space Use – Analysis of the Past and Lessons for the Future; Proc. of the 31st ITA-AITES World Tunnel Congress, Istanbul, Turkey, 9-12 May 2005: 1239-1244. Great Britain: A.A. Balkema Publishers Guenot, A., Panet, M. & Sulem, J. 1985. A New Aspect in Tunnel Closure Interpretation. In Ashwoth, E. (ed). Re-search and Engineering Applications in Rock Masses. Proc.

26th US Symposium on Rock Mechanics, Rapid City, South Dakota, 26-28 June, 1985: 445-460. Rotterdam: Balkema Jeon, J.S., Martin, C.D., Chan, D.H. & Kim, J.S. 2005. Predic-tion ground conditions ahead of the tunnel face by vector orientation analysis. Tunnelling and Underground Space Technology 20 (4): 344-355. Elsevier

Rabensteiner, K. 1996. Advanced Tunnel Surveying and Monitoring. In Felsbau 14 (1996), No. 2: 98-102. Essen: VGE

Schubert, W. 1993: Erfahrungen bei der Durchörterung einer Großstörung beim Inntaltunnel. In Felsbau 11 (1993), No.

6: 443-447. Essen: VGE

Schubert, W. & Budil, A. 1995. The importance of longitudi-nal deformations in tunnel excavation. In Fujii, T. (ed.), Proc. 8th int. Congress on Rock Mechanics, Vol. 3, Tokyo, Japan, 1995: 1411-1414. Rotterdam: A.A. Balkema Schubert, W. & Steindorfer, A. 1996. Selective Displacement Monitoring during Tunnel Excavation. In Felsbau 14 (1996), No. 2: 93-97. Essen: VGE

Schubert, W., Steindorfer, A. & Button, E. 2002. Displacement Monitoring in Tunnels. In Felsbau 20 (2002), No. 2: 7-15.

Essen: VGE

Sellner, P.J. 2000. Prediction of Displacements in Tunnelling.

Ph.D. Thesis, Graz University of Technology, Austria. In Schubert, Riedmüller & Semprich (eds.). Schriftenreihe der Gruppe Geotechnik Graz, Heft 9

Sellner, P.J. & Grossauer, K. 2002. Prediction of Displace-ments for Tunnels. In Felsbau 20 (2002), No. 2: 24-30. Es-sen: VGE

Steindorfer, A.F. 1998. Short Term Prediction of Rock Mass Behaviour in Tunnelling by advanced Analysis of Dis-placement Monitoring Data. PhD thesis, Graz University of Technology, Austria. In G. Riedmüller, W. Schubert, S.

Semprich (eds.). Schriftenreihe der Gruppe Geotechnik Graz, Heft 1. Graz

Sulem, J. Panet, M. & Guenot, A. 1987. Closure analysis in deep tunnels. In Int. Journal of Rock Mechanics and Min-ing Science 24: 145-154. Amsterdam: Elsevier Vavrovsky, G.M. & Schubert, P. 1994. Interpretation of Moni-toring Results. World Tunnelling Vol. 7 (1994), No. 8: 351-356. London: The Mining Journal Ltd.

Vavrovsky, G.M. & Schubert, P. 1995. Advanced analysis of monitored displacements opens a new field to continuously understand and control the geotechnical behaviour of tun-nels. In. Fujii, T. (ed.), Proc. 8th int. Congress on Rock Me-chanics, Vol. 3, Tokyo, Japan, 1995: 1415-1419. Rotter-dam: A.A. Balkema