INTERNATIONAL SOCIETY FOR SOIL MECHANICS AND GEOTECHNICAL ENGINEERING
This paper was downloaded from the Online Library of the International Society for Soil Mechanics and Geotechnical Engineering (ISSMGE). The library is available here: https://www.issmge.org/publications/online-library
This is an open-access database that archives thousands of papers published under the Auspices of the ISSMGE and maintained by the Innovation and Development Committee of ISSMGE. Geotechnical Aspects of Underground Construction in Soft Ground, Kusakabe, Fujita & Miyazaki (eds) © 2000 Ba/kema, Rotterdam, ISBN 90 5809 1 066
High-speed railway tunnelling in soft granular Frankfurt ground
R.Katzenbach, U.Arslan, G. Festag & A. Rtickert Institute of Geotechnics, Darmstadt University of Technology Germany
ABSTRACT: Two examples show the experience with highspeed railway tunnelling in F rankfuit soft ground. At the ”Frankfurter Kreuz” the railway line Cologne-Frankfurt/Main crosses the highways Fra.nkfLu"t-Basel and Colo g;ne Wiirzburg in a tunnel in near vicinity to the Frankfurt/Main airport. The construction of the tunnel was composed of a two step support. ln the first step the crown was excavated under a jet grouted screen. The working face was reinforced with a support core and eight additional horizontal jet grouted columns. ln a second step the lower part of the tunnel was excavated. To verify the soil parameters two series of in-situ large scale shear tests were performed. Within the project ”Frankfurt 21” a new railway tunnel undercrossing the city of Frankfurt/Main is planned. For that various underground tunnels as well as numerous buildings and skyscrapers with their fotmdations, the river Main and the East docks will be undercrossed. ` l INTRODUCTION high speed A ' to FrankfuV railway The new railway line Cologne-Frankfurt/Main is the N , .JA centrepiece ofthe Gennan and the future European high \on%_d\§\““C5 7? Gateway Gardens `={~ B42 speed traffic network and links important national and \(\?,\\5Wed \o%“e S\.ar\°“ f `Wg~g- ’ '~-'_' ef ‘ S_ international long-distance lines. The new line will act Yd\W,@1 \° Co PS, _ tunnel in V' as the central north-south connection between G P`\“D,o.A\\\\ mining method5, Kreuz rankfuner Scandinavia, Benelux and the UK in the north and Frankfurt/Main Airport J; Switzerland, Austria, Italy and Greece in the south. With - 1-6, km E? gpggd railway a maximum speed of 300 km/h travelling time for :T to Mannheim passengers will be reduced si gnificantly. At the central Figure l. Tunnel "Frankfurter Kreuz“. hub ”Frankfurter Kreuz” two highways are undercrossed. The tunnels are constructed by open cutting method as well as by mining method and by Main runs parallel to the ”Autobahn A3”. A new long top-down method. At the same time the run of the distance station is constructed at Frankfurt/Main airport. motorway is changed at ”Frankfu1'ter Kreuz”. Also the The line runs east of this new station in a tunr1el_ ln the undercrossing of downtown Frankfurt/Main by along area ofthe Frankfurt highway intersection a level-free distance railway tunnel is planned in the context ofthe branching southward in the direction of Mannheim and so-called proj ect ”Frankfurt 2 l ”. northward into the direction of Frankfurt/Main central station takes place. The tunnel project is situated in the northern Upper 2 TUNNEL”FRANKFURTER KREUZ” Rhine Graben, actually in the area of the so-called Kelsterbach deep block. In the Frankfurt airport area 2.1 Situation - ground ana' ground-water conditions the elder Pleistocene Main sediments in the west, at Deutsche Bahn AG builds the new high speed railway the Kelsterbach connection, reach thicknesses of 33 m line Cologne-Frankfurt/l\/Iain. The ”FrankfLu'ter Kreuz”, to 44 m under the ground surface and in the east, at the which is situated in the immediate vicinity of the Frankfurt highway intersection, 22 m to 29 m under Frankftut/Main airport between the highway ”Autobabn ground surface. Below the elder Pleistocene Main A3”, the four lane road B43 and the US-housing estate sediments follow Pliocene sediments with thicknesses Gateway Gardens, is undercrossed in several tunnels of 100 H1 to l4O m. The top layers mainly consist of up (Fig. 1). to l m thick younger Pleistocene fine- to medium The new high speed railway line Cologne-Frankfurt/ grained dune sand layers. In the area of the traffic roads carriageway slab "Autobahn A3" there are also fills, which reach thicknesses up to 8 m ...... _ ._...... » ..~. .-,.‘ ...., ,... _ .,__..__.._ ._...._ .._._..._. _ _._...... _... _ in the area ofthe embankments. The Kelsterbach terrace consist of an alternating 1'm f' 1-Q'35 f-: ,_." `_-_Y -<.»- #§‘?5>Z.f§ ¢- < -'- -- --Z» _-$' -_ ,- _; _ _ - _' _ ’ _ -j »’£: “gf-fi'_ -V _ __ sequence of medium- to coarse-grained sands and __ j€§t_;gron;¢{j _mo-5 ¢OV¢f______.__ ._ _ &€__¢_t_____mnS gravels. This is characteristicaly of fluviatile sediments. 1' - _, M .... I 'i'§;»< 7-5ii ~= V " Y _ _.... 1 V V' Below the elder Pleistocene Main sediments mainly .__u_~.,,__.__..___,_~______W __ ~~'"“'i sholcmm L top-~»---~ h<,<1<_¥rng ' ______fine- and medium grained sands of the Pliocene with ""final_ , concrete ,, __ mum] _ lmmg. . . f'UPP0’f . _ " C011 ' x _ _ .__jg irregular wavy interfaces are deposited. In some parts ____.____~lmmg bench______invcr' . A 'V 3 '_ A' __f I' ` of the upper 10 m to 15 m of the Pliocene sequence ltfgitli 3l5 §'5 ll'1|=|l‘5|i'E|E|l l l:l=l¢l=lsl2lflrl; sl=l=l=lil1!al1 _' __ '34 layers of silt and clay with thicknesses of only some centimeters up to several decimeters do appear. The ground-water level lies at approximately 15 rn to 16 m below ground surface. The general direction _____ . of flow is directed to the north-west and north to the Main, but flow is also oriented towards wells west IiiID m and north ofthe railway line. Figure-"'"""` 2. Soil profile and longitudinal section. 2.2 Tunnelling method r.Y 1'ground " r" surface _" 1-,"' J" "¢¥""4' '- ~ ’ ’_' 1. r =' A 280 m long part ofthe tunnel situated in the Frankfurt ?f55;~'1§?»€_ ei; n-_'fr ri _ 1" 7’ '_ "ff ._‘ " 7 .v 1 1 .V-. ‘Ev -_"vi " "'~: _ "FV ' ` Z, _` -~ `__é,‘.€_l1;:'“:flE""¢__ highway intersection area beneath the ”Autobahn A3” _;?";§é ~' _ -/__ ¢_§r..::§1%§f§,§u,{,i-4; _f;__fg.}‘ ' __ _ ‘ ‘!§¢ v!~1FfJ= "7*‘f Ti; was built in mining technique. The tunnel crosses the H ___,Q _ r - 1. ‘fsyv-:_ '_ }_>~_ _' _ iz- §:3.__.§:5'j1`f:__§'_'e ”Autobabn A3” in a smooth angle. A modified spread concrete method with a driving in several parts was ‘-'»/\».-,`=;;§;Q2"2 - : \' '.\;f= ;:: 'Zu /f' ._ chosen with spread concrete initial lining and trailing Sand/gravel T -'-'a;f'¢'>i4l"2¢S§; .1 ?G>i€-'Z S-T :fi final concrete lining. The tunnel construction is planned 1E2--> 2 _rer::::.f;;_=‘dn# ','Ff`_;', _.F ¥?F.5:.-'-_= Si2 -"Z +L' 4-_'EZ as a mouth-profile with a constant cross-section of 150 square meters. This shallow tunnel is driven with a minimum cover of 15 m. Surface settlement was ' ' limited to a maximum value of 5 cm. For fullfilling this _ _.,,f.__ ,.,,-ji.riff j_` ‘Gif »j“:*" -»’2.:7` ‘“ _ __;_~r_:rm¢».¢¢_rr~-=;§;f~>2_>rz@:A~r»¢'=.»z»; ififtp' " demand and in order to guarantee the safety ofthe tunnel ¢._§.wa=»:_5ra@*@fS““' _ '1 #__ _mv =- init*-~ pi `f`f"1 _ ` A \ and to protect the staff from material falling down %‘ -?§&1?€f’ C -=;~f¢,;_5=¢m_=~_¢ 4 additional constructive support for the driving was -»;r,¥i4 =2m;.i‘-,"=\‘r»A=‘T,i2*"*fl - -V _'_-» 4 , necessary. First the crown (top heading) is driven along the entire tunnel route under the protection of a securing Figure 3. Stability analysis forthe working face with jet grouted roof cover hurrying on ahead (Fig. 2). After rigid plastic computational model. the completion of a 14.5 m long jet grouted roof cover the crown drivage was carried out on a stretch of about 10.8 m. The length of the cuts varied between 0.8 m the crown~feet’s area. and 1.0 m. The tunnel’s bench and inverts lying in the To guarantee the stability of the crown all work was groundwater are excavated under the protection of a done with a supporting core and a plugged face by waterproof j et grouted floor cover. The bench and invert means of eight jet grouted columns. The constructive excavation with the closing ofthe spread concrete initial support consisting of Support core, jet grouted roof lining’s ring' was carried out after the cutthrough ofthe cover and jet grouted face columns was required to crown with lengths of up to 2 m. The 280 m long tunnel ensure the stability of the working face. In Figure 3 a line was subdivided into several drivage sections by possible failure mechanism ofa collapse is shown. This vertical jet grouted transverse bulkheads with distances failure mechanism is used for a stability analysis with of 30 m resp. 60 m. The tightness of this jet grouted a rigid plastic, empirical calculation approach (Belter floor was proofed with a test lowering in each et al., 1999). bulkhead. Because of the cementation of some gravel Following the course of each jet grouted roof cover in the bench the test lowering was not able to establish the crown was widened in a cone-like manner as the a safe statement about the thightness of the whole driving progressed. In order to create a closed bearing bulkhead. The other tunnel lines were built under ring and an additional stiffening element the 20 cm thick protection of impermeable building pits with anchored top heading floor was gone over with spread concrete underwater concrete floors. immediately after excavation. 4 m long piles/anchors leaned outwards were integrated in distances of 60 cm in order to ensure stability and to reduce settlements in
118 ¢,,,‘.‘,’. ‘ ' jet ‘ grouted‘QQ 45roof cover_ , ,gp cl_ri~.fing.scCti011S Q'| ~ll* ;U A`,O "fl <11 me ‘ tunml. 1 3 gWé l O support ‘ `°Qf°uI@d1 1==`$core 0 VVZ. A V-‘. five 2 V:__ iolwnns 2.12 ‘..=_ .;` . ____Q r A.. U§ ---'_-"' ‘> il1 l i*_'°’Lf H' 'B0 'IVA*Q EAA 6 .~ 2 lim test_ fSample rLib *V ‘ 2=_-.Q ~A_`” _.`-_ 5El A-'= _» *lr __ » .,.A rA. _L __ . _.:. __.¢,¢ :A A: . .:- in-situFigure 4. Cross-section large of the scaletest installation shear of the VAZA tests. “‘_,A >_-~V _": ;,_QQ..,i.3..l.§=;._.;.: i'ifi"i3iQ _~"1» »-,‘:1 i;_ff3;.*
2.3 Soil investigation Figure 5. Map of ground surface settlements [cm]. Before and during soil investigation large scale triaxial tests were carried out to determine the strength of the The in-situ large scale shear tests have shown, that soil at the area ofthe tunnel Lmdercrossing the ”Autobahn the limit surface of maximum stresses ofthe sands and A3”. The triaxial tests were carried out with cylindrical gravels could be described by a non linear power law. samples of 25 cm in diameter and 63 cm in height. For The non linear failure criteria in Mohr’s plane can be determination of Young’s modulus a series with loading Written as and reloading cycles was performed (Arslan et al., 1998). During the construction stage two series of in-situ 'c=Ao"+Bo (1) large scale shear tests were performed in the crown to A, B and n are material parameters. Bishop (1966) verify the soil parameters from triaxial tests. Also a and Jaeger (1971) suggested the value of the material new series of triaxial test was performed with material parameter n = 0.5 for sand and gravel. The parameter received out ofthe tunnel face. In the in-situ large scale A and B were detennined by the in-situ large scale shear shear tests samples of a dimension of 1 m >< 1 m >< tests to A = 9.82 and B = 0.34 (Katzenbach et al., 1998). 0.3 m were tested in a newly constructed in-situ shear The maximum shear stress is reached after large apparatus (Fig. 4). deformations. The use of these shear stresses in the The in-situ shear apparatus consists of devices to calculation is only allowed if the large deformations apply normal force and shear force and a computer don’t cause any damage. This was not possible in case based measuring system recording forces and of tunnel ”Franlcfu1ter Kreuz” because ofthe interaction defonnations. With a large steal construction it was between settlements and serviceability of ”Autobahrr possible to distribute the loads in the tunnel’s top. In A3”. that area, in which the loads were concentrated, the The resulting shear parameter of the encountered initial spread concrete lining had a double thickness. sands and gravels were very close to those obtained With this construction it was possible to obtain loads by the laboratory triaxial tests. The large scale triaxial up to 600 kN. The test samples were placed near the tests and the in-situ large scale shear tests lead to a support core in a region of undisturbed soil. To achieve very reliable and economic turmel construction. a region of undisturbed soil the support core was left intact about 2.5 m longer in the last round. The test samples were carried out ofthe support core by manual 2.4 Measuring Work. The bond of the samples with the surrounding During tunnel driving an extensive measuring program soil was left intact. The surrounding shear surface was was carried out in order to control stability and to covered with steel plates to achieve a plane surface document occurring movements. The settlements at the with a small angle of friction. After the preparation of ground surface are documented by a fully automatic, the test samples the shear case was placed around the electronically controlled digital measurement sample and the gap was filled with mortar. After summon documentation-system. A multiple-pole extensometer at up the normal stresses with manually controlled presses two defined measuring cross-sections rectangular to the the shear case was lifted, so that a gap of 1 cm was tunnel axis aiid at a measuring cross-section in the central achieved between case and shear surface, because of reservation of the ”Autobahn A3 ” controlled the this the friction between shear case and surface was settlements of the ground above the turmel and ground successfully prevented. After a preloading phase the movements in the area ofthe crown-feet. The horizontal shear force was heighten, so that the test samples could ground deformations as well are measured at two be sheared of with a constant speed of 1 mm/min. defined measuring cross-sections vertical to the tunnel
119 WI . est _ hast_ Frankfurt Market m_a_s.l_¢;@mf;11 gfanon Mann Tl 1131] I _ East~ docks Oticnbach station
_x ";;r:`;`;=¢;:;""gp_ggi---_;3<~>5;;1;;_.;.-=» ':>'-:_', I .Zg;;;~:k;Q._iff ,______'_ *_ __ ,__ :_ ' 'fr _bf I-_[EE¢f=§E=ff‘°E§-3'?{? _ _
-921%-><¢$'<"=¢;¢.1:=~=:>1=:==1¢1-'fr"_ f;$:;;;¢>31=z;_=:=):€'=;¢=:'<;;Isl-E=;1<:2:;S's:=;3:=:=;_¢~ .i;I?,<<-a_-.»'f'.-v:-'~5.-:iz-.3-:E < ' §?f`_>¢_- 1 9-"Q f,, _ -_»_~2<~ _<- ;..w¥ ‘-°' /‘ _ '2<4>2=f:’;=-§- '11-,;-‘§f_=. "" -QQ -_¢1;_-jjj"--_s_=§j~' gg,:¢.-:-,;_;-:¢;<»:,;1. 100wwe-4<-»4A_¢~,<
3 ”FRANKFURT 21” _ ramp
The project ”Frankfurt 21” is the' construction of a new S stress-level railway tunnel undercrossing the city of Frankfurt/Main. ol, G3 biggest resp. smallest value of the principle This tunnel will be 6 km in length, consisting of four stress tubes each containing one track and being 10.6 m in rp friction angle of the soil diameter. Thus, Frankfurt central station, today a c cohesion of the soil. railhead, will be tLu'ned into a through station. The tracks in and near the station will have to be lowered 15 m The effect of a ground-water lowering can be approximately. Various S-Bahn lines and underground observed. ln Figure 7a the level of the water head tunnels as well as numerous buildings and skyscrapers exceeds that of the tunnel roof by far, whereas in with their foundations, the river Main and the East Docks Figure 7b it is lowered until its level is situated below will be undercrossed (Katzenbach et al., 1997). Parts tunnel floor. As a consequence, buoyant force of the track tunnel are situated up to 35 m below correspondingly decreases in size, thus effective stress groundwater level. Figure 6 shows the geotechnical increases in size. Shear resistance will increase in size profile along the tunnel axis. As the result of the and the zones with plastic deformations (S = 1.0 in changing soil conditions from soft to hard material (from Fig. 7b) become rarer. The computer simulation shows west to east), there is a wide range of bearing capacities a much higher rate of convergence (Fig. 7c). According concerning the different soil layers. to the stress level the settlements at the tunnel roof in A tunnelling project of this scale is without parallel case of lovgpred ground-water (case b) are smaller. in Frankfurt. Hence it follows that, before the more A great number of track tunnels of the underground detailed design steps are carried out, it has to be system and the S-Bahn system in Frankfurt/Main has examined weather such a project is practicable (this been constructed using spread concrete lining and full is, whether stability can be achieved) concerning the face drilling under the protection of ground-water geotechnical problems and the challenge put upon the lowering that were extended below the base of the underground tunnelling method. Using numerical tunnel. The face-drilling generally had been divided
120 BW to the three-dimensional arch action at the working face. -_l=- . . The sturounding material has to transfer his dead load / fi -1 ‘ ,__ _ n _ along and across the axis of the tunnel in front of the PAY: / ` arch action and above and beside the tunnel. Both the compaction of the soil in front of the arch ._ ~-> 5313+ action and the destressing of the soil above the tunnel PO? .if roof cause settlements ofthe surface. These settlements |\\\.04 ‘ calotte _,ge " alone' ' go ahead and follow the driving of the tunnel. They f"~\ ~».-_-. V can’t be helped neither by spread concrete lining nor | 1\ \ \ \Q6 IQ2 \ \ \ \ by the shield driving method. I J¢Qg \ \ 0.2-f 0\\ \\a ¢¢. \\Q4\_ "` |_ 1 >__;_| _ 4 _ 4 CONCLUSION
a) b) The realisation of the projects at ”Frankfurter Kreuz” and ”Frankfurt 21” means an extraordinary challenge E number of iterations to geotechnical engineering and the power of modern ‘H2 0 25 50 L-4 tunnelling. Great geotechnical demands are made on the construction ofthe undercrossing of existing streets gs: caseb as well as for undercrossing of high-rise buildings and 344 1.5 I 4-4G1 case a the river Main. Carrying out laboratory tests and in 1: O situ tests connected with numerical simulations the so E 3 S50 proofed design model leads to a reliable and efficient ‘ai construction. cn C) Figure 7. Effect of hydraulic pressure and water. a) including water pressure REFERENCES b) without water pressure c) convergency of settlements at the tunnel roof in Arslan, U., G. Festag & M. Vogler 1998. Beitrag der [cm]. Versuchstechnik am Beispiel eines Grofitunnels der DB Neubaustrecke Koln-RheinfMain am Frankfurter Kreuz. Felsbau 5/98: 300-305. (crown, bench and invert excavation) and had been Belter, B., W. Heiermann, R. Katzenbach, B. Maidl, arranged in a construction cycle consisting of five steps H. Quick & W. Wittke 1999. NBS Koln-Rhein/1\/Iain including a top heading. By undercrossing the Main - Neue Wege bei der Umsetzung von tunnelling experience dating from the underground Verkehrsprojekten. Bauingenieur 74, No. 1: 1-7. projects is available. In addition to the ground-water Bishop, A. W. 1966. The strength of soils as engineering lowering using gravity wells the soil located in the materials. Géotechnique 16, No. 2: 89-130. tunnels’s crown area had been stabilised by freezing Falk, E. 1997. Underground works in urban the soil at the heading works. environment. Proc. of the XIV Int. Conf on Soil The widening of the mining operations in the area Mechanics and Foundation Engineering, A. A. of railway stations was undertaken in spread concrete Balkema, Rotterdam: 1401- 1406. lining, e.g. at the Schweizer Platz in Frankfurt’s district Jaeger, J. C. 1971. Friction of rocks and stability of of Sachsenhausen. Apart from using the mining method rock slopes. Géotechnique 21, No.2: 97-134. as an efficient way of ttmnelling in Frankfurt/Main, some Katzenbach, R. 1985. The influence of soil strength and projects have taken advantage of the shield driving water load to the safety of tunnel driving. Proc. 5"’ method. International Conference on Numerical Methodes The S-Bahn section ”Lot 13” for example used the in Geomechanics, Nagoya, Japan, Vol. 2: 1207-1213. shield driving method. The shield machine was open, Katzenbach, R., U. Arslan, Chr. Moormann & O. Reul therefore a ground-water lowering was necessary The 1997. Studien zur Baugrund Tragwerk lnteraktion. used shield machine had an overbreak of 13 cm to Vortrdge zum Workshop Baugrund-Tragwerk 20 cm. As a result of the late gap backfilling the Interaktion am 21. November 1997, Mitteilungen settlement on the ground surface had a total of maximum des Institutes und der Versuchsanstalt fiir 22 cm. The buildings concerned had to be protected, Geotechnik der Technischen Universitat e.g. by injections (Falk, 1997). Darmstadt, Heft 3 8. lt is known by measurements and by computations Katzenbach, R., U. Arslan, G. Festag, & A. Ruckert that the spread concrete initial lining transfers the load 1998. Soil-structure interaction influenced by in the Frankfurt clay only after closing the ring. The shallow tunneling in urban areas, Darmstadt short period before closing the ring stability is reduced Geotechnics, No. 4: 97~111.