Figure 2 Vertical cross-eeetion of the test sectien for the dynamic tests.
Two measurements are<i$lll"edistinghuished;the instrumentedring and the surroundings (ins~ed piles, instrumented cones and surface stations).
The instturnentedljng is equipped witb aecelerometers and strain gauges. Ring3ç7, located
550
m from tbe western en-trance oftbe soutberntube, was ehosea to be instrumented, as its surrounéiags (soHatlrlsurface) are ratheruadisrurbed, 'Ibis is beneficia! fer thequality oftbe vibratioo transmission meesure-ments.0/360o
Figure 3 Cross-section.ofthe instrumented ring (nr. 367). Positions of theaccelerometers (V),sttain gauges (R), water preesure devices (W) and soilpressures devices(G) are indicated,
The. Betlek Railwaytunnel is made of rings consisting of seven segments andakeystone,.as shown in flgUte 3. The ceatre ofeachsegment in tN;:instrumented ring is eqaipped witb 3 ac-celerometerstbat measm-e eitber radial, tangential or axia! secel-erations oftheeoncr~$urface. The strain gauges are embedded in tbe conereteand arealso located intbe centre of the segments.
Six deviees havcbeen inserted in tbesoil tbroughtbe !iniog of tbe instrumentedring. There are three waterpressare deviees near tbe ring, as indi~ed in. figure 3. Also, tbree soit pressnre devices have been in~edthroughtbe lining. These deviees meesure variatien in pressure levels due to vibration emission in tbetunnel.
Table I summarisee the49 data channels in thetunnel. The first channel is used to measure the sine function of the harmonie excitation, This signal is also used to synchronise the two data aequisition systems, The latter is essenrial for the proper deter-minetion of transmission and coherence functions,
Tab1eISummary of data cbanne1sin the instrumented tunnel ring.
channei instruments type nrs water pressure devices ground pressure devices
The secend measuringsystem is formed bythe instruments in the surrcundings of the instmmented ring. The following instru-ments are used to determine tbe response of tbe surroundings:
• 3 cones at 20 m below grsde, iastrumested witb aeceler-omesers (30);
• 3 prefab pUes, instrurnented witb 9 accelererneters (top and bottom);
• 9 accelerometers on the surf ace.
Table 2 summarises thechannels thatare measured in the
SUT-roundings during a test. The positions of the instruments relative to ring 367 are indicated in tbe plan given in tigure 4. Table 3 links the positions in figure 4 tothe insttuments in tbe the sur-roundiags,
Table 2 Summary of data channels in the surroundings of the instru-mented tunnel ring.
channel instruments accelerometers pile heads accelerometerspile points
Table 3 Summary of instrument positions in the surroundings of the in-strumentedtunnel ring. Position numbers !"eferto figure 4.
positron instruments channel
nr nrs
The two systems contain 84channels te messure the response of the tunnel and its surroundings.
3 TEST PROGRAM
Three testsessiens are plaaned. In test sessioas I and 2, a shaker is located near tbe instrumented ring and wiUserve as tbe souree of tbe vibrations. The positionof tbe shaker will be varied as in-dicated in figure 4. The variable posiriea of the shaker eaebles a study of dîstaaee effects in tbe transmission of vibrations tbrough a segmented lining.
The shaker is made from rwoeecentric rotating masses that are powered by a frequency-controlled motor. The force applied to tbe tunnel depende on the angular frequency oftbe messes. A statie pre-lead of 5 kN is used to prevent uplift of the shaker.
Bolting is therèfore not reqaired, but it limits the force amplitude to 4 kN (single peak),
Test sessions 1 and2 will he conducted in various stages of the tunnel huildingprocess. The first session was performed in April 2000 whenthe$Outhern tube was finished, but without in-lay and before constrllclion of the northern tube. The secend ses-sion is planeed for$pring 2001, when both tubes will becorn-pleted, It wiJl consistof two scenarios, In the first, the shaker is placed in the southemtube, including inlay. In the second, the shaker is placed in thenorthern tube.
In test sessien 3,theresponse. of tunnel and surroundings to freighe trains will~lneaSured. The instrumentation will be the same as in the previous sessions, exeept for the train dereetion sensor that will beuSC(jto trigger the data acqnisition system,
3m
Figure 4 Plan of the testsection. Positions of the harmonie force inthe test programare indicatl;(jbythe grayrings, Eachring bas a length of 1.5 m, The six posîtîOllSoftheinstruments in tbe surroundings are also given,
During the tests .with the shaker, the frequency of the har-monie force is vari«ifrom 5 to 85Hz. From 5to 40 Hz, ineer-mediate steps of I •
Hz
are taken. Above40
fu,steps of 2 Hz are taken. At eaeh fre(}ueney step, the response ofatl channels is measured during.32secondsat 500 Hz sampling frequency.By means of FoUTiertransfonnation, the frequency-dependent response ofeach. ehailne1 maybe detennined ..Since the forcing frequency is known,tbis isa veryaeeurate testingmethod for the assessment of vibmtiontransmission using only a small shaker force.
4 PRELlMlNARY RESULTS
Thefirst test sessionwas cenducted in April 2000. About 2 Gb of data was collected.Althougha comprehensive analysis is still te beconducted,somcpreliminary results caa be given.
Fitst, .the qualityofthe measurementll isvery high. This may be detennlnedfrmnthevery high coherence(largertban 0.99 for most of the frequenc;ysteps) between forceand response signaIs.
See, for example, figllI'e6, whieh showstheeoherencefunetions that eorrespond to thetmnsmission functions glvenin figure 5.
The vibmtion .transmission was detennined fromtbe shaker 10-cated inting 367tO(hevertical component of the vibration ve-locity in tbe threesurface stations. Above 2()Hz, theeoherence is near perfeetandeven at low frequencies the coherence is aboveO.9.
This proves the power of harmonie excitation testscompared to impact tests. The latter type of testing norrnally shows low to moderate eoherence for most of thefrequencies and high eoher-ence near the peak in the traasmission. An impact test will, therefore, give reliable results that are limited to the peaks in the transmission. For a reliable viaration prediction, however, the coherence .should be high for the complete frequeney range (ap-prox. 5 to 80 Hz) that is important in the assessment of vibration transmission.
V;t,3 __ Vz.,5
30 frequency (Hz)
Figure 5 Vertical vibration transmission fancrions for the three surface stations. Shaker position: ring367.
_ ... Vz,J
Figure6Conerenee tunenons corresponding to the transmission func-tions in figure 5.
Secondly, ir may be stated that there is a general trend toward symmetry in vibration transmission from a souree in a seg-mented lining to soil and surfaee, In tbe first test session,· the shaker was positioned 3 ringg left or right ofthe instrulnented ring (rings 364 and 370 respectively In figure land 4). The vi-bmtion transmission functions that were cakulated from the two shakerpositionsshowed, in general,good agreement.
Thirdly, itseeass that tbe force that tbeshaker. applies to the tunnelfloorquickly migratèS tothesoil. This issuggested by the low vibration transmission tothe accelerometers and strain gauges in the eciling of the instrumented ring. It caneven be stated that the· nuijorityoftbe vibration transmission runs threugh thetwo segmeate in the hottom. Thesesegments can be recognised in figure 3 by tbe positioe of accelerometers V7 and VI.
Thc test program of tbe seeoad test session will be sIightly different from the fitst sessies, based. onthe experîences in the first session, The preliminary results suggest that larger fre-queney steps are allowed, making the testing program more effi-cient.
5 FUTURE RESEARCH
The tests are a part of general research program on dynamics in The Netherlands, initiated by the Centre for Underground Con-struction (F300, 1997). Predietiens of the tunnel behaviour due to a train passage are made using numerical models, Also, nu-merical models are u$èdto predier the transmission of vibrations through the (soft) soilto the piles, In the evalaation phase of the research program, predictions and measurements wiJl he com-pared. It is anticipated that this comparison win lead to a hetter understanding of the dynamies of tunnels in soft soils,
6 CONCLUSIONS
The use of harmonie excitation yields vibration transmisston functions of high quaJity. Even at low excitation force levels, a reliable vibration ttansmission function can be determined from a force in a tunnel topoints in the surroundings.
7 REFERENCES
F300, 1997. Genera! project plan experimental research at the Botlek Railway Tunnel. Gouda.
Petronio,L.,Poletto, F. & Schleifer, A., 1999. "Seismie-While-DriJling using the tunnel boring machine noise", Geo-physics, 64 (2),Pl'.452-456.
Sattel, G. Sander. B.K., Amberg, F. & Kashiwa, T., 1996.
"Tunnel seismie prediction TSP - some case histories". Tunnels and TunneJing, Pl'. 1-16.
-" ...•..•.•.••.•...•..•.•...••.••....•....•...•... -
.••....••..•••.•...•.•..••...•···-~···III ...
IIIIIIIIIIIII····=I·~IIII· •..•... 1IiII.·... ••1IIIIIIIl~.:(U
.".:..: '._:•• Ta
::.... .
•_II:IIIIIIIJI.·IIIIIIIIII·iIlIIIl···~iIIIIiilllll:II•••••-·.·•••
i .•.•.. ·.··· . •.•....••... . .
....
..-
• -.... ':. 1.11
ar)
\0 ("")-.:::t ...-1 N
00 00 00 00 00 00
0 c, 0 c, 0 p.. 0 c, o, 0 0 ~
ar) s .
00
N
;:~&@nMMrlit~:i~
("")
--. I .~ J ~ J .
~-~---;r-·--·--7·_·--!ill@m_iW ~
("")
• •
~. 8 •. ~ S "'~S ••.
00 00 00
CI':J
0
~iiiIm~mT:f;:
0..
._._._
..._._
...~ _._
..._._
•...._._
...•
._u••• u •••••-.u...
• Vtffi~~
.••...
o ~
00 ~ ~ ~ ~ ~N - ~
- 000 0 0 000 0 0
o
00o
Vio ~
o
No
-o
0
0 0\ 00