Available online at www.sciencedirect.

com
Tunnelling and
Underground Space
Technology
incorporating Trenchless
Technology Research
Tunnelling and Underground Space Technology 24 (2009) 14–21
www.elsevier.com/locate/tust

Numerical simulation of rock deformation for support design

in tunnel intersection area
F.Y. Hsiao a,b,*, C.L. Wang b, J.C. Chern a
a
Geotechnical Engineering Research Center, Sinotech Engineering Consultant, INC, Taipei, Taiwan
b
Department of Resource Engineering, National Cheng Kung University, Tainan, Taiwan

Received 4 July 2007; received in revised form 19 December 2007; accepted 18 January 2008
Available online 7 March 2008

Abstract

Passenger and vehicle adits and ventilation shafts are commonly used for emergency access and ventilation in long tunnels. Increasing
support load and additional tunnel deformation may endanger tunnel stability during construction in the intersection of the access and
main tunnels. To understand the mechanical behavior of rock masses in the intersection area, 75 cases of 3D numerical analysis were
conducted. These analyses were conducted under various tunneling conditions including rock strength, rock mass rating, overburden
depth, and intersection angle. Following analysis results, a criterion for assessing the effect of intersection angles on tunnel behavior
has been established, and three categories of support design suggestions for different geological conditions have also been proposed.
Ó 2008 Published by Elsevier Ltd.

Keywords: Tunnel intersection; 3D numerical analysis; Support design suggestions

1. Introduction having a rational approach for the support design in the

intersection area is important.
Passenger and vehicle adits and shafts are often con- Most of the previous studies on tunnel behavior in the
structed for emergency access and ventilation in long tun- intersection area were conducted by studying stress concen-
nels. Rock behavior at the intersection area of the main tration factor using elasticity theory. Photo-elasticity exper-
and access tunnel is a complicated, three-dimensional prob- iments were conducted by Riely (1964) and Pant (1971) to
lem. Increasing support load and additional tunnel defor- study stress surrounding tunnel intersection areas. Due to
mation were often observed during construction due to rapid development of modeling techniques, 3D analysis
further disturbance of rock masses surrounding the inter- were used more widely by researchers, such as Thareja
section area. Conventionally, a support system heavier et al. (1980, 1985) and Takino et al. (1985), on the displace-
than the normal tunnel section was adopted to counter ments and liner stresses in the intersection area by consider-
the adverse effects of complicated stress conditions. In long ing various rock properties and intersection angles.
tunnels, such as the 12.9 km long Hsuehshan tunnel in Tai- As for the design of the support system, Tsuchiyama
wan, 42 intersections were constructed for passenger and et al. (1988) examined the excavation of an access tunnel
vehicle adits and ventilation. The heavy support system with an oblique angle of 45° to the existing main tunnel
adopted in this project created significant cost. Therefore, through a 3D numerical analysis. It was found that the
affected area along the main tunnel, which requires addi-
tional support work, is about one tunnel diameter on the
*
obtuse angle side and about three tunnel diameters on
Corresponding author. Address: Department of Resource Engineer-
the acute angle side from the point of intersection. The
ing, National Cheng Kung University, Tainan, Taiwan. Tel.: +886 2
27580568; fax: +886 2 27290273. practice adopted by JNC (Japan Nuclear Development
E-mail address: [email protected] (F.Y. Hsiao). Institute, 1999) requires that additional reinforcement be

0886-7798/$ - see front matter Ó 2008 Published by Elsevier Ltd.

doi:10.1016/j.tust.2008.01.003
 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21 15

installed at the intersection area extending four tunnel While it is impossible to have design suggestions for tun-
diameters on the acute angle side and around the same dis- nel intersections under various geological and geometrical
tance on the obtuse angle side. More recently, the area with conditions, degrading the rock mass rating and support
additional reinforcement was suggested by Nonomura system in the intersection area are used in practice. Never-
et al. (2004) for the rough estimate of support requirement. theless, unexpected tunnel closure and support failure have
Chen et al. (2002) and Hsiao et al. (2004) conducted 3D been observed in several cases in Taiwan. A more rigorous
elasto-plastic analysis of tunnel behavior in the intersection design method is, therefore, needed.
area. However, these studies were only limited to the case
study.
2. Numerical study
Table 1
Case conditions of tunnel intersection
For comprehensive understanding of rock behavior in
Rock strength Soft rock (rc = 100 kg/cm2), medium strength rock the tunnel intersection area, a series of 3D numerical
(rc = 500 kg/cm2), hard rock (rc = 1000 kg/cm2)
analysis were conducted by Hsiao et al. (2005). The analy-
Rock mass RMR = 30, 50, 70 ses were conducted under various tunneling conditions,
rating
including rock strength, rock mass rating, rock covering,
Rock covering 100, 400, and 700 m
Lateral stress 0.5, 1.0, 1.5 and intersection angle, are shown in Table 1. Tunnel span
ratio (k) of 12 m in the main tunnel and 8 m in the adit were used in
Intersecting 30°, 60°, 90° numerical simulation. 75 numerical cases of different
angle tunneling conditions were analyzed, as shown in Table 2.

Table 2
Case number and tunnelling conditions
Rock strength RMR K=1 k = 0.5 k = 1.5
H = 100 m H = 400 m H = 700 m H = 100 m H = 400 m H = 400 m H = 700 m
Hard rock 50 #1 M22 38 #2 M23 39 #3 M24 40 – – #66 #67
70 #4 M25 41 #5 M26 42 #6 M27 43 – – #68 #69
Medium strength rock 30 #7 #8 #9 #54 #55 #70 #71
50 #10 M28 44 #11 M29 45 #12 M30 46 #56 #57 #72 #73
70 #13 M31 47 #14 M32 48 #15 M33 49 #58 #59 #74 #75
Soft rock 30 #16 #17 – #60 #61 – –
50 #18 M34 50 #19 M35 51 – #62 #63 – –
70 #20 M36 52 #21 M37 53 – #64 #65 – –
#: Case of 90° intersection angle; M: case of 60° intersection angle; : case of 30° intersection angle.

Table 3
Excavation and support design of tunnel intersection
RMR 70 50 30
Excavation process I ? II ? III ? IV
M (Main Tunnel)
A (Adit)

III I

IV II

Round length 2.5 m 2.0 m 1.5 m

Shotcrete M: 15 cm M: 20 cm M: 25 cm
A: 10 cm A: 15 cm A: 20c
Rock bolt M: 25 mm £, M: 25 mm £, M:25 mm £,
L = 4 m @ 2.0 m  2.0 m L = 4 m @ 1.5 m  1.5 m L = 6 m @ 1.5 m  1.5 m
A: 25 mm £, A: 25 mm £, A:25 mm £,
L = 4 m @ 2.0 m  2.0 m L = 4 m @ 1.5 m  1.5 m L = 6 m @ 1.5 m  1.5 m
Steel rib – – M: H150  150 @ 1.5 m
A: H100  100 @ 1.5 m
Forepoling – – M: 32 mm £, L = 3 m, @ 1.0 m
A: 32 mm £, L = 3 m, @ 1.0 m
16 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21

Excavation process and support design are shown in simulation. The strength and deformation characteristics of
Table 3. the rock mass were evaluated by the empirical methods
The computer program, FLAC-3D, was used for the proposed by Hoek et al. (2002) and Hoek and Brown
simulation of tunnel construction. Tunnel meshes for the (1998). These empirical methods have been widely used in
different intersection angles are shown in Fig. 1. The mesh design practice and verified by back analysis of case histo-
boundary from the main tunnel is about 4.8 tunnel diame- ries in Taiwan (Chern et al., 2005).
ters in the horizontal direction and about 5.2 tunnel diam-
eters in the vertical direction. The length in the main tunnel 3. Results of analysis and assessments
axis direction is varied different from 12 to 15 tunnel diam-
eters for various intersection angles. The elasto-plastic The results of tunnel displacements and plastic zones at
model of Mohr–Coulomb criterion was used for numerical different locations and construction stages were obtained.
The symbols for tunnel displacement at different locations
and construction stages are explained in Table 4.
The strength/stress ratio (rcm/P0) of rock mass, where
rcm is the uniaxial compressive strength of rock mass and
P0 is the field stress, was found to be an important factor
for tunnel behavior (Chern et al., 1996; Hoek and Marinos,
2000). According to the results of analysis, the roof settle-
ment of tunnel in the intersection area increases with
decreasing strength/stress ratio of rock mass, as shown in
Fig. 2. The increasing roof settlement was found to acceler-
ate sharply when the rcm/P0 is less than 0.5. The strength/
stress ratio was used as an index for rock condition in fol-
lowing discussions.
To estimate the effect of intersection angles on tunnel
deformation, relationships between additional roof settle-
ment (Ddm/dm0, Dda/da0) and intersection angles are shown
in Fig. 3. The additional roof settlement in the cases of 30°
intersection angle was significantly greater than the cases of
60° and 90°. Moreover, the increasing tendency became
obvious for all intersection angles when the strength/stress
ratio of rock mass was less than 0.5.
Plotting the turning points of the regressive curves in
Fig. 3, a curve which delineates the severely affected area
from the slightly affected area due to access tunnel con-
struction can be obtained, as shown in Fig. 4. This provides
a criterion for judging the potential effect of access tunnel
construction. In the shaded area of Fig. 4, the effect of
access tunnel construction is expected to be severe. On
the other hand, above the shaded area, no significant effect
is expected due to the access tunnel construction.
In practice, the access tunnel will be located in area with
better rock quality. From these results, 90° intersection

Table 4
Symbols for tunnel displacement at different locations and construction
stages
Symbol Description
dm0 Roof settlement of main tunnel after main tunnel was excavated
dm Roof settlement of main tunnel after main tunnel and access
tunnel were excavated
Ddm Additional roof settlement of main tunnel due to the excavation
of access tunnel (Ddm = dm–dm0)
da0 Roof settlement of access tunnel far from the intersection area
da Roof settlement of access tunnel in the vicinity of intersection
Fig. 1. Numerical meshes for different intersection angles: (a) 30° area
intersection angle; and (b) 60° intersection angle; and (c) 90° intersection Dda Additional roof settlement of access tunnel in the intersection
angle. area (Dda = da da0)
 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21 17

40 40
Main Tunnel -- Roof Settlement Adit -- Roof Settlement

30 30
δm (cm)

δa (cm)
20 20

10 10

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
σcm/ P0 σcm/ P0

Fig. 2. Relationship between roof settlement and strength/stress ratio of rock mass in tunnel intersection area: (a) roof settlement of main tunnel; and (b)
roof settlement of access tunnel.

80 160
Main Tunnel -- Roof Settlement Intersection angle=90 Adit -- Roof Settlement Intersection angle=90
Intersection angle=60 Intersection angle=60
140 Intersection angle=30
Intersection angle=30

60 120

100
Δδm/δm0 (%)

Δδa/δa0 (%)

40 80

60

20 40

20

0 0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0
σcm/ P0 σcm/ P0

Fig. 3. Relationship between the additional roof settlement and the intersection angle of tunnels: (a) additional roof settlement of main tunnel; and (b)
additional roof settlement of access tunnel.

angles seem to be the best arrangement for access tunnels. used to solve the problem in practice. However, the appro-
Only when rock mass in highly squeezing condition, i.e., priateness of tunnel design from experience still required
rcm/P0 6 0.25, will it has significant effect. On the other more examination by using numerical analysis for impor-
hand, in situations with sharper intersection angles, the tant projects. A guideline for the support design in tunnel
access tunnel must be located in areas with better rock con- intersection area is discussed as follows using the results
dition in order to avoid the adverse effect of access tunnel of numerical simulation obtained.
construction.
4.1. Supplementary support works
4. Support design for tunnel intersection
The assessments of tunnel stability for the intersecting
A critical issue for tunnel intersection is the design of cases were conducted first. An empirical safety criterion,
supplementary support, including the increase of support- which was proposed by Sakurai (1983, 1993) and success-
ing strength and range with additional reinforcement. fully applied to jointed rock masses in Taiwan by Chern
The experience of engineers and past cases were commonly et al. (1998), were adopted to evaluate tunnel stability.
18 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21

2.00
According to the increment of roof settlement at main tunnel Main tunnel excavation
According to the increm ent of roof settlement at adit 10.00 Adit excavation
1.75

1.50 Case9

Case8
1.00 Case12
1.25
Case21

εc=δ/r (%)
III
σcm/ P0

Case11
1.00 Slightly affect Zone Case18
II

0.75 I
0.10

0.50

0.25
Gray Zone
Severely affect Zone
0.01
0.00 0.01 0.10 1.00 10.00 100.00 1000.00
30 60 90 σcm (MPa)
Intersection Angle
Fig. 6. Stability assessment of tunnel with perpendicular intersection for
Fig. 4. Criterion for judging the effect of intersection on tunnel behavior. moderately squeezing rock (0.25 6 rcm/P0 6 0.5).

Three warning levels: I, II, III, were established from the

relationship between tunnel crown strain (ec) and uniaxial Main tunnel excavation
10.00 Adit excavation
compressive strength of rock mass (rcm). They were pro-
posed as an empirical criterion for tunnel safety assess- Case17

ment. A stable condition is expected when the data point

falls below warning level II; on the contrary, severe tunnel Case19

closure and support damage may be encountered when the 1.00

εc=δ/r (%)

data point lies above the warning level III. Between warn- Case16 III
ing levels II and III, is a transition area. Some tunnel sec-
tions may be in stable condition while others develop II

minor support damage. The tunnel safety of the numerical I

0.10
cases with 90° intersection were assessed and the results for
different rock conditions are illustrated in Figs. 5–7. It may
be seen that for non-squeezing or slightly squeezing rock
(Fig. 5), the tunnels will be in stable condition. For moder-
0.01
0.01 0.10 1.00 10.00 100.00 1000.00

σcm (MPa)
Main tunnel excavation
10.00 Adit excavation
Fig. 7. Stability assessment of tunnel with perpendicular intersection for
highly squeezing rock (rcm/P0 6 0.25).

1.00 ately squeezing rock and highly squeezing rock, the tunnel
εc=δ/r (%)

III may be in stable or unstable condition, depending on

Case3 Case15

Case7
actual rock condition encountered.
Case2 Case14 II
Case6 Nowadays, three categories of design philosophy are
Case20

0.10
Case10
Case5 I usually used in the tunnel intersection, and are described
Case13
briefly as follows:
Case1

Case4 (1) Only local strengthening of the support system is

designed in the intersection area, such as increasing
0.01 the thickness of shotcrete, and the density or length
0.01 0.10 1.00 10.00 100.00 1000.00
of rock bolt.
σcm (MPa)
(2) There is no difference in support design between the
Fig. 5. Stability assessment of tunnel with perpendicular intersection for intersection and normal tunnel area. But the measure
slightly or non-squeezing rock (rcm/P0 P 0.5). of degrading rock mass rating is used to install a
 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21 19

heavier support system in the intersection area. For the so-called ‘‘good geological area”. At such an area,
instance, one third of the Q-value is suggested for severe tunnel deformation is not expected. Strengthening
the tunnel intersection. of support system is mainly aimed at avoiding local poten-
(3) Heavier support system is directly designed in the tial wedge failure. According to Fig. 5, the tunnel deforma-
intersection area. For example, type IV support sys- tion for the slightly or non-squeezing rock were almost
tem is used to sustain the type III rock. below warning level II. The tunnel will stay in stable con-
dition with no excessive tunnel closure or severe stress
The first design philosophy is due to the fact that the relaxation. Therefore, the first design philosophy is reason-
location of a tunnel intersection is commonly chosen in able for the ground of slightly or non-squeezing condition
(rcm/P0 P 0.5).
The second and the third design philosophies adopt the
Table 5
heavy support system to strengthen the ground around the
Guidelines for support design in tunnel intersection area
intersection. Basically, the designs are generally too conser-
Geological Support design suggestions
vative for slightly or non-squeezing rock. The design phi-
condition
losophy is more suited for moderately squeezing ground
Slightly or non-  Only strengthening support system to prevent
by reducing tunnel deformation whose values are between
squeezing rock potential wedge failure in the intersection area,
(rcm/P0 P 0.5) such as increasing the thickness of shotcrete, the warning levels II and III (see Fig. 6). For highly squeez-
and the density or length of rock bolt ing conditions, the tunnel deformation has exceeded
 Monitoring instruments should be installed to warning level III (see Fig. 7). Merely reducing tunnel dis-
examine tunnel stability placement is insufficient for tunnel safety. Adoption of
Moderately  Reducing rock mass rating in order to install hea-
heavy support design alone may not overcome extremely
squeezing rock vier support work or design a more conservative
0.25 6 rcm/ support system in the intersection area. poor rock conditions. Auxiliary measures, such as special
P0 6 0. 5)  Monitoring instruments should be installed to excavation arrangement and/or ground improvement by
examine tunnel stability. Carry out detailed grouting, should be used to improve tunnel stability. Based
visual inspection on tunnel conditions and on prior discussions, three categories of support design
increase monitoring frequency during
suggestions for different geological condition are proposed,
construction
Severely  Conservative excavation processes and support are shown in Table 5.
squeezing rock system should be used in the intersection area.
(rcm/ Auxiliary measures, such as ground improve- 4.2. Range of additional reinforcement
P0 6 0.25) ment, should be made if necessary. 3D numerical
analysis is suggested to assess the appropriateness
The range requires additional reinforcement in the tun-
of tunnel design.
 Monitoring instruments should be installed to nel intersection area is another important topic. The rela-
examine tunnel stability. Carry out detailed tionship between additional roof settlement and the
inspection on tunnel conditions and increase distance from the intersection is indicated in Fig. 8. The
monitoring frequency during construction relationship can be used to evaluate the range, which

60
Main Tunnel -- Roof Settlement 90º 60º 30º
Acute Side
0
Δδm/δm0 (%)

Obtuse Side
-Y Y B 50

40
D
Area requires additional
reinforcement
30

20

10

Obtuse angle side Acute angle side

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
-Y / D Y/D

Fig. 8. Relationship between additional roof settlement at main tunnel and distance from intersection center.
20 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21

Table 6 tion area are proposed based on the squeezing condition of

Suggestions for area with additional support rock mass. Three categories of support design suggestions
Geological Extent suggestions for additional system for different geological conditions are proposed. For
condition support slightly squeezing rock or non-squeezing rock, no system-
Slightly or non- No additional system support need in atic additional support is required. Heavier support system
squeezing rock intersection area, only support in preventing is required for moderately to severely squeezing rock.
(rcm/P0 P 0.5) potential wedge failure
The area which needs additional support is also pro-
Moderately Acute angle Obtuse angle posed based on results of numerical analysis. It mainly
squeezing rock side side
depends on geological condition, intersection angle, and
(0.25 6 rcm/P0 6 0.5) 90° 1.0D* 1.0D
60° 1.5D 1.0D its location in the intersection area. However, the sugges-
30° 2.5D 1.0D tions proposed are only intended to serve as a preliminary
Severely squeezing Acute angle Obtuse angle
design guide. Re-examining the appropriateness of sup-
rock side side port design by using 3D analysis for important tunnel
(rcm/P0 6 0.25) 90° 1.5D 1.5D project or in poor geological condition is recommended.
60° 2.5D 1.5D Further field monitoring and visual inspection will prove
30° 3.5D 1.5D to be valuable in ensuring tunnel stability during
*
D: span of access tunnel. construction.

References
requires additional reinforcement. Fig. 8 shows that the
additional roof settlement in the intersection area Chen, Y.C., Chang, M.C., 2000. Study of the deformation behavior and
decreases with the distance from the intersection center. lining stresses of horizontal tunnel intersection during excavation –
On the acute angle side, smaller intersection angle causes numerical investigation. J. Chin. Inst. Civil Hydraul. Eng. 12 (3), 487–
497.
much larger increase in roof settlement than that with lar-
Chen, C.N., Chen, J.S., Lu, E., 2002. The mechanical behavior of 45°
ger intersection angle. On the obtuse angle side, however, tunnel intersection inducing by construction through various rock
the same tendency occurs only in the vicinity of intersec- masses. J. Chin. Inst. Civil Hydraul. Eng. 14 (1), 159–165.
tion center. Chern, J.C., Yu C.W., Hsiao, F.Y., 1996. A study on the current tunnel
In the design practices used in Taiwan, the area support design practice in Taiwan and effects of construction proce-
dures on tunnel behavior. Sinotech Engineering Research Foundation
requires additional reinforcement ranges from 0.5 to 1.0
Report SEC/R-GT-97-01 (in Chinese).
tunnel diameter (D) for both the main tunnel and access Chern, J.C., Hsiao, F.Y., Yu, C.W., 1998. An empirical criterion for
tunnel. This practice is based mainly on experiences. Sug- tunnel construction. In: Regional Symposium on Sedimentary Rock
gestions have been made by several researchers (Chen and Engineering, November 20–22, Taipei, pp. 325–330.
Chang, 2000; Chen et al., 2002) that the area with 10% Chern, J.C., Chang, Y.L., Tan, C.H., 2005. Re-examination on the
empirical deformation modulus relationship using back analysis results
increase in tunnel deformation should be reinforced with
of underground excavation. In: International Symposium on Design,
additional support. This is considered to be a rational Construction and Operation of Long Tunnels, November 7–10, Taipei,
approach. Comparing the results obtained in this study Taiwan, pp. 1247–1254.
(Fig. 8), the current practice adopted corresponds roughly Hoek, E., Brown, E.T., 1998. Practical estimates of rock mass strength.
to 10–20% increase in roof settlement in perpendicular Int. J. Rock Mech. Min. Sci. 34, 1165–1186.
Hoek, E., Marinos, P., 2000. Predicting tunnel squeezing. Tunnels and
tunnel intersection. For smaller intersection angle, how-
Tunnelling International, Part I-November, Part II- December.
ever, the area requires additional reinforcement is much Hoek, E., Carranza-Torres, C., Corkum, B., 2002. Hoek–Brown failure
larger, especially on the acute angle side. Based on results criterion-2002 edition. In: Proceedings of the 5th North American
shown in Fig. 8, with geological condition, intersection Rock Mechanics Symposium, Toronto, July 2002, pp. 267–273.
angle, and the location in acute angle side or obtuse angle Hsiao, F.Y., Yu, C.W., Chern, J.C., 2004. Feedback analysis on stability
of intersection area of seven tunnels based on monitoring and
side further considered, 10–15% increase in roof settlement
construction data. Chin. J. Rock Mech. Eng. 23 (22), 5012–5018.
was selected as the criteria for additional reinforcement. Hsiao, F.Y., Yu, C.W., Lin, C.C., Chern, J.C., 2005. The effect of
Suggestions for area with additional reinforcement are intersecting angle on tunnel stability and mechanical behavior in
shown in Table 6. These suggestions proposed intend only tunnel intersection area. In: Proceedings of the 11th Conference on
to serve as a preliminary design guide. It is necessary to Current Researches in Geotechnical Engineering in Taiwan, Septem-
ber 8–10, Wanli, Taiwan, pp. F06-1–F06-8.
perform 3D analysis for important tunnel projects or poor
Japan Nuclear Development Institute (JNC), 1999. H12 project to
geological conditions to ensure the safety of the tunnel establish the scientific and technical basis for HLW disposal in Japan.
intersection. Support Report 2-Repository Design and Engineering Technology,
Appendix C., pp. C31–C35.
5. Conclusion Nonomura, S., Kimura, H., Nakamura, J., Tamura, S., 2004. Design and
construction of a tunnel intersection area with special structure.
Tunnelling and Underground Space Technology, WTC 2004, G13, pp.
Results show that strength/stress ratio of rock mass 1–7.
(rcm/P0) plays an important role in tunnel intersection Pant, B., 1971. Analysis and design of pressure tunnel intersection. Indian
behavior. Guidelines for support design in tunnel intersec- Society of Engineering Geology, Tunnelling Seminar, Part I.
 F.Y. Hsiao et al. / Tunnelling and Underground Space Technology 24 (2009) 14–21 21

Riely, W.F., 1964. Stress at tunnel intersections. J. Eng. Mech. Div., Proc. Thareja, D.V., NataRajan, R., Madhavan, K., 1980. Structure analysis
ASCE, 96, Part I. and design of bifurcating tunnels. Water Power Dam Constr. 32, 42–
Sakurai, S., 1983. Displacement measurement associated with the design 46.
of underground openings. In: Proceeding International Symposium Thareja, D.V., Madhavan, K., Sharma, K.G., Natarajan, R., 1985. Three-
Field Measurements in Geomechanics 2, Zurich, pp. 1163–1178. dimensional finite element analysis of branching tunnel. In: Fifth
Sakurai, S., 1993. The assessment of tunnel stability on the basis of field International Conference on Numerical Methods in Geomechanics,
measurements. In: Proceedings of the XVIII Convegno Nazionale Di Nagoya, pp. 1193–1199.
Geotecnica, Rimini, Italy. Tsuchiyama, S., Hayakawa, M., Shinokawa, T., Konno, H., 1988.
Takino, K., Kimura, H., Takeda, N., Ito, F., 1985. Three-dimensional Deformation behavior of the tunnel under the excavation of crossing
behavior of tunnel intersection. In: Fifth International Conference on tunnel. In: Numerical Methods of Geomechanics. Balkema, Rotter-
Numerical Methods in Geomechanics, Nagoya, pp. 1185–1192. dam, pp. 1591–1596.