Numerical Modelling Fault Rupture Hazard of the Silver Creek fault, San

Jose, California
B. Ghosh1, V. Kumar1 J. Lee 2 and M. Walker3

ABSTRACT
As part of a major infrastructure project located in downtown San Jose, California, an investigation was performed to evaluate
the potentially active, 48-km-long Silver Creek fault (SiCF). This paper presents the modelling assumptions for the tunnel fault
crossing and presents the results from the analysis undertaken in LS-DYNA. The results from this analysis informed further
design decisions at the fault location and aid in the seismic design of the tunnel fault crossing.
Introduction
Tunnels built in areas subject to earthquake activity must withstand both seismic and static loading. In some
cases, a tunnel may be constructed across a fault zone as it is not always possible to avoid crossing active faults
at grade. In these situations, earthquake-induced fault movement may subject the tunnel to differential
displacements and generate stress concentrations in the tunnel liner. Additionally, the tunnel must tolerate the
expected fault displacements and seismic action and allow only minor or reparable damage. The seismic
behavior of a tunnel across a potentially active fault has been analyzed in several publications (Wang et al.
2012) and methods of designing flexible lining for tunnels across active faults include the following:
1. Excavating a larger diameter section to provide enough space for the conditions of earthquake or faulting.
2. Grouting the ground in order to increase the strength and
ductility of the faulted zones.
3. Isolating the tunnel by filling an annulus between the ground
and the lining with some soft materials.
4. Installing flexible joints to increase longitudinal flexibility of
the tunnel.
This paper presents the modeling assumptions for the tunnel fault
crossing and presents the results from the analysis undertaken in
LS-DYNA in an alluvial deposit and fault crossing. The results
from this analysis informed further design decisions at the fault
location and aid in the seismic design of the tunnel fault crossing.
Similar studies have been conducted by Donikian et al. (2020)
where tunnel design has been considered across a fault.
A companion paper (Baldwin et al 2022) presents the
probabilistic and deterministic fault displacement hazard analysis
(DFDHA/PFDHA) of the Silver Creek fault (SiCF), performed as
part of a major infrastructure project. The study identified that the
SiCF (Figure 1) lies on the eastern to central area of Santa Clara Figure 1. Uncertainty zone and tunnel
Valley, California, and intersects project’s bored tunnel model after Baldwin et al (2022).
alignment. The fault study provides information on fault activity, Approximate area of intersection
style and pattern of deformation, potential for near-surface identified in red circle.
deformation, and estimates of coseismic fault displacement for
the SiCF. These parameters were used to in the planning, mitigation concepts, and preliminary engineering
design considering fault rupture for the project.

1
Technical Principal, Mott MacDonald, London, (email: [email protected])
3
Technical Principal, Mott MacDonald, San Jose, ( email [email protected])
 Fault Location and Zone of Uncertainty
Previously, Wentworth (2018),
Chaussard et al. (2015), NORCAL
(2010) and Wentworth et al. (2010)
have conducted studies to constrain
the SiCF fault using geological
profiles, LiDAR data and
geophysical analysis. The location of
the fault splays determined by these
studies varied, thus, an uncertainty
zone have been determined by
Baldwin et al. 2022 to constrain these
variabilities.
The inputs to the numerical model
identified that a rupture could occur Figure 2. Modelling Scenario for Case 2
within the uncertainty zone shown in
Baldwin et al. (2022) also identify the
SiCF strikes at azimuth 315° (±15°) with a dip that could vary ±20º (from vertical). The PFDHA recommended
a displacement of 8 inches and a shear width of 1 foot. For numerical modeling, the three fault scenarios
summarized in Table 1 were studied. Figure 2 presents the actual fault modelling scenario for Case 2: the fault
strike is at 330°, the angle between the strike and tunnel alignment at fault crossing is 89.3°, and the dip of the
fault is 90°. However, it is also noted that that there is insufficient data to determine whether potential future
faulting would occur on a single strand or across multiple fault strands.
It is also postulated that SiCF is potentially Holocene active. It is however noted that the fundamental soil
characteristics and material mineralogy at the site does not seem to support the presence of softened material
(post-yielding) in Holocene deposits. The trend of fault offsets inferred in Baldwin et al. 2022 interpretation
suggest the presence of flower-shaped pattern with the offset magnitude decreasing with depth.
Fault Tunnel Interaction

The fault-tunnel interaction analysis is developed in two phases:

1) Development of a free-field model to simulate permanent fault displacement offset without the
presence of the tunnel.
2) Introduction of segmental tunnel and assessment of tunnel response to the offset imposed by the ground
deformations from the fault rupture event.
This paper focuses on the results of Step 1 and the various modelling approaches adopted.
Table 1: Possible fault rupture scenarios (from Baldwin et al. 2022)
Scenario No. Fault Strike Angle between Fault Dip H:V Slip Ratio
Azimuth Fault Strike Angle
and Tunnel
Alignment

1 300° 119.3° 90° 8inches:2inches

2 330° 89.3° 90° 8inches:2inches
3 330° 89.3° 70° 8inches:2inches
 Free Field Model
To model the uncertainty in the fault displacement,
three different scenarios were developed based on
Table 2 as shown in Figure 3. It was seen that for
Scenario 1 and Scenario 3 applied external loading is
greater than the fault offset displacements
recommended. Scenario 2 was the most realistic
scenario. Depending upon the material behaviour in
the shear zone, there could be there could be
significant difference in the displacement gradient
across the 1ft shear zone width which could have an
impact on the lining design.
Quasi-static movement of ground block (permanent
offset displacement after fault rupture) is achieved by
prescribing deformations along the vertical lateral
boundaries and at the base of the ‘moving ground
block and Ground and shear zone modelled with
elasto-plastic soil model with depth dependent
stiffness.
Figure 3: Fault displacement modelling scenarios

Table 2: Modelling approach for different scenarios

Scenario Description Remarks
1 Free Field 3D model with ground material For the applied fault offsets at model boundaries
properties assumed to be the same on (4H:1V) and without the assumption of 1ft ‘soft’ shear
either side of the fault rupture plane zone in Holocene deposits, the relative displacements
Modelled depth of Holocene deposits. across the fault rupture plane are likely to be lower
No preferential shear zone than (8in:2in)

2 Same as Scenario 1 but with For the applied fault offsets at model boundaries
‘softer’ material properties in the 1ft (4H:1V) and WITH the assumption of 1ft ‘soft’ shear
shear zone (over full model depth) zone in Holocene deposits, the relative knife-edge
to simulate knife-edge displacements displacements across the shear zone could be achieved
Modelled depth of Holocene deposits
with preferential shear zone
3 Free field 3D model with no shear zone in For the applied fault offsets at the model boundaries
Holocene deposits but with 1ft shear zone (4H:1V) and with the assumption of 1ft ‘soft’ shear
in Pleistocene deposits. Soil depth greater zone in Pleistocene deposits, the deformation pattern
than in Scenarios 1 & 2. tend to evolve from knife-edge to a more gradual
Modelled depth of Holocene deposits + distribution with decreasing depth below ground
a finite depth of Pleistocene deposits 70ft). surface (i.e. going from Pleistocene to Holocene
deposits)

The 3D, free field, permanent fault displacement developed for the simulation is based on Case 2 ( Scenario
2) and represented in LS DYNA as shown in Figure 4 with a ‘moving block’ and a ‘fixed block’ (see Figure
4). The 3D model is 1000ft long and 354 ft wide and has a depth of 178 ft to represent the soil profile at the
fault intersection of the tunnel. The 3D FE model developed consisted of 647,960 solid elements and 7,000
shell elements for the tunnel lining
The strata identified in the fault uncertainty zone are generally granular with fine-grained interbeds, cross-cuts,
and interlayering; however, in the middle of the uncertainty zone, the strata identified in geotechnical
investigations were majority fine-grained, exhibiting clay-like behavior. The soil profile taken for this fault-
tunnel interaction analysis uses the clay-like soil profile with a 20-25 ft thick sand layer below the tunnel invert.
The moving block and fixed block are modeled with an elasto-plastic soil model. The 1ft shear zone is
modelled with soil stiffness assumed as 1/1000 of the in-situ material to concentrate the relative motion of the
blocks within the shear zone only.
After the establishment of geo-static stresses
in the soil and the model reaching
equilibrium, fault displacement offsets of
8inches (horizontal) and 2inches (vertical)
are applied at the boundary nodes of the
moving block. The prescribed displacements
are slowly increased from zero to their final
value (8in and 2in depending on the
direction) to keep the loading process as
quasi-static. Figure 4 shows the horizontal
deformation in the 3D free field model. The
ground deformations are relatively uniform
within the ‘moving block’ and the ground
deformations are largely concentrated
around the shear zone. Subsequently to
assess the effect of offset imposed on the
tunnel by the ground deformations resulting Figure 4. Horizontal Deformation Contours in the Moving
from the fault rupture event, segmental Block (Blue) and Fixed Block (Pink) Showing Ground
tunnel lining was introduced into the free- Movement Along the Fault Rupture Plane (50x magnified)
field model and the deformations around the
shear zone was assessed to understand
possible mitigation solutions.
Conclusions

This paper summarizes the approach developed to numerically model the permanent offset due to fault rupture
hazard of the Silver Creek fault, San Jose, California. Fault rupture studies showed that there was insufficient
data to determine whether potential future faulting would occur on a single strand or across multiple fault
strands and thus knife-edge style ground displacement discontinuity was recommended for design. The 3D FE
model developed consisted of 647,960 solid elements and showed that a potential knife edge displacement
could only develop by considering a thin shear zone in Holocene deposits. Depending upon the material
behaviour in the shear zone, there could be there could be significant difference in the displacement gradient
across the 1ft shear zone width which could have an impact on the lining design.
 References
1. Livermore Software Technology Corporation (LSTC). 2020a. LS-DYNA Keyword User's Manual, Volume 1.
https://www.dynasupport.com/manuals/ls-dyna-manuals/ls-dyna_manual_volume_i_r12.pdf. Retrieved 2021-04-06.
2. Baldwin J, Givler, R, Bloszies C., Lettis W, Johnson Zeiger K, Miller R and Blake W (Published 2022), Fault Rupture
Hazard Investigation of the Silver Creek fault, San Jose, California,12th National Conference on Earthquake Engineering,
3. Chaussard, E., Bürgmann, R., Fattahi, H., Nadeau, R., Taira, T., Johnson, W., & Johanson, I. (2015). Potential for Larger
Earthquakes in the East San Francisco Bay Area due to the Direct Connection between the Hayward and Calaveras Faults.
Geophys. Res. Lett., 2734-2741.
4. Donikian, R., Fowler, M., Cook, R. & Fong, M.(2020). Numerical modeling of special linings of tunnels crossing active seismic
faults. Proceedings of 17th World Conference on Earthquake Engineering, Sendai, Japan – September 13th-18th 2020.
5. Langenheim, V., Jachens, R., Wentworth, C., Graymer, R., Stanley, R., Mclaughlin, R., Ponce, D. (2015). A Summary of the
Late Cenozoic Stratigraphic. California Geosphere, 50-62.
6. Lettis Consultant International. (2020). Silver Creek Fault Investigation Report. San José, California : Valley Transportation
Authority VTA BSV Phase II Project.
7. NORCAL Geophysical Consultants (NORCAL). (2020). High-Resolution Seismic Reflection Survey Report, North Silver
Creek Fault, Santa Clara Valley, California, prepared for WSP USA, San Francisco, California. 28.
8. Hashash Y. M. A., Park D. and Yao J.I.C.. 2005. Ovaling deformation of circular tunnels under seismic loading, an update on
seismic design and analysis of underground structures. Tunnelling and Underground Space Technology, 20(4): 435-441.
9. Schwartz, D., Lienkaemper, J., Hecker, S., Kelson, K., Fumal, K., Baldwin, J.,Niemi, T. (2014). The Earthquake Cycle in the
San Francisco Bay Region: AD 1600-2010. Bulletin of Seismological Society of America, 1299-1328.
10. Wentworth, C. (2018). Surface Trace of Silver Creek fault. 2015-5017.
11. Wentworth, C., Jachens, R., Williams, R., Tinsley, J., & Hanson, R. (2015). Physical Subdivision and Description of the Water-
bearing Sediments of the Santa Clara Valley. California: U.S. Geological Survey Scientific Investigations Report, 2015-5017.
12. Wentworth, C., Williams, R., Jachens, R., Graymer, R., & Stephenson, W. (2010). The Quaternary Silver Creek Fault beneath
the Santa Clara Valley. U.S. Geological Survey Open-File Report.
13. Working Group on California Earthquake Probabilities (WGCEP). (2003). Earthquake Probabilities in the San Francisco Bay
Region. San Francisco: U.S. Geological Survey Open-File Report.
14. Working Group on California Earthquake Probabilities (WGCEP). (2008). The Uniform California Earthquake Rupture
Forecast, Version 2 (UCERF 2). USGS Open File Report 2007-1437, CGS Special Report 203.
15. Young,D., Dean,A.,Warren,S, Haldin,A. & Gregor,T. (2007). Preliminary seismic design considerations for a highway tunnel
in Coronado, California.pg 405-416
16. W.L. Wang, T.T. Wang, J.J. Su, et al. 2001. Assessment of damage in mountain tunnels due to the Taiwan Chi-Chi earthquake
[J]. Tunnelling and Underground Space Technology, 16(1): 133-150.