This is a repository copy of Influence of corners in excavations on damage assessment.

White Rose Research Online URL for this paper:

http://eprints.whiterose.ac.uk/133352/

Version: Accepted Version

Article:
Fuentes, R orcid.org/0000-0001-8617-7381 (Accepted: 2018) Influence of corners in
excavations on damage assessment. Geotechnical Research. ISSN 2052-6156 (In Press)

Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.

Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.

eprints@whiterose.ac.uk
https://eprints.whiterose.ac.uk/
 1

1 Influence of corners in excavations on damage assessment

2 Fuentes, R.
3 School of Civil Engineering, University of Leeds
4
5 Abstract
6 This paper provides guidance to quantify the extent of corner effects in excavations and their
7 impact on damage assessment. The corner effects extent is of great importance to make early
8 decisions during project planning and preliminary design, especially in relation to stakeholder
9 engagement and placement of instruments. By using empirical relations we are able to provide
10 an equation, validated against the literature and additional numerical models, to estimate the
11 extent of corner effects for a particular excavation geometry. Furthermore, two more equations
12 to quantify damage of excavations to adjacent structures is presented and validated against
13 two case studies in the literature. The proposed equations are also useful in the context of
14 early stages of project development. Finally, a simple study shows the different effect of
15 corners in sections parallel and perpendicular to a retaining wall. This highlights that corner
16 effects may actually induce additional damage due to the introduction of a movement gradient,
17 as opposed to common previous perception that assumed they were always conservative as
18 they reduced absolute movements.

19 Keywords: Excavation, Retaining Walls, Geotechnical Engineering

20
21 1. Introduction

22 Corners affect the distribution of ground movements behind an excavation, as demonstrated

23 extensively in the literature, e.g. (Finno et al. 2007; Finno and Roboski 2005; Fuentes and
24 Devriendt 2010; Hong et al. 2015; Hsiung et al. 2018; Lin et al. 2003; Moormann and
25 Katzenbach 2002; Ou et al. 2000; Roboski and Finno 2006; St John et al. 2005; Tan et al.
26 2014). However, less work has been published on their effects on damage assessment on its
27 surroundings (Finno et al. 2007; Finno and Roboski 2005; Roboski and Finno 2006) despite
28 this being a critical consideration for design and construction of excavations.

29 The most accurate way to approach damage assessment is using three-dimensional (3D)
30 sophisticated numerical models that can cope with all of the complexities of behaviour
31 surrounding an excavation. A two-dimensional approach (2D) can clearly not cope with corner
32 effects without additional empirical considerations. However, at early stages of projects,
33 especially during planning or outline design, the 3D approach may be both time-consuming
34 and expensive, and the 2D only an approximation.
 2

35 Therefore, a simple, economic and reliable approach to have an early estimation of the extent
36 and impact of corner effects on building damage assessment is of great value for practising
37 engineers.

38 This paper presents a short literature review of previous work where the impact of excavation
39 geometry was considered in relation to corner effects. Using validated published empirical
40 methods, a relationship is derived and compared against case studies to show the extent of
41 corner effects for different excavation geometries. This expression is critical to allow a final
42 presentation of a formulation that allows estimating building damage assessment very simply
43 that can be used by practising engineers. The paper finishes with a comparison of the inclusion
44 of corner effects against ignoring them in building damage assessment and highlights their
45 importance and relevance for structures with various position and orientation with respect to
46 the corners.

47 2. Effect of excavation geometry on corner effects

48 An interesting concept used in the literature to evaluate the effect of excavation geometry in
49 the presence of corner effects is the plane strain ratio (PSR), that was firstly introduced by (Ou
50 et al. 1996) based on numerical analysis. PSR is defined as 3D / 2D, or the ratio between the
51 maximum lateral movement of the wall, max, calculated using 3D simulations ( 3D) and 2D
52 simulations ( 2D). PSR can be used as a proxy for the presence of corner effects by assuming
53 that when PSR is lower than 1.0, corner effects must be present.

54 Fig. 1 shows the excavation geometry nomenclature that will be used throughout the paper,
55 where L is the length in plan of the long-side of an excavation, B is the length of short-side
56 and H is the retained height.

57 Using the PSR approach, (Moormann and Katzenbach 2002) found that excavations with a
58 L/H ratio greater than 4.0 exhibited a PSR value of 1.0: in other words, corner effects were not
59 evident at the centre of excavations. (Finno et al. 2007) carried our similar work and estimated
60 a value of L/H greater than 6.0 instead; they also showed that PSR reduces very rapidly when
61 L/H is lower than 2.0.

62 In terms of the effect of the width of the excavation, (Finno et al. 2007) showed that the effect
63 of width was minimal for values of L/B greater than 2.0 whilst (Lin et al. 2003) suggest values
64 of L/B greater than 3.5 and (Faheem et al. 2004) a value of 6.0. (Tan et al. 2014) analysed
65 multiple Metro stations in Shanghai and observed only small variations for a large range of
66 L/B values from ~2 to ~20. Other authors (Hsiung et al. 2018; Ou et al. 1996) have also shown
67 that for increasing values of L/B, the PSR value also increases generally. In general, although
68 they do not agree on a single limiting value of L/B, all indicate that for narrower excavations,
 3

69 the 2D wall displacement predictions in the longer wall are more accurate than for wider
70 excavations if the same retained height is considered.

71
72 Figure 1. Excavation nomenclature and graphical representation of empirical methods

73 In summary, the above contributions indicate that, as a lower bound, 2D calculations provide
74 similar results to 3D calculations in the estimations of max, for excavations where L/H is greater
75 than 4.0 to 6.0, and L/B is greater than 2.0 to 3.5. For excavations with lower L/H and L/B
76 ratios the presence of corner effects is therefore assumed. Although all contributions seem to
77 agree in that corner effects occur for all excavations, less agreement, and very little guidance,
78 emerges when considering their extent.

79 3. Extent of corner effects

80 The use of empirical relations allows calculating the extents of corner effects simply and when
81 combined with case studies provide confidence in their relationships. Two main methods are
82 used here: (Roboski 2004) and (Fuentes and Devriendt 2010) (F&D).

83 (Roboski 2004) introduced an ingenious use of the complimentary error function (erfc) to
84 model the ground movements (both horizontal and vertical) behind an excavation in a section
 4

85 parallel and along the wall (see Fig. 1). For the remainder of this paper, the author uses
86 movements meaning both horizontal and vertical ground movements, unless specified,
87 because the empirical methods used here were validated for both. The method of (Roboski
88 2004) was calibrated against different case studies in different ground conditions, and
89 provided satisfactory results for both horizontal and vertical ground movements. This method
90 is fully summarised and further validated by comparison to an excavation in Chicago, USA, in
91 (Roboski and Finno 2006).

92 Eq. 1 shows the horizontal movements behind an excavation wall (Finno and Roboski 2005).

93 Eq. 1

94 where x is the distance from the corner along the wall in plan (see Fig. 1).

95 From Eq. 1, the ratio of movement at any point along the wall is

96

97 Eq. 2

98

99 from which it follows that at the corner where x = 0,

100

101 Eq. 3

102

103 which is only a function of the excavation geometry expressed as the ratio L/H.

104 (Finno and Roboski 2005) through the use of the proposed function shown in Eq. 1, inherently
105 admit that a constant lateral movement of the wall (defined herein as plane strain) occurs in
106 all cases. This is due to the function’s shape that is used to approximate the movements.

107 Eq. 2 allows calculating movements for each ratio of L/H, and therefore, the distance to plane-
108 strain movements can be observed. In order to have a consistent framework for comparison
109 to cases from the literature, the author defines the distance to plane strain movements, x =
110 dmax, as the distance from the corner of the excavation to the point where 90% of the maximum
111 movement behind the wall occurs (see Fig. 1). The value of 90% was selected somewhat
112 arbitrarily to allow for a 10% variation around the maximum value. However, checks were
113 made and taking another percentage level between 80% and 90% changes the figure only
 5

114 slightly and does not affect the process and results. It is important to note that a value between
115 90% and 100% may produce different results although it would be less conservative. Since a
116 value of 90% is used, this is not strictly plane strain, although the author uses plane strain for
117 simplicity. Applying the above definition, i.e. corner / max = 0.9, Eq. 2 can be rewritten as

118

119 Eq. 4

120

121 From Eq. 4, it can be seen that the ratio dmax/L is independent of the absolute values of L, H
122 and max (i.e. the value of dmax/L is unique for a given ratio of L/H). The same equation allows
123 representing the relationship between dmax/L and L/H with a logarithmic expression shown in
124 Eq. 5, which is also shown as a dashed line in Fig. 2 for simplicity of use. For validation
125 purposes, another logarithmic line was fitted to the observed points marked in Fig. 2,
126 independently of Eq. 4, and gave close results with coefficients 0.0581 and 0.1341 for Eq. 5.

127

128 Eq. 5

129

130 In order to investigate the adequacy of Eq. 5, the case studies presented in Table 1 were
131 observed for different values of L/H and plotted in Fig. 2, where FM stands for field
132 measurements and FE for finite elements respectively, referring to the origins of the
133 observations. The 12 FM case studies available from the literature in Table 1 were not
134 sufficient to draw robust conclusions. Hence, additional case studies were created using a
135 finite element approach, the details of which can be found in Appendix A.
 6

0.5

0.45

0.4

0.35

0.3
dmax / L

0.25

0.2

0.15
FM
0.1 FE
0.05 Proposed equation

0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5 7 7.5 8 8.5 9 9.5 10
L/H
136
137 Figure 2. Position of plane strain movements for case studies and derived line

138 The proposed line constitutes a lower bound of the 26 case studies and was calculated by
139 using the marked points for fitting. Fig. 2 shows a significant scatter of points, with values of
140 dmax/L ranging between 0.15 and 0.44. This means that between 30% and 88% of the wall
141 length in plan is affected by corners. It also shows that for all the different conditions covered
142 in the case studies, plane-strain conditions occur (i.e. dmax / L < 0.5) confirming what was
143 shown by other authors using the PSR method and the empirical methods of (Roboski 2004)
144 and (Fuentes and Devriendt 2010). Eq. 5 could be hence used as a conservative estimate of
145 the position of plane strain movements, dmax, in an excavation.

146 Having the value of dmax is of double practical relevance to understand the extent of corner
147 effects, because it allows calculating indicative levels of damage assessment at a very early
148 stage of projects, and also deciding where to install monitoring stations to capture this
149 phenomenon.

150 3. Building damage assessment – results and discussion

151 Traditionally methods to establish damage assessment have been based on the theories
152 presented by (Burland et al. 1977) and (Boscardin and Cording 1989). These frameworks
153 have been revised over time by many, such as (Burland 2001), (Cording et al. 2001) and (Son
154 and Cording 2005), but the variables used for the different assessments remain the same as
155 in Fig. 3. The (Burland et al. 1977) method combines deflection ratio and horizontal strain to
 7

156 produce charts that provide guidance on the level of damage a building is undergoing.
157 (Boscardin and Cording 1989) arrive at similar results using angular distortion and horizontal
158 strains instead.

1 2 3 4 1 2 3 4

2v max 3v 4v 14 4v
1v
12
34

L 23 L

Maximum deflection Slope / rotation

max = max ( 1 , 2 , 3 , 4)
= V / L (i.e. 12 ≈ ( 1V – 2V)/L12

max = max( 12 , 23 , 34) sometimes

Deflection ratio taken as
max / L = 14

159

1 2 3 4 1 2 3 4

4H 34 4v
1v 12 23
1H

L L
δ’ Tilt
= H / L (i.e. 12 ≈( 1H – 2H)/L12
L= 1H – 4H = L – δ’ = max( , , )
max 12 23 34

Lateral / horizontal strain

Angular distortion
hor = L / L
= –

160
161 Figure 3. Typical damage assessment parameters definition

162
 8

163 As shown above, the main focus has been on the combination of horizontal strain and, either
164 angular distortion or deflection ratio. This section of the paper focuses on the deflection ratio
165 and angular distortion as they are immediate results of both empirical methods’ formulations.

166 (Roboski and Finno 2006) proposed Eq. 6 to estimate the maximum slope / rotation (as
167 defined in Fig. 3) parallel to the wall. It must be noted that Roboski and Finno used the term
168 distortion as opposed to Slope / rotation, and in doing so, were assuming that tilt was zero.

169

170 max Eq. 6

171

172 A similar equation, Eq. 7, can be derived for the method proposed by (Fuentes and Devriendt
173 2010), when combining it with the derived dmax in Eq. 5. Due to the method’s assumption,
174 rotation, angular distortion and deflection ratio are all equal for zero tilt (see Fig. 3).

175

176 max Eq. 7

177

178 The value of 0.33 in Eq. 7 comes from the assumption that corner / max = 0.67 as recommended
179 by (Fuentes and Devriendt 2010). However, the use of another ratio, i.e. using Eq. 3 or another
180 calibrated value, allows adapting this to other conditions as will be shown later.

181 Both Eq. 6 and 7 provide an estimate of maximum slopes in sections parallel to the retaining
182 wall for both horizontal and vertical movements. Both also depend only on the geometrical
183 parameters of the excavation and the maximum movement; close observation shows that, in
184 fact, both are independent of the absolute values of L and H. This allows a simple and quick
185 comparison of various geometries as shown in Fig. 4.
 9

2000 0.1 - Rob 0.1 -F&D 67% 0.2 - Rob

0.2 - F&D 67% 0.3 - Rob 0.3 - F&D 67%
0.5 - Rob 0.5 - F&D 67% CHI observed
1750 CHI Rob CHI F&D 67% CHI - F&D 33%
RLMRB w obs RMLRB w Rob RMLRB w F&D 67%
RLMRB w F&D 33% RMLRB s obs RMLRB s Rob
1500 RMLRB s F&D 67% RMLRB s F&D 33% RMLRB n obs
RMLRB n Rob RMRLB n F&D 67% RMLRB n F&D 33%
1250
Slope (1 in)

1000
Very Slight Damage

750
Slight Damage

500

250 Moderate Damage

Severe to V. Severe Damage

0
0 1 2 3 4 5 6 7 8 9 10
L/H
186
187 Figure 4. Slope vs L/H estimation using both methods compared to two case studies (please
188 refer to Table 1 for details of case studies shown). Key: Obs stands for observed and Rob refers
189 to Roboski’s method.

190 Fig. 4 shows the results from both methods for different excavation geometries, L/H. Both
191 methods differ in their predictions when a value of corner / max equal to 67% is used in the F&D
192 method, although the differences are reduced significantly when this value is taken as 33%.
193 The results from two sites are also included for validation purposes (see Table 1 for details).
194 In the CHI site, a maximum slope in the direction parallel to the wall of 1 in 690 was observed
195 for a basement with a L/H ratio of 3.87 and max/H of 0.33%, which agrees better with the
196 prediction of the F&D method, 1 in 713, than by Roboski, 1 in 347, (underestimated by almost
197 50%). Conversely, the data presented for the RMLRB excavation shows the opposite trend.
198 Here Roboski’s method provides a much better fit to the measurements than that presented
199 by Fuentes & Devriendt for a value of corner / max equal to 67%, although the match is similar
200 when this value is chosen as 33%.

201 Fig. 5 shows a similar graph for increasing maximum movement behind the wall. As expected,
202 the slope reduces with increasing movement. It indicates that for values of max/H (%) greater
203 than 0.3 and any values of L/H up to 10.0, the values of slope are normally close to 1 in 1,000
204 which would traditionally constitute a warning level for many structures as incipient ‘very slight
205 damage’, e.g. as presented by (Boscardin and Cording 1989). The dashed lines in Figs. 4 and
 10

206 5 show the damages thresholds for angular distortion for a lateral strain equal to zero as
207 recommended by (Boscardin and Cording 1989).

2000 L/H=10 / Roboski

L/H=10 / F&D 67%
1750 L/H=10 / F&D 33%
L/H=5.0 / Roboski
L/H=5.0 / F&D 67%
1500 L/H=5.0 / F&D 33%
L/H=2.0 / Roboski
1250 L/H=2.0 / F&D 67%
Slope (1 in)

L/H=2.0 / F&D 33%

1000 RLMRB
CHI
Very Slight Damage
750
Slight Damage
500

250 Moderate Damage

Severe to V. Severe Damage

0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
max/H(%)
208
209 Figure 5. Slope vs max /H estimation using both methods compared to two case studies

210 For lower values of L/H, close to 2.0, the slope reaches concerning levels for lower values of
211 max/H (%). This means that for excavations with shorter walls (lower values of L/H), more
212 potential for damage is predicted for the same retained height and maximum movement. This
213 is because the movements need to accommodate a similar total change over a shorter
214 distance. The RLMRB values seem to fall within the range of values calculated by both
215 methods for L/H = 5 when using Roboski and F&D with 33% value of corner / max. A similar
216 agreement is observed for the case study CHI which has a value of L/H equal to 3.87.

217 The above confirms that both methods provide similar values of building damage but that
218 however, a value of corner / max = 0.33, seems to provide more accurate results despite the
219 value of 0.67 that was recommended by the authors in (Fuentes and Devriendt 2010).

220 The relevance of being able to use the F&D method with confidence is that it allows calculating
221 damage assessment also in the shaded area of Fig. 1 which has not been demonstrated for
222 other empirical methods. Hence, the remainder of this section covers additional analyses that
223 was carried out using the F&D method only.

224 These extra analysis cases are shown in Fig. 6 and were done as building damage
225 assessment. The figure shows an excavation of dimensions 100m x 50m and 15m deep and
 11

226 13 simulated buildings for which damage assessment was calculated using two options,
227 considering corner effects and not considering them.

228 Parallel to the wall, buildings were placed along the long side of the wall covering the full length
229 and the short length, and straddling the corner. Perpendicular to the wall, buildings were
230 placed at the centre and at the corner of excavations on both sides. A final building was placed
231 in the corner of the excavation at 45º angle. It must be noted that the problem is fully
232 symmetrical.

233
234 Figure 6. Excavation details and position of analysed structures.

235 The analyses were run in the software Oasys XDISP 19.4 (“Oasys XDISP” 2016) where the
236 F&D method is implemented. A value of corner / max = 0.33 was used and dmax, necessary for
237 the calculations, was estimated using Eq. 5, giving a value of 8m and 3.64m for the long and
238 short sides respectively. The building damage assessment criteria follows the limits
239 established by (Burland 2001): negligible(N), very slight (VS), slight (S), moderate (M) and
240 severe (SE), with transition values of horizontal ground strain between zones of 5E-4, 7.5E-4,
241 1.5E-3 and 3E-3.
 12

242 Two ground deformation profiles were used for input as shown in Fig.7: one with a large
243 movement using the Taipei National centre (Ou et al. 2000) and another with smaller
244 movements using the House of Commons site (HOC) (Burland and Hancock 1977). These
245 deformation profiles are necessary for the F&D method and describe the deformation of the
246 ground in a cross section perpendicular to the wall at the centre as shown in (Fuentes and
247 Devriendt 2010).

y/H
0 1 2 3 4
0
0.05
0.1
H(%)

0.15
0.2
max /

0.25
0.3 House of Commons

0.35 Taipei

248
0.4
249 Figure 6. Ground movement profiles (horizontal and vertical)

250 Table 2 summarises the interesting results. For small deformations, i.e. HOC where max /
251 H(%) = 0.1, the consideration of corner effects reduces the level of damage around the corners
252 in sections perpendicular to the wall. In all other cases, the damage is negligible due to the
253 small movements. This shows the potential savings of including corners for structures around
254 corners.

255 For larger deformations, i.e. Taipei where max / H(%) = 0.35, the behaviour is more complex.
256 For sections parallel to the wall, the inclusion of corner effects increases the level of damage
257 for almost all structures and therefore, is not conservative. In the perpendicular direction the
258 differences are also significant near the corners. When corner effects are not considered, the
259 damage level is severe for all sections, except the bisector section which is moderate.
260 Conversely, including corner effects reduces the level of damage at the corner and the bisector
261 significantly.
 13

262 Table 2. Building damage assessment results

Taipei House of Commons

With Without With Without

LF-2.5m N N N N

LF-5m S N N N

SF-2.5m S N N N

SF-5m S N N N

LS-2.5m S M N VS

LS-5m S M N N

SS-2.5m S VS N N

SS-5m S VS N N

L-CENTRE SE SE VS VS

L-CORNER S SE N VS

S-CENTRE SE SE VS VS

S-CORNER S SE N VS

CORNER S M N N

263

264 The above shows the importance of considering the corner effects but highlights how both
265 sections parallel and perpendicular to the wall must be considered at the same time and how
266 the behaviour may vary for different magnitude of ground movements.

267 4. Conclusions

268 All excavations presented in the literature and studied in this paper present corner effects as
269 defined here. However, for excavations where L/H is greater than 4.0 to 6.0, and L/B is greater
270 than 2.0 to 3.5, these can be neglected in terms of the estimation of the maximum movement,
271 although this does not mean corner effects do not exist.

272 The extent of corner effects, expressed as the distance from the corner, is important to make
273 early decisions of the potential impact of an excavation on surrounding buildings, as well as
274 helping to decide the location of instrumentation. A conservative relationship validated against
275 26 case studies has been provided.
 14

276 The above relationship also allows the estimation of maximum slope, as indicator of potential
277 damage, which can be estimated easily using two equations. One requires an estimate of the
278 maximum movement and the other requires the ratio between corner and centre movements
279 to be used.

280 Furthermore, the inclusion of corner effects in damage calculations can have very different
281 implications whether considering sections parallel or perpendicular to the wall in plan. For
282 large ground movements max / H(%) = 0.35, the damage increases in sections parallel to the
283 wall but reduce in areas near the corner for perpendicular sections when comparing it to the
284 case of not considering corner effects. In the cases of small movements, max / H(%) = 0.1,
285 the damage is lower when including corner effects for both directions. This highlights the
286 importance of considering corner effects, and also, both directions.

287 References

288 Altair. (2009). “Hypermesh.”

289 Boscardin, ε. D., and Cording, E. J. (1989). “Building Response To Excavation-Induced

290 Settlement.” Journal of Geotechnical Engineering, 115(1), 1–21.

291 Burland, J. B. (2001). “3 Assessment methods used in design.” Building response to

292 tunnelling, 23–44.

293 Burland, J. B., Broms, B. B., and de εello, V. F. B. (1977). “Behaviour of foundations and
294 structures: State of art report.” Proceedings of the International Conference on Soil
295 Mechanics and Foundation Engineering, Tokyo, Japan. July 10-15, 1977, 495–536.

296 Burland, J. B., and Hancock, R. J. R. (1977). “Underground car park at the House of
297 Commons, δondon: Geotechnical aspects.” Structural Engineer.

298 Cording, E., Long, M., Son, J., and Laefer, D. (2001). “εodeling and Analysis of Excavation-
299 Induced Building Distortion and Damage Using a Strain-based Damage Criterion.”
300 Proceedings of the Conference on Response of Buildings to Tunnelling - Case Studies
301 from Construction of the Jubilee Line Extension, CIRIA.

302 Faheem, H., Cai, F., and Ugai, K. (2004). “Three-dimensional base stability of rectangular
303 excavations in soft soils using FEε.” Computers and Geotechnics, 31(2), 67–74.

304 Finno, R. J., Blackburn, J. T., and Roboski, J. F. (2007). “Three-Dimensional Effects for
305 Supported Excavations in Clay.” Journal of Geotechnical and Geoenvironmental
306 Engineering, 133(1), 30–36.

307 Finno, R. J., Bryson, S., and Calvello, ε. (2002). “Performance of a Stiff Support System in
308 Soft Clay.” Journal of Geotechnical and Geoenvironmental Engineering, 128(8), 660–
 15

309 671.

310 Finno, R. J., and Roboski, J. F. (2005). “Three-Dimensional Responses of a Tied-Back

311 Excavation through Clay.” Journal of Geotechnical and Geoenvironmental Engineering,
312 131(3), 273–282.

313 Fuentes, R., and Devriendt, ε. (2010). “Ground movements around corners of excavations:
314 Empirical calculation method.” Journal of Geotechnical and Geoenvironmental
315 Engineering, 136(10), 1414–1424.

316 Fuentes, R., Pillai, A., and Ferreira, P. (2018). “δessons learnt from a deep excavation for
317 future application of the observational method.” Journal of Rock Mechanics and
318 Geotechnical Engineering, Elsevier, 10(3), 468–485.

319 Grammatikopoulou, A., St John, H. D., and Potts, D. ε. (2008). “Non-linear and linear
320 models in design of retaining walls.” Proceedings of the ICE - Geotechnical
321 Engineering, 161(6), 311–323.

322 Hong, Y., Ng, C. W. W., δiu, G. B., and δiu, T. (2015). “Three-dimensional deformation
323 behaviour of a multi-propped excavation at a ‘greenfield’ site at Shanghai soft clay.”
324 Tunnelling and Underground Space Technology, 45, 249–259.

325 Hsiung, B.-C. B., Yang, K.-H., Aila, W., and Ge, δ. (2018). “Evaluation of the wall deflections
326 of a deep excavation in Central Jakarta using three-dimensional modeling.” Tunnelling
327 and Underground Space Technology, 72, 84–96.

328 Jen, δ. (1997). “The design and performance of deep excavations in clay.” εassachusetts
329 Institute of Technology.

330 St. John, H. D., Potts, D. ε., Jardine, R. J., and Higgins, K. G. (1993). “Prediction and
331 performance of ground response due to construction of a deep basement at 60 Victoria
332 Embankment.” Predictive soil mechanics. Proc. of the Wroth memorial symposium,
333 Oxford, 1992, 581–608.

334 δee, F. H., Yong, K. Y., Quan, K. C. N., and Chee, K. T. (1998). “Effect of corners in strutted
335 excavations: Field monitoring and case histories.” Journal of Geotechnical and
336 Geoenvironmental Engineering, 124(4), 339–349.

337 Lin, D. G. G., Chung, T. C., and Phien-Wej, N. (2003). “Quantitative evaluation of corner
338 effect on deformation behaviour of multi-strutted deep excavation in Bangkok subsoil.”
339 Journal of the Southeast Asian Geotechnical Society, 34(1), 41–57.

340 δong, ε. (2001). “A case history of a deep basement in δondon Clay.” Computers and
341 Geotechnics, 28(6–7), 397–423.
 16

342 LS-DYNA. (2008). “(Civil Engineering Application Program) Version 940.” Oasys δimited,
343 Livermore Software Technology Corporation.

344 εoormann, C., and Katzenbach, R. (2002). “Three dimensional effects of deep excavations
345 with rectangular shape.” International Conference on Soil Structure Interaction in Urban
346 Civil Engineering , Zürich.

347 “Oasys XDISP.” (2016). Oasys, Arup.

348 Ou, C.-Y., Chiou, D.-C., and Wu, T.-S. (1996). “Three-Dimensional finite element analysis of
349 deep excavations.” Journal of Geotechnical Engineering, 122(May), 337–345.

350 Ou, C.-Y., Shiau, B.-Y., and Wang, I.-W. (2000). “Three-dimensional deformation behavior of
351 the Taipei National Enterprise Center (TNEC) excavation case history.” Canadian
352 Geotechnical Journal, 37(2), 438–448.

353 Roboski, J. (2004). “Three-dimensional Performance and Analyses of Deep Excavations.”

354 Northwestern University.

355 Roboski, J., and Finno, R. J. (2006). “Distributions of ground movements parallel to deep
356 excavations in clay.” Canadian Geotechnical Journal, 43(1), 43–58.

357 Simpson, B. (1992). “Retaining structures: displacement and design.” Géotechnique,

358 Thomas Telford Ltd , 42(4), 541–576.

359 Son, ε., and Cording, E. J. (2005). “Estimation of building damage due to excavation-
360 induced ground movements.” Journal of Geotechnical and Geoenvironmental
361 Engineering, 131(2), 162–177.

362 St John, H. D., Zdravkovic, δ., and Potts, D. ε. (2005). “εodelling of a 3D excavation in
363 finite element analysis.” Géotechnique, 55(7), 497–513.

364 Tan, Y., Wei, B., Diao, Y., and Zhou, X. (2014). “Spatial corner effects of long and narrow
365 multipropped deep excavations in shanghai soft clay.” Journal of Performance of
366 Constructed Facilities, 28(4), 4014015.

367 Whittle, A. J., and Hashash, Y. ε. A. (1992). “Analysis of the behaviour of propped
368 diaphragm walls in a deep clay deposit.” Retaining Structures, Proceedings of the
369 Conference, Thomas Telford, Cmabridge, UK, 131–139.

370 Yeow, H., and Feltham, I. (2008). “Case histories back analyses for the application of the
371 Observational Method under Eurocodes for the SCOUT project.” 6th international
372 conference on case histories and …, 1–13.

373
 17

374 Appendix A - THREE-DIMENSIONAL FINITE ELEMENT MODELLING – ADDITIONAL

375 CASE STUDIES

376 A parametric study was undertaken using three-dimensional finite element model. The
377 parametric study consisted of different combinations of the L/H and L/B ratio. The choice of
378 these parameters is based on the findings of (Finno et al. 2007) and (Moormann and
379 Katzenbach 2002), who showed these to be the most relevant parameters.

380 A computer program called HYPERMESH v. 10.0 (Altair 2009) was used to create the mesh
381 and the topology of the problem, whilst LS-DYNA (LS-DYNA 2008) was used to undertake the
382 numerical analysis of the problem.

383 Problem geometry and finite element mesh dimensions

384 The retained height, H, was selected as 12m throughout the study, and the excavation had a
385 length, L, and a width, B. The L/H ratio chosen were: 2.5, 3.5, 4.5 and 7.0. This resulted in
386 lengths of the wall, L equal to 30m, 42m, 54m and 84m. The values of L/B varied between 1.0
387 and 2.0. The problem is symmetric about the two axes that pass through the centre of the
388 excavation, hence, only one quarter of the excavation was modelled, as shown in Fig. A.1.

389
390 Figure A.1. FE model geometry
391 The retaining wall dimensions were also fixed to a width of 1.0m and a depth of embedment
392 of 4.0m below final excavation level. The thickness of the wall was kept constant for all the
 18

393 different analyses, after the findings of (Moormann and Katzenbach 2002), who showed it has
394 little influence on the movements and the corners. The excavation depth of the wall was also
395 kept constant, following (St John et al. 2005) conclusions: “The effect of the embedment depth
396 of a wall on movements and structural forces in the excavations analysed is negligible”.
397 (Whittle and Hashash 1992) support this latter statement and concluded that the wall
398 embedment has a negligible effect in ground and wall movements, especially in situations of
399 soil pre-failure.

400 (Roboski 2004) recommended the lateral boundaries of the mesh to be placed at a distance
401 of 5H from the retaining wall. (Lin et al. 2003) showed that a distance of 3H to the boundary
402 was sufficient. The recommendations of (Roboski 2004) were adopted in this study to allow
403 for a greater distance behind the wall over which to obtain ground movements free of boundary
404 effects. The lower boundary of the problem was also located at a distance 5H from the bottom
405 of the excavation.

406 The vertical boundaries of the mesh were restrained in the horizontal direction and free to
407 move in the vertical direction. The bottom of the mesh was fixed in all directions.

408 The finite element model was built using hexahedral and tetrahedral elements. The remainder
409 elements accounted for less than 1% of the total number of elements.

410 Elements within the excavation were modelled with dimensions of 0.5m x 0.5m x 0.5m. The
411 rest of the elements outside this area varied in dimensions, with coarser elements appearing
412 with greater distance from the excavation as shown in Fig. A.1.

413 Soil profile, material models and groundwater

414 The soil profile shown in Table A.1 consisted of a layer of 5m of granular material overlying
415 stiff clay, similar to that taken by (Grammatikopoulou et al. 2008), which represents a typical
416 situation in the City of London. Underlying the clay, a 4m, very stiff, granular material layer
417 was used to provide a rigid boundary. Furthermore, (Jen 1997) showed in a parametric study
418 that the depth of overlying granular materials over an overconsolidated clay has a small
419 influence in the ground movements, especially when the top of the clay is above the
420 excavation level; (Yeow and Feltham 2008) confirmed this.

421 The granular materials were modelled using a linear elastic perfectly plastic Mohr-Coulomb
422 material. The clay was modelled using the non-linear small strain stiffness model BRICK
423 developed by (Simpson 1992). The validated ‘most probable’ parameters for BRICK presented
424 by (Fuentes et al. 2018) were used. The parameters used for the simulations are presented
425 in Table A.1.
 19

426 Table A.1. Soil stratigraphy and parameters.

Top of layer E’
Soil ' K0
(m bgl) (kN/m³) (GPa)

Granular soil 0 20 25 30 0.5

Clay 5 20 See (Fuentes et al. 2018)
Very stiff
granular 68 20 150 36 0.41
material
427
428 The groundwater table was placed 0.5 m below ground level (bgl) and followed a hydrostatic
429 profile to the bottom of the problem.

430 Construction sequence

431 The same construction sequence was followed for all the different simulations. The type of
432 analysis in terms of drainage is shown in brackets for each stage in the clay. The remaining
433 soils were granular materials and were always treated as drained.
434 1) Initialisation (drained)

435 2) Wish in place retaining wall (undrained)

436 3) Excavate to 1 mbgl. (undrained)

437 4) Install temporary prop at capping beam level and excavate to 5 mbgl. (undrained)

438 5) Install temporary prop at level 2 and excavate to formation level. (undrained)

439 6) Install base slab. (undrained)

440 7) Remove temporary prop at level 2 (for bottom-up construction sequence only)

441 (undrained).

442 Support systems modelling and properties

443 Different construction sequences, top-down and bottom-up were modelled by simply changing
444 the stiffness of the temporary props used in the model. For the top-down construction, the
445 stiffness of a concrete slab was used. For the bottom-up model the stiffness was calculated
446 using 660mm external diameter circular hollow sections placed at 10m intervals. This
447 arrangement is typical of London, as shown in (Fuentes et al. 2018).
 20

448 The weight of the prop elements was reduced to a very small value to avoid imposing
449 significant loading on the wall, which would be unrealistic. The base slab was however
450 modelled with full weight as its effect on the behaviour of the formation level is important.

451 A stiffness of 2.1E+07 kN/m2 (that of concrete in the short term) was used in the top-down
452 case. The top two levels and the slab were taken as 0.3m and 1.2m thick respectively.

453 In the bottom-up construction sequence, the stiffness of the tubular props mentioned above
454 was 4.2E+06 kN/m2 and a thickness of 0.4m.
455
 21

Table 1. Case studies details

Plan Vmax Hmax G/W

Retained Construction dmax H/V
Source dimensions approx approx
Case study FM height Corner Wall method /
Soil (m) dmax/L
(abbreviation) See note a) / FE LxB angle type Propping (mm) (mm)
H (m) system see note d)
(m x m) See note b) see note c)

New Palace (Burland and 12.5v (east)

Top down /
Yard Car Park Hancock FM London Clay 18.5 66 x 50 90° D-wall 13 (east) 303 0.25
Permanent slabs 16.5v (south)
(NPYCP) 1977)
19 (south)
35 x 35 22.5 7.0v 0.2
Moorhouse Bottom up /
(St John et v
FE London Clay 40 70 x 35 90° N/A Multi-prop (7no 31.0 18.9 0.27
(MOOR) al. 2005) N/A
prop levels)
140 x35 38.0 28.0v 0.2

Immigration Marine Clay Bottom up /

Building (Lee et al.
FM (soft, high 17.3 75.6 x 50.4 90° D-wall Multiple level 110 60W N/A N/A
1998)
(IMM) plasticity) props

Silty Clay (low

Taipei National 25S
(Ou et al. plasticity Top down /
Enterprise FM 19.7 107 x 45 90° D-wall 78 24V 0.225
2000) overconsolidat Permanent slabs 100W
Centre, (TNEC)
ed)

60 Victoria
Embankment (St. John et Alluvium and Secant Top down /
FE 19 67 x 29 90° 28 343,W N/A N/A
al. 1993) London Clay pile wall Permanent slabs
(60 VIC)

Top-down /
Basement 1 Arup FE 89.3 x 150 Horizontal
Secant
LSDYNA FE London Clay 14.26 90° temporary props 15.9 12.0 31 0.34
(BAS 1) Approx. pile wall
results in diagonal
arrangement

24.4 x 29.2 Bottom up /

Basement 2 Arup FE Multi-level 7.8 8.9 10.7H 0.44
(in one Secant
LSDYNA FE London Clay 15 75° temporary
(BAS 2) corner of the pile wall 9.3 11.4 H
0.39
results corner props and
excavation) horizontal prop
 22

Plan Vmax Hmax G/W

Retained Construction dmax H/V
Source dimensions approx approx
Case study FM height Corner Wall method /
Soil (m) dmax/L
(abbreviation) See note a) / FE LxB angle type Propping (mm) (mm)
H (m) system see note d)
(m x m) See note b) see note c)
Robert H. Lurie 80 (west - w) 83W (w)
Glacial 17H (w) 18H
Medical (Finno and 60 (south - s) 51W (s)
clays(soft to Sheet Bottom-up / Tie- (s) 32H (n) 0.25 (w) 0.23
Research Roboski FM 15 68 x 80 90
stiff varying pile backs (3 levels) 50 (north - n) 53W (n) (s) 0.39 (n)
Building 2005)
with depth)
(RLMRB)

Takagi Silt, peat, and

(Roboski
FM sandy and silty 28 30 x ? 90 N/A N/A 52.5 N/A 4.5V 0.15
(TAK) 2004)
layers

Sand overlying
Chicago State a thin stiff crust
Bottom-up /
Subway (Finno et al. over stiff clays Secant
FM 12.2 47.3 90 Tubular props 40 38 18V 0.38
2002) of increasing pile wall
(CHI) and tie-backs
stifness with
depth

Flagship Wharf Silt deposit,

Bottom-up / 3
Building (Roboski glacio-marine
FM 15.0 30 x 76 90 D-wall levels of tubular 28 N/A 10.5H 0.35
2004) deposit, and
(FWB) props
glacial till

Clay layered
(Moormann soil with sand
Frankfurt Clay and bands and Secant ? / Tie-back at 4
FM 15.8 34 x 54 90 175 175 17H 0.31
(FRA) Katzenbach rocky pile wall levels
2002) limestone
embedded
30 x30 2.89 3.80 10.0H 0.33
30 x 15 2.75 3.08 10.5H 0.35
Granular
material 15 x 30 0.67 1.10 4.5H 0.30
FE additional Shell
This paper FE overlying stiff 12 90 Top-down
case studies 42 x 42 element 4.14 5.6 13.5 H
0.32
overconsolidat
ed clay H
54 x 54 4.97 6.8 16.5 0.30
H
54 x 27 4.10 6.0 17 0.31
 23

Plan Vmax Hmax G/W

Retained Construction dmax H/V
Source dimensions approx approx
Case study FM height Corner Wall method /
Soil (m) dmax/L
(abbreviation) See note a) / FE LxB angle type Propping (mm) (mm)
H (m) system see note d)
(m x m) See note b) see note c)
27 x 54 2.90 3.7 8.5H 0.31
84 x 84 6.30 8.5 21.5H 0.25

Granular 30 x30 4.05 5.05 10.5 H 0.35

material
FE additional Shell H
This paper FE overlying stiff 12 30 x 15 90 Bottom-up 3.81 4.12 10.5 0.35
case studies element
overconsolidat H
ed clay 15 x 30 0.72 1.20 3.5 0.23
H
84 x 84 8.5 12.0 23.5 0.28

NOTES
a) Unless otherwise stated within the table and the notes below, all the information included in the table comes from the reference listed in the second column.
b) A number superscript numbers indicates where the information was taken from:
1 (Roboski 2004), 2 (Roboski and Finno 2006), 3 (Long 2001)
c) The superscripts W and S represent whether the value was derived from observation of the wall (W ), or the soil (S).