OSF Paper.pdf

This article was downloaded by: [Kang Fong Pong] On: 18 July 2012, At: 18:36 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

The IES Journal Part A: Civil & Structural Structural Engineering Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tiea20

Design considerations for one-strut failure according to TR26 – a practical approach for practising engineers a

a

a

a

K.F. Pong , S.L. Foo , C.G. Chinnaswamy , C.C.D. Ng & W.L. Chow

b

a

Geotechnical Engineering Department, Meinhardt Infrastructure Pte Ltd (Member of Meinhardt Group), Singapore

b

Formerly of Geotechnical Amberg and TTI Engineering Pte Ltd

Version of record first published: 18 Jul 2012

To cite this article: K.F. article: K.F. Pong, S.L. Foo, C.G. Chinnaswamy, C.C.D. Ng & W.L. Chow (2012): Design considerations for onestrut failure according to TR26 – a practical approach for practising engineers, The IES Journal Part A: Civil & Structural Engineering, 5:3, 166-180 To link to this article: http://dx http://dx.doi.org/10.1 .doi.org/10.1080/19373260.2012.700 080/19373260.2012.700790 790

PLEASE SCROLL DOWN FOR ARTICLE Full terms and conditions of use: http://www.tandfonline.com/page/terms-and-conditions This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, redistribution, reselling, loan, sub-licensing, systematic supply, supply, or distribution in i n any form to anyone is expressly forbidden. The publisher does not give any warranty express or implied or make any representation t hat the contents will be complete or accurate or up to date. The accuracy of any instructions, formulae, and drug doses should be independently verified with primary sources. The publisher shall not be liable for any loss, actions, claims, proceedings, demand, or costs or damages whatsoever or howsoever caused arising directly or indirectly in connection with or arising out of the use of this material.

The IES Journal Part A: Civil & Structural Engineering Vol. 5, No. 3, August 2012, 166–180

TECHNICAL PAPER Design considerations for one-strut failure according to TR26 – a practical approach for practising engineers K.F. Ponga*, S.L. Foo a, C.G. Chinnaswamya, C.C.D. Nga and W.L. Chowb a

Geotechnical Engineering Department, Meinhardt Infrastructure Pte Ltd (Member of Meinhardt Group), Singapore; bFormerly of Geotechnical Amberg and TTI Engineering Pte Ltd (Received 22 May 2012; final version received 13 June 2012 )

2 1 0 2 y l u J 8 1 6 3 : 8 1 t a ] g n o P g n o F g n a K [ y b d e d a o l n w o D

Technical Reference 26: 2010 (TR26: 2010) requires the design of an earth retaining and stabilising system (ERSS) to be structurally safe, robust and has sufficient redundancy to avoid catastrophic collapse of the ERSS system resulting from an isolated case of overloading or failure of any particular member which may lead to the failure of adjacent members thus leading to progressive failure. One such redundancy check is the condition, where failure of a single strut, anchor or tie-rod occurs or more commonly known as one-strut failure (OSF) stated in Clause 3.7.4 of TR26: 2010 at each stage of the construction works. Analysis for OSF is actually a three-dimensional (3D) problem and carrying out such 3D analyses covering all the cases of wall stiffness, properties of the soil layers, friction between retaining wall panels in the case of diaphragm wall, soil arching effect due to the deflection of the retaining wall, etc. is very time consuming. In the conventional approach for OSF using two-dimensional (2D) plane strain analysis, the whole layer of failing strut is removed and thus provides paths to distribute the forces in the vertical direction only. This usually leads to more conservative design with heavier struts sections. In this paper, a procedure to rationally idealise OSF from a 3D analysis to a 2D plane strain analysis is presented. This simplified approach will be more practical for practising engineers to arrive at a more efficient design without the need for rigorous 3D analysis. The results of this simplified approach are compared with the conventional approach and results incorporating appropriate strut stiffness from 3D analysis. The comparison showed that the approach is reasonable. Keywords: numerical modelling; deep excavation; practical approach; three-dimensional analysis; one-strut failure; redundancy check; catastrophic collapse; overloading; progressive failure; TR26

1.

Introduction

Due to scarcity of land aboveground especially in urban area, underground infrastructure projects such as underground rails, roads and utilities tunnels networks are increasingly built and are often in close proximity to existing structures and buildings. The construction of these underground structures inevitably requires the use of safe and robust earth retaining and stabilising system (ERSS) for the deep excavations to minimise ground movement and impact to the surrounding structures and buildings. Hence, the structural safety and robustness are the prime considerations and requirements of the ERSS design and construction. Furthermore, the ERSS should have sufficient redundancy to avoid any catastrophic collapse of the supporting system resulting from an isolated case of overloading or failure of any particular supporting element. In an ERSS system, overall failure as described by Puller (2003) is more likely to occur either as a result of inadequate strutting or passive soil failure if the key-in depth is inadequate. Sometimes, this inadequacy in the strutting system may occur due

*Corresponding author. Email: [email protected] ISSN 1937-3260 print/ISSN 1937-3279 online  2012 The Institution of Engineers, Singapore http://dx.doi.org/10.1080/19373260.2012.700790 http://www.tandfonline.com

to bad fitting or strut failure due to disproportionate loads. The load of the failed strut will be redistributed to the adjacent struts which are not designed for these additional loads, thus resulting in the progressive failure of the whole strutting system. Clause 3.7.4 of Technical Reference 26: 2010 (TR26: 2010) states that the design for deep excavations should accommodate possible failure of any individual strut, tie rod, ground anchor, structural member or connection at each stage of the construction works. The wall and remaining supporting members, including walings and connections, should be capable of carrying the load from the failed member. The remaining structural system and wall should continue to be safe without causing any danger to surrounding adjacent structures and properties. This requirement is commonly known in Singapore as one-strut failure (OSF). It is necessary to consider the requirement of OSF in the design of the ERSS for deep excavation. In practice, it may not be necessary to check all the combinations of OSF. For example, if there are six struts in a deep excavation, the total number of

The IES Journal Part A: Civil & Structural Engineering


possible cases of OSF is 21. It is very onerous to check all 21 cases. In reality, only six cases need to be analysed. The first case is OSF for strut at level 1 (S1) when the excavation reaches strut at level 2 (S2) but before installing S2. The second case is OSF for S2 when the excavation reaches strut at level 3 (S3) but before installing S3, etc. The sixth case is OSF for strut at level 6 (S6) when the excavation reaches the final excavation level. Analysis for OSF is actually a three-dimensional (3D) problem and carrying out such 3D analyses covering all the cases of wall stiffness, properties of the soil layers, friction between retaining wall panels in the case of diaphragm wall, soil arching effect due to the deflection of the retaining wall etc. is very time consuming. In the conventional design approach for OSF using two-dimensional (2D) plane strain analysis, the whole layer of failing strut is removed and thus provides paths to distribute the forces in the vertical direction only. This usually leads to more conservative design with heavier struts sections. In this paper, a procedure to rationally idealise the OSF case from a 3D analysis to a 2D plane strain analysis is presented. This approach will be more practical for practising engineers to arrive at a more efficient design without the need for rigorous 3D analysis. The results of this approach are compared with the conventional approach and results incorporating appropriate strut stiffness from 3D analysis of OSF.

Figure 1.

2.

167

Modelling of strutting system

Figure 1 shows a typical layout of the strutting system for an ERSS commonly used in deep excavation projects. The soil and water pressures acting on the retaining wall are partially transferred to the struts through the walers and partially to the soil support below the excavation level. In this force transfer mechanism from soil to strut, the waler acts as a load distributing member and the strut acts as compression member to balance the soil and water pressures from both sides of the ERSS to maintain the force equilibrium in the system and thus stabilises the ERSS. In order to carry out the plane strain analysis, the equivalent stiffness of the strut supporting system is needed in the 2D numerical analysis. This equivalent stiffness can be obtained as two springs in series as shown in Figure 2 and can be derived as follows: When two springs are serially connected, the total spring displacement due a given load is the sum of the individual spring deflections while subjected to the same loading. In other words, deqvt ¼ d w þ d s

ð1Þ

where d represents the deflections and the suffixes eqvt,w and s represent equivalent, waler and struts, respectively.

Strut layout of a typical ERSS for deep excavation projects.

K.F. Pong et al.

168

Figure 2.


Simplified model of strut-waler support system for retaining wall in deep excavation.

Re-writing the spring displacements in terms of spring stiffness, the following equation in terms of spring stiffness is obtained 1 1 1 ¼ þ K eqvt K w K s

ð2Þ

where K represents the stiffness. Generally, in normal case, walers are supported by struts and splays which leads to a very short span of 2– 2.5 m for the walers and thus the term K w becomes too large and therefore, the first term on the right hand side of Equation (2) becomes negligible. Therefore, soil and water pressures on the retaining wall are partly transferred to the struts through the walers and the rest to the soil support below the excavation level, Equation (2) can be simplified as: In normal case, K eqvt ¼ K s :

ð3aÞ

In OSF case, K eqvt ¼ 3.

K s  K w : ðK s þ K w Þ

ð3bÞ

Modelling with one-strut failure (OSF)

The requirement of design for deep excavation to accommodate possible failure of any individual strut, tie rod, ground anchor is stated in Clause 3.7.4 of TR26. A similar clause is stated in BS 8002 (1994) as well. For projects of Singapore Land Transport Authority (LTA), consideration of OSF is a design requirement stated in the LTA Civil Design Criteria (2010). One-strut failure (OSF) is a 3D problem and it will not be an easy task to model the problem using 3D modelling every time a design of ERSS is carried out. Therefore, it is necessary to idealise OSF in 2D analysis to obtain a solution that is comparable to the 3D

modelling approach. The load combination factors for limit states design of the structural elements are shown in Table 3 of Clause 3.10.2 of TR26. These requirements of design of temporary supports accommodating possible OSF cases can be complied with using the following two approaches: (1) The first approach is to consider removal of one row of struts in the geotechnical modelling using computer software (such as PLAXIS). In this case, the contribution of struts in the same layer as the failed strut in the horizontal direction is totally ignored and the force from the failed strut can only be distributed vertically to the adjacent layers of struts. This conventional approach may lead to a more conservative design with higher reinforcement for walls and heavier sections for the struts. (2) The second approach is to consider failure of a single strut within a layer of struts in the horizontal direction. The force of the failed strut will be transferred both vertically and horizontally. However, the waler at the level of the failed strut has to be designed to transfer the load horizontally. As such the span of the waler for design would be increased compared with the normal span. This is a more realistic approach. A case study based on the second approach using the analysis of a cross-over cut-and-cover tunnel from the Downtown Line Stage 3 (DTL 3) Project was carried out. The objective was to assess that the second approach is not too conservative for both the wall and the strutting system but yet remain robust. Twospecific aspects were investigated: (1) Reduced stiffness of the failed strut to be adopted in the geotechnical finite-element modelling.

The IES Journal Part A: Civil & Structural Engineering (2) Waler span for the design of waler due to OSF.

4.

Methodology

Both structural and geotechnical modelling were carried out in an attempt to study the force distribu-

Table 1. Cases for different wall stiffness to study effect on waler span. Cases 1 2 3 4 2 1 0 2 y l u J 8 1 6 3 : 8 1 t a ] g n o P g n o F g n a K [ y b d e d a o l n w o D

Table 2. ness. Cases 1 2 3 4 5

Figure 3.

Wall Seventy per cent stiffness of diaphragm wall of 1200 mm Seventy percent of CBP wall 800 mm diameter Sheet Pile Wall KSP IV One hundred per cent stiffness of diaphragm wall 1200 mm

Geotechnical modelling with different strut stiffDescription Removal of one row of struts Removal of one strut – stiffness at the particular layer was reduced (Lwaler ¼ 2L; Sstrut ¼ 2L) Removal of one strut – stiffness at the particular layer was taken as that from 3D STAAD Pro Analysis Removal of one strut – stiffness at the particular layer was reduced to Kstrut/1.5 (or 1.5 reduction factor) Removal of one strut – stiffness at the particular layer was reduced (Lwaler ¼ 2L; Sstrut ¼ 1.5L)

169

tion in the wall-waler-strut system. The structural analysis was carried out using STAAD Pro program and the geotechnical modelling was carried out using PLAXIS program. For the structural analysis, the actual strut and waler stiffnesses were modelled with four different wall stiffnesses as shown in Table 1, to study the waler span for the design of waler due to OSF. This has been done to determine the load distribution pattern for various types of ERSS system. For the geotechnical modelling, cases as indicate in Table 2, e.g. removal of one layer of strut and different strut stiffnesses were modelled in an attempt to study the force distribution in the wall-strut system. The example chosen in this paper aims to compare the results of analyses from PLAXIS 2D model and STAAD Pro 3D model. However, one should note that it may not be necessary to check all combinations of OSF. As there are six struts in the example, the critical cases to analyse are the case where OSF occurs at S6 after the excavation reached the final excavation level, the case where OSF occurs at S4 after the excavation reached the S5 level but before installing S5, etc.

5.

3D Analyses with structural models

Figures 3–12 illustrate the models of 3D structural analyses using the STAAD Pro structural finiteelement program. The objective of the 3D analyses carried out using the structural program is to verify the effect due to the removal of a single strut and compare with 2D PLAXIS plane strain analyses results.

Structural model of a space frame with soil spring and releases between plate elements.

K.F. Pong et al.

170


Figure 4.

Illustration of strut-waler connection adopted in the structural model.

Figure 5. Loading onto the wall.

As shown in Figure 3, the retaining wall was modelled with plate elements with release between plate elements to simulate the joints between the retaining wall such as contiguous bored piles wall, secant bored piles wall and diaphragm wall. Soil springs were modelled to simulate the presence of passive soil resistance. The soil spring constants were derived using Vesic’s equation from Vesic (1975). Figure 4 shows the struts and walers sizes and connections adopted in the structural models. Figure 5 shows the application of earth lateral pressure on the retaining wall. The soil parameters for

the calculation of the lateral earth pressure are shown in Figure 5. Figure 6 shows the area of the study for the effect of removing a single strut from the level S4 struts. Figure 7 shows the deflection mode of the temporary work system and Figure 8 shows the deflection of waler at level S4 in normal case and OSF case. Figures 9–19 show the bending moment result of waler with different wall stiffness for both normal case and OSF case. Figure 9 shows the effects of the removal of a single strut at S4 level on the bending



171

moment induced in the waler at S4 level for Case 1, where the wall is a 1.2 m thick diaphragm wall and cracked section properties (70% of full stiffness) are used for the stiffness of the wall. As shown in Figure 9, the maximum bending moment at the mid span and support of the waler increases to about 1.5 times of the bending moment in normal condition without OSF.

Figure 10 shows the effects of the removal of a single strut at S4 level on the bending moment induced in the waler at S4 level for Case 2 where the wall is a 0.8 m diameter contiguous bored pile wall and cracked

Figure 6.

Area of study for the 3D structural analyses.

Figure 7.

Figure 8.

Deflection of waler at level S4 with and without one-strut failure.

Deflection mode of the temporary work system.

172


Figure 9.

Figure 10.

K.F. Pong et al.

Waler bending moment with and without one-strut failure (Case 1: 70% stiffness of 1200 mm thick diaphragm wall).

Waler bending moment with and without one-strut failure (Case 2: 70% of CBP wall 800 mm diameter).



173

Figure 11.

Waler bending moment with and without one-strut failure (Case 3: Sheet Pile Wall KSP IV).

Figure 12.

Waler bending moment with and without one-strut failure (Case 4: 100% stiffness of diaphragm wall 1200 mm).

174

K.F. Pong et al.

section properties (70% of full stiffness) are used for the stiffness of the wall. As shown in Figure 10, the maximum bending moment at the mid span and support of the waler increases to about 1.8 times of the bending moment in normal condition without OSF. Figure 11 shows the effects of the removal of a single strut at S4 level on the bending moment induced in the waler at S4 level for Case 3, where the wall is KSP IV sheet pile wall. As shown in Figure 11, the maximum bending moment at the mid span and support of the waler increases to about 2.2 times of the bending moment in normal condition without OSF.


Figure 12 shows the effects of the removal of a single strut at S4 level on the bending moment induced in the waler at S4 level for Case 4, where the wall is 1.2 m thick diaphragm wall with full (100%) wall stiffness. As shown in Figure 12, the maximum bending moment at the mid span and support of the waler increases to about 2.6 times of the bending moment in normal condition without OSF. The comparison of bending moment ratio and span length ratio for different wall stiffness are presented in Figures 13 and 14, respectively. It is observed that the ratio of maximum bending moment in OSF case to maximum bending moment in normal case without OSF decreases when the retaining wall stiffness increases.

Figure 13.

Graph showing the ratio of waler bending moment with OSF to Normal case.

Figure 14.

Graph showing the ratio of waler span with OSF to Normal case.



Figure 15.

6.

175

Geometry of the retaining wall and strutting system with details of strut and waler sizes.

Plaxis modelling

A case study based on a cross-over cut-and-cover tunnel from the DTL 3 Project was carried out using PLAXIS program. The geometry of the retaining wall and strutting system of the cross-over cut-and-cover tunnel from DTL 3 Project is shown in Figure 15. Figure 16 illustrates the model of geotechnical analysis modelled in PLAXIS program for the normal case without OSF. The various cases of geotechnical modelling carried out are shown in Table 2. These various cases of analyses attempt to study the force distribution in the wall-strut system for the 2D PLAXIS model based on different input of strut stiffness derived from different scenarios. In the PLAXIS 2D analyses, the effects of OSF on the ERSS were studied by the scenario of failure of a single strut of the ERSS which consists of

total six layers of struts. Figure 17 shows a typical section, where a single layer of strut was considered for OSF case in the analysis. There were five OSF analyses carried out as listed in Table 2. For OSF Case 1, an entire layer of struts was completely removed. For OSF Case 2, the removal of one strut was simulated by reducing the stiffness at the strut. For OSF Case 3, the removal of one strut was simulated by using the stiffness for the strut taken from 3D structural model analysis. For OSF Case 4, the removal of one strut was simulated by reducing the stiffness for the strut to Kstrut/1.5, i.e. reduction factor ¼ 1.5. For each of the five cases listed in Table 2, six scenarios of OSF have been carried out. The first case is OSF for strut at level 1 (S1) when the excavation reaches strut at level 2 (S2) but before installing S2. The

176


K.F. Pong et al.

Figure 16.

Sectional view of the plain strain 2D model in PLAXIS program.

Figure 17.

PLAXIS model showing the OSF for strut at level 4 (S4) for Cases 1–5.



second case is OSF for S2 when the excavation reaches strut at level 3 (S3) but before installing S3. The third case is OSF for S3 when the excavation reaches strut at level 4 (S4) but before installing S4. The fourth case is OSF for S4 when the excavation reaches strut at level 5 (S5) but before installing S5. The fifth case is OSF for S5 when the excavation reaches strut at level 6 (S6) but before installing S6. The sixth case is OSF for strut at level 6 (S6) when the excavation reaches the final excavation level. The strut stiffness comparisons for all the cases of analyses mentioned above are tabulated in Figure 18. The results of strut forces for all the cases of analyses mentioned above are tabulated in Figure 19. The strut forces tabulated are based on the maximum strut forces for each layer of the struts for the various scenarios and sequence of OSF in each case of analysis with different approach of simulation of OSF analysed using 2D PLAXIS model. As shown in Figure 19, the results of strut forces from analyses of Cases 3 and 4 are similar. This means that the assumption of strut stiffness equivalent to the original strut stiffness multiplied by a reduction factor of 1.5 is a reasonable simplified assumption of the reduced strut stiffness as a result of a single strut failure as compared to the strut stiffness derived from the 3D structural analysis.

Figure 18.

Stiffness comparison for different cases of analyses.

177

Figure 13, which shows the comparison of waler span length with different wall stiffness, indicates that the increase in span length due to OSF ranges from 1.6 L to 1.2 L. Hence, it is reasonable to adopt 1.5 L for waler design. As shown Figure 20, in the comparison of diaphragm wall bending moment with different strut stiffness, the diaphragm wall design is generally governed by normal case, though there are locations where removal of one layer of struts governs. However, it may be too conservative to consider removal of one layer of struts in geotechnical modelling, hence adopting combined stiffness as illustrated in Figure 21 was found to be reasonable. It is also noted that the bending moment diagram of the diaphragm wall in Cases 3 and 4 analyses are very similar. This shows that the assumption of strut stiffness equivalent to the original strut stiffness multiplied by a reduction factor of 1.5 is a reasonable simplified assumption of the reduced strut stiffness as a result of a single strut failure as compared to the OSF strut stiffness derived from the 3D structural analysis. Based on the results shown in the study, for modelling of OSF in PLAXIS, the approach of using equivalent strut stiffness for OSF as illustrated in Figure 21 is appropriate and can be used.

K.F. Pong et al.

178


Figure 19.

Comparison of strut forces (in kN/m) for Upper Changi cross-over cut-and-cover tunnel.

Thus, the following derivation shown in Equations (4) and (5) can be adopted for waler design: 2

Mw1 ¼ 1 =8 Wreduced ð1:5 LÞ ! Factor ¼ 1 :05 2

Mw2 ¼ 1 =10 WOSF ðLÞ ! Factor ¼ 1 :05

ð4Þ ð5Þ

where Wreduced: Removed strut force (with reduced stiffness) WOSF: OSF strut force WOSF: OSF strut force accounting for removal of strut above/below it.



Figure 20.

Comparison of diaphragm wall bending moment with different strut stiffness.

Figure 21.

Recommended approach for derivation of OSF strut stiffness to be used in PLAXIS modelling.

179

K.F. Pong et al.

180 7.


Conclusions

In conclusion, analysis for OSF is actually a 3D problem and carrying out such 3D analyses covering all the cases of analyses considering the variability of the different parameters is very time consuming. In the conventional approach for the analysis of OSF case using 2D plane strain analysis, the entire layer of failing strut is removed and thus the forces are redistributed in the vertical direction only. This usually leads to more conservative design with heavier strut sections. A procedure to rationally idealise OSF from a 3D analysis to a 2D plane strain analysis which is a more practical approach for practising engineers has been presented. The comparisons of results of various cases have been carried out. The results of strut forces and bending moment of diaphragm wall from analyses of Cases 3 and 4 are similar. This means that the assumption of strut stiffness equivalent to the original strut stiffness multiplied by a reduction factor of 1.5 is a reasonable simplified assumption of the reduced strut stiffness as a result of a single strut failure as compared to the strut stiffness derived from the 3D structural analysis. The increase in span length due to OSF ranges from 1.6 L to 1.2 L. Hence, it is reasonable to adopt 1.5 L for waler design. In the comparison of diaphragm wall

bending moment with different strut stiffness, the diaphragm wall design is generally governed by normal case, though there are locations where removal of one row strut is governing. However, it may be too conservative to consider removal of one row of strut in geotechnical modelling, hence adopting combined stiffness was found to be reasonable. For design to accommodate possible OSF according to Clause 3.10 of TR26, the results presented in this paper show that the proposed approach of adopting a combined stiffness and using a simplified method in deriving the reduced equivalent strut stiffness for simulating OSF case in 2D numerical analysis is reasonable and appropriate. References BS 8002., 1994. Code of practice for earth retaining structures. London, British Standards Institution. LTA Civil Design Criteria – Revision A7 for Road and Rail Transit Systems, Land Transport Authority, 2010. Puller, M., 2003. Deep excavations: a practical manual , 2nd ed. Thomas Telford Publishing. TR26: 2010. Technical reference for deep excavation, Spring Singapore, Singapore. Vesic, A.S., 1975. Foundation engineering handbook, 1st ed. Van Nostrand Reinhold Company, Chapter 14, Pile Foundation, 561–563.

OSF Paper.pdf

Recommend Documents