PREVENTION OF FAILURE OF BRIDGE FOUNDATION AND APPROACH EMBANKMENT ON SOFT GROUND

Transcription

1 PREVENTION OF FAILURE OF BRIDGE FOUNDATION AND APPROACH EMBANKMENT ON SOFT GROUND Ir. Dr. Gue See Sew & Ir. Tan Yean Chin Gue & Partners Sdn Bhd 39-5, Jalan 3/146, The Metro Centre Bandar Tasik Selatan Kuala Lumpur Tel.: + (603) Fax: + (603) gnp@gueandpartners.com.my; ABSTRACT The success of bridge construction on soft ground relies on proper planning, design, construction control and site supervision. However, this is usually easier said than done and therefore there are still repeated geotechnical failures of bridge projects on soft ground. Most of the approach embankment failures are rather similar in nature and were induced by bearing capacity and stability of the embankment. The construction methods employed at site also have significant effect on the failures of adjacent piers. This paper presents case histories of two failures on bridge projects investigated by the Authors. The causes of failure, remedial works proposed and lessons learned are discussed. 1. INTRODUCTION The success of bridge construction on soft ground relies on the following major factors :- - Planning of Subsurface Investigation - Analysis and Design. - Construction Control and Supervision. However, most of the approach embankment failures investigated by the Authors are quite similar in nature and were induced by bearing capacity and stability of the embankment. The investigations carried out also clearly showed that construction methods employed at site also have significant influence. The failures can be prevented if the design consultant had taken care in geotechnical consideration in the analysis, design and construction. Two case histories of bridge failures investigated by the Authors are presented with causes of failure, remedial works proposed and lessons learned. In order to prevent history from repeating itself, this paper presents a brief guide to ensure successful construction of bridge foundation and approach embankment on soft ground. 2. CASE HISTORY Background The project is an access road with a reinforced concrete (r.c.) bridge over a river in Selangor. The proposed heights of the approach embankments on both sides of the abutments were about 8m with side slope of 1v(vertical) to 1.5h(horizontal). These embankments were to be constructed over a layer of very soft silty clay or clayey silt of 3m to 9m thick with Standard Penetration Tests values (SPT N ) of zero. Underlying the top very soft layer is 3.5m to 5.5m thick of medium dense silty sand followed by completely weathered shale with SPT N values vary between 30 to 50 blows/300mm. The liquid limit (LL) of the clay is about 78% and average moisture content is about 106%. Fig. 1 shows the general layout of the project. Fig. 2 shows a schematic profile of the subsoil. At abutment A, the wing walls were designed on piles but Abutment B was designed with cantilever wing walls.

2 Fig. 1 Layout of the bridge in case history 1 River Fig. 2 Subsoil profile of the site

3 2.2 Slip Failures of the Approach Embankments The first slip failure occurred on the right side of the approach embankment behind Abutment B when the fill was about 3m high above ground level. The designer inspected the failure and gave instructions to build a berm and continue filling. Three days later, the second slip failure at the site occurred. This time, it was on the left side of the same embankment as shown in Fig. 3a. Despite the failure, the designer instructed the Contractor to dig a ditch of 0.5m wide by 1m deep at the toe of left side of the embankment and allowed earth filling works to continue towards Abutment B at slower rate. Piling works at Abutment B started about 6 weeks after the first slip and the construction of the whole abutment was completed in about 5 weeks. After two months and when the embankment reached a height of 7m (1m below formation level), the third slip failure occurred as shown in Fig. 3b. Fig. 3 (a) Second slip failure at the left (b) Third slip failure The third slip occurred on the same location as the first slip and together with the embankment directly behind Abutment B. There was a drop of about 2.4m at the slip interface. On further investigation, Abutment B was found to have moved forward by 0.31m and 1.12m on the left and right sides respectively of the abutment. The abutment had also tilted vertically about 2 degrees clockwise towards embankment. Fig. 4 View of exposed pile cap and piles on the right side of Abutment B

4 In the failure investigation of Abutment B, the piles immediately below the pile cap were examined by excavating a trench on the right side of the Abutment B. Fig.s 4 and 5 show the condition of the piles. Cracks on pile B appear to have propagated from left side of the pile towards the fill. This indicates that the piles had been subjected to high lateral stresses imposed by the fill. This pile also shows a slight curve which suggests that there has been some restrain at the lower end of the pile. Fig. 5: Close-up view of the three exposed piles Fig. 6 shows the crushing of Pile C, which further indicates lateral compression due to the fill. Fig. 7 shows schematically the movements of pile cap and piles. Fig. 6 Pile C showing crushing of pile near pile cap

5 Fig. 7 Movements of pile cap and piles (view from right side of Abutment B) The forth slip occurred on the left side of the embankment behind Abutment A about three months after the third slip. At the time of the failure, the height of the embankment was about 6.5m above its ground level (1.5m below proposed formation level). Fig. 8 shows the slip failure. Fig. 8 Fourth slip failure on the left side of embankment behind Abutment A 2.3 Geotechnical Investigation In order to ascertain the causes of the failures, geotechnical analyses and investigations were carried out. Independent soil investigation was carried out after the failure with the following objectives :- i) to obtain physical and strength properties of the subsoil; ii) to determine the extent of the soft materials below the fill; iii) to ascertain the depths and dry density of fill at various positions of the embankment; iv) to study the ground water table and trace the slip lines if possible; v) to check and compare the results with those obtained during previous soil investigation. The results of the independent soil investigation generally confirm the results of the previous soil investigation except that the very soft silty clay layer appeared to have gained some strength and decrease in thickness due to some consolidation.

6 Generally, the average undrained shear strength (s u ) of the very soft clay is 10kPa with a lower bound value of 7.5kPa as shown in Fig. 9. The sensitivity (St) of the clay ranges from 2 to 8. Fig. 9 Undrained shear strength (s u ) of the soft clay Bearing capacity and limit equilibrium stability analyses carried out indicate that the subsoil could not support embankment height in excess of 2.7m if there is no ground treatment or strengthening. From the analyses, it is very clear that the embankment height of 8m proposed by the designer is not safe. Further back-analyses carried out on the failed embankment indicated that the s u of the clay was 11kPa. This is in good agreement with the average s u of 10kPa obtained from the S.I.. The analyses on the foundations for Abutments A and B were also carried out. The lateral earth pressure on piles was calculated using stress distribution behind piles proposed by Tschebotarioff (1973) as shown in Fig. 10. Since the spacing of piles was about three times the width of the pile, therefore, the group of piles and soil can be assumed to act as a unit. Fig. 10 Additional forces on abutment (soil undergoing lateral movement)

7 The ultimate lateral resistance, Ru was calculated using assumption of Poulos & Davis (1980) which assumed 4s u at the surface and increases to a constant value of 9s u at three times the width of the pile. The critical height of embankment that will induce a lateral force equivalent to the ultimate resistance was evaluated for the different thickness of soft clay. For abutment B the lateral resistance of the pile group and soil would be exceeded when the height of embankment was 5m to 5.5m. Therefore, the movements of abutment B could have happened when the height of embankment was about 5m, i.e. 3m below the proposed formation level. The calculations have also shown for Abutment A, that the lateral resistance of the pile group and soil would be exceeded when the height of embankment was 7.5m to 8.0m. However, Abutment A was only subjected to a height of 6.5m (1.5m below the proposed formation level), therefore it did not fail. Abutment A could withstand a higher embankment height than Abutment B because of its shape and dimensions. Abutment A was designed and constructed with piled wingwall. The total number of piles used was 88, and had 36 piles more than Abutment B. 2.4 Proposed Remedial Works Following many meetings with the designer, the final accepted proposals are as follows :- (a) Underpinning of Abutment B with 48 Nos. of micropiles, Abutment A with 10 Nos. of micropiles. (b) Use of reinforced soil wall behind Abutment A. (c) Piled embankment to be used for embankment fill exceeding 2.5m high. (d) Geogrid reinforced embankment for fill between 1.5m to 2.4m high. The total estimated cost of remedial works was about 3.7 million ringgit. 3. CASE HISTORY Background Similar to Case History 1, this project was under construction when failures occurred. This project is also an access road with prestressed concrete bridge over a river in Sarawak. The proposed heights of the approach embankments on both sides of the abutments were about 5m with side slopes of 1v(vertical) to 1.5h(horizontal). These embankments were constructed over 25m thick of soft coastal and riverine alluvium clay followed by dense silty Sand and very stiff silty clay. The soft alluvium generally has SPT N value of zero and average moisture content of more than 70%. Fig. 11 shows the partially completed bridge after failure and removal of failed materials. The layout is shown in Fig. 12 and the subsoil profile in shown in Fig. 13. Fig. 11 Overview of partially completed bridge.

8 Fig. 12 Layout of the Bridge in Case History 2 Fig. 13 Subsoil condition In the construction drawings, the approach embankments using local fills were supported by 200x200mm RC piles with pilecaps. In addition, 6m length wood piles were also used between the RC piles for further support of the embankment fill. More wood piles were also installed on the banks of the river trying to stabilize the lateral displacement of the soft alluvium. The abutments and piers are generally supported by 400mm diameter spun piles. 3.2 Slip Failures of the Approach Embankments

9 A deep seated slip failure occurred at the approach embankment about 25m from Abutment II. It happened when the fill reached about 3m high. Fig. 14 shows the shear drop after removal of some of the fill near the abutment. Fig. 14 Shear Drop at about 25m from Tilted Abutment Abutment II has tilted away from the river with the magnitude of about 550mm at the top of the abutment at the time of the site inspection by the Authors who were carrying out geotechnical investigation of the failure. The tilt translates into an angular distortion of 1/6. Due to the excessive angular distortion, the integrity of the spun piles driven to set into the stiffer stratum has also been affected as it exceeds the normal threshold of about 1/75. Due to the tilt of the Abutment II away from Pier II, a gap of about 300mm wide was observed between the two bridge decks at the pier s pilecap. Fig. 15 shows the photograph of the tilt at the Abutment II and the gap between two bridge decks. The failure also caused the pilecap at Pier II to tilt as shown in Fig. 16. Fig. 17 shows the schematic diagram of the possible slip plane relative to the deformed structures. Fig. 15 Tilted Abutment and Observed Gap between Bridge Decks These observations infer that the slip of the Approach Embankment near Abutment II is deep seated and is consistent with the depth of the soft alluvium. The cause of the rotational slip failure is due to the weak subsoil unable to support the weight of the approach embankment. The weight of embankment initiated the consolidation settlement of the soft subsoil and mobilised the low shear strength of the slip failure plane. The use of the 6m wood piles and RC piles offers little lateral resistance and instead, extends the

10 rotational slip deeper into the soft subsoil. At the pier, the bridge decks, being simply supported and fixed to the abutment via bearing pad, had moved along with the displacement of the abutment. Fig. 16 Titled pilecap at Pier II Fig. 17 Schematic of the slip failure At the start of the construction, the contractor observed that their personnel could not walk on the riverbanks without their feet sinking into the soft subsoil to a depth of about a foot. This observation infers that the upper subsoil undrained shear strength of the subsoil is estimated at about 10 kpa. As preliminary check using correlation of 5*s u to obtain ultimate bearing capacity of 50 to 60kPa, the estimated maximum height of fill that can be supported is about 3m which is consistent with the observed failure during embankment filling to 3m high. Therefore, if the designer and contractor had carried out simple checks, failure could have been prevented.

11 Additional subsurface investigation after the failure shows that the undrained shear strength from the vane shear tests range from 18 kpa to 51 kpa with remoulded strength of 7 kpa to 12 kpa. The higher s u obtained from the additional S.I. is due to the gain in strength during the whole period of filling. 3.3 Remedial Measures The rotational slip failure of the approach embankment is due to many factors such as inadequacy in the design, construction control, the soft subsoil and absence of adequate ground treatment. Several remedial options were explored for the embankment. The first remedial option is to remove the failed embankment fill and re-construct a new RC ramp (bridge) with ground beams for increased rigidity. This option avoids the weight of the fill bearing on the soft subsoil. The second option is to surcharge the soft subsoil in combination with prefabricated vertical drains to accelerate the consolidation process of the clayey subsoil and allow the subsoil to gain strength with time. The third option is to use piled embankment with slab to transfer the embankment load to the stiffer soil stratum instead of the soft upper clay. After much consideration by the client, the third option was chosen for the shortest construction time in order to put the bridge into service and no long term risk of further subsoil settlement. However, in this option, the soffit of the RC slab should be at or below the original ground level to avoid additional load on the soft upper subsoil stratum that can generate negative skin friction on both the abutment and embankment piles. At the tilted abutment, analyses of the pile head movement of the existing piles showed that integrity of the piles is doubtful and shall be compensated. However, there are two options of installing the compensation piles; firstly at the sides of the existing pile group and secondly, behind the abutment. The first option requires demolishing the existing abutment and enlarging the pilecap. The second option minimizes modification of the abutment but requires longer I-beams for support of the bridge deck. In addition, there is also a risk the compensation piles might be impeded by the wood and RC piles since the location of these piles might have displaced along with the slip failure. Option 1 was chosen for the minimal remedial cost by reusing the existing I-beams and minimizes risk of the new spun piles striking the existing wooden and RC piles. For the pier foundation, the existing spun piles are fully compensated by demolishing the existing pier and pilecap and installing new ones at the sides. The total estimated cost of remedial works is about 1.3 million ringgit. 4. LESSONS LEARNED AND PREVENTIVE MEASURES From the two case histories presented in this paper, it is obvious that they are very similar in nature and can be categorized to be caused by the following factors:- - Inadequacy of geotechnical design for the approach embankments or abutments. - Lack of understanding of the subsoil condition and awareness on the possible problems/failure that could happen during construction. - Lack of construction control and site supervision by the Consultant It is very obvious that the two failures are due to inadequacy of geotechnical design for the approach embankments and abutments. If proper geotechnical analyses and designs were carried out, the failures could have been prevented. The designs of approach the embankment and abutment are quite similar to normal fill embankment where key issues like stability and settlement shall be properly addressed. Although settlement calculation is equally important, they are not discussed here as it is not the main factor causing the failures but rather a long term serviceability problem that required regular maintenance. For embankment and abutment stability, both circular and non-circular (wedge) failure surfaces shall be evaluated using a limit equilibrium analysis (Tan & Gue, 2000). It is very common to wrongly assume that as long as the structural design of an abutment has considered both vertical and lateral pressures, slip failure would not occur. A good example of abutment instability is shown in Fig. 17 where a deep seated instability of the embankment fill behind the abutment seriously affect the stability of the abutment. The most critical condition that usually triggers failure is during filling where the stability of an embankment

12 shall be analysed based on undrained shear strength (s u ) of the subsoil. The recommended Factor of Safety (FOS) against instability is at least 1.2 for short term. Long term stability is usually less critical as the subsoil increases its strength with time. The stability of embankments and abutments in long term shall be checked using effective stress strength parameters (c and φ ) and the minimum FOS required is 1.4. A quick preliminary check on the stability of the embankment is possible using modified bearing capacity equation below : q allow = (s u. N c / FOS) where : q allow = allowable bearing pressure = (γ fill.h + 10) (kn/m 2 ) γ fill = bulk unit weight of the compacted fill (kn/m 3 ) H = allowable height of embankment (m) s u = undrained shear strength of the subsoil (kpa) N c = 5 (suggested by Authors for ease of hand calculation) FOS = Factor of Safety Note : The 10kPa allowance in the q allow is to cater for traffic load. Design consultant, consultant s site engineer(s) and contractor should have some fundamental geotechnical knowledge which include understanding of the subsoil condition and awareness on the possible problems or failures that could happen during construction. A good example is shown in Section 3.2 where the contractor were aware that their personnel could not walk on the very soft riverbank and could have used the simple bearing capacity equation to check the allowable height of the fill the subsoil can support. More often than not, failures were due to bad temporary works that were never considered for in the design. One serious problem that usually occurs for bridge project is the temporary fill placed by the contractor to form a temporary platform to facilitate their piling or other construction works. If not careful, slip failure in subsoil could be triggered by the load from the temporary fill. Therefore, it is recommended that the design consultant should consider the possible construction method to be used by the contractor and designed for it. The design consultant shall also ensure that during construction, the contractor must carry out works according to the approved method statement to prevent failure. It is also important for the design consultant to ensure the method statement including temporary works proposed by the contractor does not cause failure. Finally, the proper full-time site supervision by the consultant s representatives with adequate site experiences and knowledge are also very important to prevent failure due to temporary works and ensure permanent works are constructed according to the drawings and specifications. Another common problem caused by temporary fill over soft ground is the failure to remove the temporary fill after construction. The temporary fill will cause the compressible subsoil to settle with time (consolidation settlement). If this area has piles, then the piles will be subjected to down drag (negative skin friction) due to the settling subsoil and reduce the capacity of the piles. If the down drag is not catered for in the design, the piles will have lower allowable capacity and larger settlement causing distortion to the structures. Therefore, the design consultant shall ensure the removal of temporary fill after construction by the contractor or to design the piles to accommodate negative skin friction. 5. CONCLUSIONS The success of bridge construction on soft ground relies on proper planning, analysis, design, construction control and site supervision. However, from the two case histories presented in this paper, it is obvious that they are very similar in nature and can be categorized to be caused by the following factors:- - Inadequacy of geotechnical design for the approach embankments or abutments. - Lack of understanding of the subsoil condition and awareness on the possible problems/failure that could happen during construction. - Lack of construction control and site supervision by the Consultant To prevent embankment and abutment failure due to instability, both circular and non-circular (wedge) failure surfaces shall be checked using limit equilibrium analyses. A quick preliminary check on the stability of the embankment is possible using modified bearing capacity equation of

13 q allow = (s u. N c / FOS) It is also important for design consultant, consultant s site representatives and contractor to have some fundamental geotechnical knowledge so that any irregularities at site can be spotted and precautionary actions carried out before failure occurs. Finally, proper full-time site supervision by the consultant s representatives with adequate experiences and knowledge are also very important to prevent failure due to un-engineered temporary works. REFERENCES GUE, S.S (1988). An Investigation into Geotechnical Failures of a Bridge Project. IEM/RRIM Joint Engineering Symposium. Johor Bahru, Malaysia, pp LADD, C.C. (1991). Stability evaluation during staged construction. J. Geotech. Eng.,ASCE. 117(4) : PECK, R.B. (1969). Advantages and limitations of the observational method in applied soil mechanics. Geotechnique. 19(2): POULOS, H.G. and DAVIS, E.H. (1980). Pile Foundation Analysis and Design. John Wildy and Sons, Inc. Canada. SKEMPTON, A.W. (1951). The Bearing Capacity of Clays. Building Research Congress, London. 1 : SKEMPTON, A.W. & NORTHEY, R.D. (1953). The post-glacial clays of the Thames Estuary at Tilbury and Shellhaven. Proc. 3rd Int. Conf. Soil Mech. Found. Eng., Zurich. 1 : TAN, Y.C. & GUE, S.S. (2000), Embankment Over Soft Clay Design and Construction Control, Seminar on Geotechnical Engineering 2000, IEM (Northern Branch), Penang, 22 & 23 September, TSCHEBOTARIOFF, G.P. (1973). Foundation, Retaining and Earth Structures. 2nd Edition McGraw- Hill Inc. USA. BIOGRAPHY Ir. Dr. Gue See Sew :- After graduating with First Class (Honours) Degree from University of Strathclyde, UK in 1979, Dr. Gue joined the Public Works Department, Malaysia as a Bridge Design Engineer. In 1981, he won the prestigious Kuok Foundation Award. He completed his doctorate at University Oxford in 2½ years and joined the Public Works Department Malaysia in In 1990, he left the government service and is now the CEO/Managing Director of Gue & Partners Sdn Bhd, a geotechnical consulting firm. Dr. Gue was the President of the Institution of Engineers, Malaysia (IEM), April 2001 April He is a Board Member of The Board of Engineers, Malaysia ( , ). Dr. Gue was the Chairman for the Taskforce Committee on Policies And Procedures For Mitigating The Risk Of Landslide On Hill-Site Development. He is the Regional Chairman of the Coordinating Committee of APEC Engineer Register ( ). Dr. Gue is also the Council member of National R&D Council for 2002 to He has published 60 technical papers related to geotechnical engineering in various conferences and seminars. Ir. Tan Yean Chin :- Ir. Tan obtained his Bachelor Degree in Civil Engineering with First Class Honours from Universiti Teknologi Malaysia in He later obtained his Master Degree in Geotechnical Engineering and The Chin Fung Kee Prize for outstanding academic performance from Asian Institute of Technology, Bangkok in He is now a director of Gue & Partners Sdn Bhd; a geotechnical consultant firm. Ir. Tan is currently the Deputy Chairman of the Geotechnical Engineering Technical Division and Council member of the Institution of Engineers, Malaysia (IEM). He has published more than 30 technical papers on geotechnical engineering in local and overseas conferences and seminars. In 1998, he won The Tan Sri Raja Zainal Prize presented by IEM for an outstanding technical paper.