Modern Touch to Historical Railway Station
Delft Railway Station, Netherlands
Delft Railway Station, Netherlands referred as the oldest railway line in the country between Hague and Rotterdam cities. Now a new rail tunnel of 2.3km is opened recently that integrates the stations hall with city municipal offices. The brilliant structure is designed by Mecanoo, international architect company in Delft.
The Delft railway tunnel project comprises of the design and construction of a 2.3 km long; four track railway tunnels, an underground railway station and an underground parking in the historic city centre of Delft. The lowest excavation level is about 10 m below the ground surface. The ground water level is 1.5 m below ground surface.
Shallow Surface Strata
The shallow sub-surface strata in the western part of Netherlands can be characterized by almost 20 m of soft soil of Holocene origin underlain by medium to dense sands of Pleistocene origin. North of the tunnel area a factory is situated that extracts large amounts of brackish groundwater from the Pleistocene sands. The drawdown results in low pore pressures in the Pleistocene sands, an advantage during construction as it reduces the risk of up-lift of the bottom of the excavation. With limited measures deep excavations can be achieved although the final tunnel construction is designed to retain higher upward water pressures in case the extraction ends.
Extensive risk assessments regarding safety and legal aspects were performed, as it was the first project in the Netherlands where large buildings supported by and integrated with a railway tunnel. Mecanoo and ABT are designers of the city hall tendered in a separate contract. All designing parties work closely together with the contractor, the client and its advisor DRB integrated their partial designs.
When working on a project with various types of people, in terms of different contracts, proper communication was indispensable. In this project it is achieved by intensive interface management. The Client’s Change Control Board decides in cases were contract changes are required to improve the overall design or to minimise risks, such as:
- Contractually, the designers of the city hall had to take measures to reduce train induced vibrations. However, it proved impossible to incorporate in their concept and it was decided to take measures underneath the railway tracks.
- It was decided to transfer parts of the scope of the city hall (an underground bicycle park directly adjacent to the tunnel) to the tunnel contract, in order to minimise construction, logistic and planning risks.
Inclinometers for Risk Evaluation
The design process and the construction of the tunnel are based on the method of systems engineering (Everaars et al., 2010).
In order to validate the applied design methodology, monitoring instruments are attached to some cross sections before the start of construction. Inclinometers are installed at close distances to the diaphragm wall and at one location directly inside the diaphragm wall. A series of settlement markers up to a distance of 25 m of the diaphragm wall is installed to monitor horizontal and vertical ground deformations. Monitoring data enables verification of design models.
As this experiences had to learn that failures of diaphragm wall trenches were often caused by joint anomalies, sonic logging equipment was installed at panel joints close to existing (historical) buildings. The equipment comprises four watertight HDPE tubes (two at both sides of the joint), fixed to the reinforcement cage. After concrete hardening, sonic logging is performed by lowering transmitters and receivers at opposite sides of the diaphragm wall for joint inspection.
Adjusted Construction Methods
Initially, a short sheet pile wall was anticipated for at a minimum distance from the facade of the historical railway station to improve trench stability. The idea was that the sheet pile wall, which is necessary to construct the heat/smoke outlet, conducts between the facade and the future tunnel would reduce the required bentonite level in the stable diaphragm wall trench. Analysing the introduction of stabilised soils between sheet pile wall and diaphragm wall however resulted in low safety factors. For this reason high bentonite levels were required and to install the sheet pile wall to schedule after concrete hardening of the diaphragm wall.
 |
Pre-stressed grout anchors in the central section of the railway station were originally planned to pass through metal tubes incorporated in the reinforcement cages. As these tubes should form an important rheological obstacle during the concreting of the diaphragm walls, this concept was dropped. The lower grout anchors with spacing of 1.9 m (two per panel) will now pass through drilled holes. Those holes will be cored during excavation. To avoid drilling through reinforcement special cages were designed. The holes for the top anchors with spacing of 3.8 m (one per panel) could pass through the diaphragm wall in the unreinforced central zone.
Northern Part
The northern part of the railway station has two sections; one section has a lower roof and as a consequence an earth covers of several metres. The second section has a 2.7 m higher roof between both the sections a solid concrete beam crosses the station transversally from one outer diaphragm wall to the other outer diaphragm wall. The span is approximately 40 m. The excavation depth is 10 m. To reduce the necessary dimensions of the beam, it is supported by the tunnel partition walls. This means that the generally applied top-down construction sequence had to be altered to a bottom-up sequence in the lower northern section in order to finish the supporting walls and transversal beam before the roof could be concreted. To avoid a too high quantity of dewatering, the necessity of compartment walls was obvious.
Underground Measured Construction
In the second section with the high roof, central columns with longitudinal spacing of 8.1 m support the roof, while at the far ends the roof slab is fixed to the outer diaphragm walls. As the high structural column loads of the city hall have to be supported by the diaphragm walls and the central columns, measures had to be taken to limit possible differential settlements. Due to the great variation of concentrated vertical structural loads (4 MN to 28 MN), large solid spreading beams had to be casted on top of the outer diaphragm walls. Conservatively, the load spread through the joints of diaphragm wall panels by mobilisation of side friction was not taken into account. Load spread could be achieved with beam heights of approximately 3.0 m.
 |
The section with the high roof was constructed top-down. After installation of the diaphragm walls and barrettes, the spreading beams were casted after partial removal of the top section of the diaphragm walls. Reinforcement of the concrete spreading beams is connected to the reinforcement cages of the diaphragm walls. Before covering the tunnel section with the concrete roof, anchor piles are installed which counterbalance uplift below the future tunnel floor.
Central Part
The central part of the underground railway station is characterised by two important aspects. At one side that is the historical building of the old railway station where as needs to be preserved. It is located at no more than 3.0 m distance from the outer diaphragm wall. On the opposite side, the future ground surface level is equal to the intermediate floor level of the underground railway station about 3.0 m below the tunnel roof slab. This asymmetrical load situation had to be addressed in both design and construction.
 |
In order to reduce the deformations of the historical railway station, the allowable horizontal diaphragm wall deflection (1/100 of the excavated height) that was stipulated for the northern and southern part has to be limited. Therefore the diaphragm walls installed with smaller panel widths of 3.8 m in comparison with panel widths of 7.3 m as applied elsewhere. By doing this, the expected horizontal ground deformations during the bentonite phase of the diaphragm wall excavations will be reduced.
Southern Part
The construction sequence of the low section was introduced to facilitate casting of five partition walls. At the extremity of this section however, a length of 14.0 m of the roof slab is cast using the top-down method, as this slab temporary supports a compartment sheet pile wall. The reinforcement in the connection roof/diaphragm wall was designed taking into account the self weight of the slab and supports by the partition walls (top-down). The high roof slab settlement which takes place during top-down construction requires compensation. Before placement of fill on top of the roof, the slab will be jacked up to reduce bending moments and shear forces at the diaphragm wall connections. Reaction force is obtained from the partition walls.
Barrettes
At middle of the tunnel roof is support by a row of columns with spacing of 8.1 m. These columns have longitudinal concrete beams above and below. The lower beam, being part of the tunnel floor is supported by barrettes instead of a continuous diaphragm wall. As it would be too expensive to cast barrettes up to roof level and shorten them afterwards down to tunnel floor level, temporary steel columns were applied in this section. Before installation, the reinforcement cages and steel profiles were welded together at ground surface. The reinforcement was lowered into the trench. The steel profiles could be used to support the roof during top-down construction. The gravel can easily be removed during the excavation. After the floor is cast, a formwork can be placed around the steel profiles to give the columns a concrete look.
|
|
| In accordance with CUR166 (2005) interaction of the Vertcal and horizontal wall stability should be analysed where vertical loads are substantial (over 12.5kN?Sqm). |
Engineering Diaphragm Walls
Dutch qc-Theory
The diaphragm walls below the city hall have combined functions. They retain ground and groundwater pressures and they support the city hall located on top of the tunnel. In accordance with CUR166 (2005) interaction of the vertical and horizontal wall stability should be analysed where vertical loads are substantial (over 12.5 kN/Sqm ). In this case loads from the city hall are over 50 kN/Sqm .
The horizontal stability of the propped diaphragm walls was assessed using the computer program MSheet (Delft Geosystems, The Netherlands) based on the Winkler theory of beams with elastic-plastic spring behaviour. Separately, assessment of the vertical stability of the diaphragm walls were performed in accordance with the Dutch standard for compression piles NEN6740 (2006) and NEN6743-1 (2006) applying the computer program MFoundation (Delft Geosystems, The Netherlands). Pile design was based on the Dutch qc-method and took account of stress reduction by trench excavation and a complex geo-hydrological profile. The described design routine proved to be very practical.
Horizontal stability
According to design approach B outlined in CUR166 (2005) a Finite Element Study (FEM) was performed using representative ground parameters, groundwater levels and geometry in the different construction stages. For critical construction stages separate calculation steps were added in the FEM using design parameters. In the FEM the direction of the friction angle (δ, o) at the concrete wall and soil interface is automatically determined for vertical loading of the wall.
In the elasto-plastic spring model the approach is different. The model uses parameters such as c’, ?’, δ’ kh (kN/Cum) is the horizontal sub-grade reaction. In the final construction stage the diaphragm wall is vertically loaded up to 1.9 MN/m’. The balance of vertical forces has to be achieved by rotating the wall friction angle bottom-up, from tip level to about the excavation level, at the active side of the retaining wall. The presented case has diaphragm wall tip levels of NAP -23.5 m penetrating 5 m into dense to medium dense Pleistocene sand.
The verification results for horizontal stability of the retaining walls show that:
- The calculated maximum bending moments (ULS) in FEM and the spring model are comparable, but appear in different construction stages.
- Introduction of vertical loads increases maximum bending moments in the retaining wall.
- In the spring model no clear increase is observed of the maximum bending moment between completion of the tunnel and application of vertical load on the retaining wall.
However a second order bending moment was added to the results of the final spring model construction stage. The additional moment was calculated using the maximum deflection of the retaining wall from the spring model calculation and the introduced vertical load from the superstructure. It was 135 kNm/m’ (0.07 m x 1,9 MN/m’). The results from the FEM and the spring model become comparable, 1,308 kNm/m’ versus 1,154 kNm/m’.
Experimental Methods
|
|
| The high ground water levels impermeable retaining walls are required. Vertical loading conditions however result in irregular settlements of the diaphragm wall panels |
In the FEM and the spring model wall friction was mobilised over the wall length including the Holocene soil layers. The qc-based method took account of wall friction from the top of the deep Pleistocene sands. In the FEM and the spring model the wall friction is assessed using the slip-method while the Dutch qc-method relates wall friction to qc using empirical pile constants (αp = 0.5, αs = 0.006 and s = 0.63 according to NEN6743-1 (2006)). End bearing of the spring model was also based on the qc method where in the FEM the end bearing was assessed by multiplication of effective stress below the tip of the diaphragm wall (including the effective weight of the wall) with the tip surface.
Vertical Support and Deformations
The city hall columns are supported by the outer diaphragm tunnel walls and the barrettes at the centre of the tunnel. Strict differential settlement criteria were prescribed in the tunnel contract to assure the structural integrity of the city hall. In transverse direction differential settlement between adjacent structural columns has to be reduced to <20mm and in longitudinal direction to <15mm. Maximum absolute vertical column settlement has to 30mm.
To control (differential) settlement three possible options were considered:
- Post base and shaft grout injection. Application of this technique increases pile stiffness as well as the bearing capacity. The technique has not widely been applied to improve the performance of diaphragm walls and barrettes. The effect of post injection for this project could only be verified by load tests.
- Spreading beam on top of the diaphragm walls and barrettes for steady load distribution at tip level.
- Varying tip levels to control vertical stiffness. A feasibility study indicated that the increase of tip levels by about 7 m is equivalent to the application of shaft and base injection. In addition, prediction of performance of diaphragm walls after injection is subject to uncertainties.
A summary of the calculation results is as follows; city hall column settlements between 5 mm and 25 mm were assessed. The maximum column loads of up to 28 MN per barrette (L/W is 3.3 m/1.0 m) requires tip levels of 46.5 m below ground surface.
The high ground water levels impermeable retaining walls are required. Vertical loading conditions however result in irregular settlements of the diaphragm wall panels. The spreading beam contributes to reduce differential vertical settlements of neighbouring panels to 5mm. Joints between diaphragm walls panels have limited differential settlement capacity to a maximum vector deformation of 25mm. Therefore, measures such as post joint injection are available for instant use during construction.
Conclusion
Diaphragm walls in the cut and cover tunnel reduces ground deformations and therefore minimises settlement of nearby buildings. The relatively high bearing capacity of diaphragm walls creates opportunities to build on top of the tunnel. A new city hall will be constructed directly on top of the railway tunnel over a length of 130 m. In consequence of construction on top of the tunnel, the diaphragm walls below have two interacting functions. They retain ground and ground water and they provide vertical support. A FEM analysis has proven that assessment of horizontal and vertical stability can be performed separately.
References: