The Minaret of the Great Mosque in Algiers, a Structural Challenge ()
1. Introduction
The Algerian state represented by ANARGEMA (Agence Nationale de Réalisation et de Gestion de la Mosquée d’Alger) has commissioned the German Joint Venture KSP/KuK (Consultant Architects and Engineers), winner of an international competition, with the planning of the Great Mosque of Algiers [1]. The building complex is situated in the central axis of the famous Golf of Algiers, facing the Mediterranean Sea, some half way between the old city and the airport and is currently being erected by the China State Construction Engineering Corporation.
The mosque itself covers a surface of 600 m × 150 m [2]. Additional buildings are provided for a cultural center, a library, a religious university and a huge underground parking. The building’s complex is seen as the future architectural landmark of the city.
The prayer hall has a squared plane with the side of 150 m, which can accommodate 36,000 prayers and has a central dome with the apex height of 70 m (Figure 1). To mitigate the highly seismic risk of the region, its structure is base isolated by means of a combination of mechanical seismic isolators and hydraulic dampers.
The minaret is a very slender parallelepiped with a total height of 265 m above ground and a squared plane with the side of 26.5 m (Figure 1). Due to this slenderness, to the particularity of the stiffening system and to the strong seismicity of the region, the structural design of the minaret has faced several technical challenges.
The paper presents the design philosophy and some main features of the structure.
2. Seismicity
The north of Algeria is a very strong seismic region. Due to the major national importance of the project, the aseis-
mic design of the minaret has been based on a micro zonation study authored by the Algerian Centre of Applied Research in Earthquake Engineering (CGS).
It recommended the design seismic spectrum given in Figure 2. The spectrum corresponds to a peak ground acceleration of 6.5 m/s2 and to a return period of 1000 years.
The fundamental period of vibration of the minaret is about 3.7 s and corresponds to a translation. The first torsional period of vibration is 1.1 s.
3. Structure
The minaret will accommodate a national history and art museum and a corresponding research institute. Over the height of the building there are 5 blocks of 5 stories each separated by sky foyers. The height of each story is 5.85 m and that of a sky foyer is 11.7 m. The transparent top of the minaret envelopes the summit small tower which is typical for Maghreb’s region. It is 41 m high and its structure is of steel and glass. There are two underground levels, with the total height of 11.2 m and a squared plane with the side of 50 m. This enlargement of the minaret foot was crucial for ensuring the foundation system.
From the ground level up to the bottom of the summit tube there are four reinforced concrete (RC) cores situated in the corners (Figure 3). They have a squared perimeter with a side varying from 7.75 m at the ground level up to 7.5 m at their top, with variable wall thicknesses over the height and with external walls thicker than the internal ones. They carry out the whole building weight of about 700 MN above ground. The cast-in-place RC floors are designed as girder grids. The main floor beams depicted blue in Figure 3 are also part of the horizontal stiffening system as they couple the cores.
Figure 2. The elastic design spectrum (normalized by means of the gravitational acceleration g = 9.81 m/s2).
The four corner cores have a height to depth ratio of about 30 and are therefore not able to stiffen the tower even if the coupling floor beams are considered. The necessary lateral stiffness and load-bearing capacity can be achieved only if the whole building width is activated. In that case the height to depth ratio becomes about 10. An “outer tube” has to be therefore created. This has been realized by coupling the RC cores by means of X-crossed façade diagonals made of steel sections (Figure 4). On aesthetical grounds the façade diagonals were not desired at the sky foyers, so that a discontinuous bracing has to
Figure 4. Coupling façade diagonals and steel members cast within the core’s external walls.
be used. In order to avoid the transfer of the high internal forces from the coupling steel diagonals to the RC walls and back, a steel construction has been provided within the exterior walls of the cores (see the diagonal and horizontal cross-bars as well as the vertical bars depicted in Figure 4). In this way a composite stiffening system has been created, which combines the RC one made of the corner cores and the coupling floor beams with a spatial steel truss (Figure 5). To accommodate the embedded steel profiles the external walls of the cores have thicknesses varying from 1 m at the ground level and 45 cm towards the top. The corresponding internal walls are 75 and 40 cm, respectively. The coupling effect of the fa- çade bracing is outlined in Figure 6. The overturning bending moment M0 induced by the seismic action yields internal axial forces N and bending moments M within the cores. Their relative magnitudes depend on the relative stiffness of the two stiffening components, i.e. the cantilevered cores, on one side, and the spatial truss with very stiff flanges, eccentric joints and missing diagonals, on the other side. The truss response to loading is similar to that of a “Vierendeel” beam, except that the coupling between the flanges is realized by means of axial forces within the diagonals instead of bending moments within the connecting members. Due to the relatively high bending stiffness of the cores and to the flexibility of the discontinuous bracing the spatial truss takes over only some 3/4 of the seismic action during an elastic seismic response.
An additional coupling element has arisen at the top of the tower cores from the walls existing over the height of the last two minaret stories, i.e. 9 m. They are aligned with both the external and the internal walls of the cores. The internal walls have been required to fix the summit tube in the minaret cores. These coupling walls were designed as composite concrete—steel and included steel trusses too, in order to control the concrete cracking during a strong earthquake and ensure a durable ductile be-
havior. The steel trusses within the external walls are connected with those existing in the core walls (see Figure 4). The steel trusses within the internal walls transfer the internal forces to the RC walls over the height of two stories.
The enlargement of the tower foot to the foundation foot required a stiff box over the height of the basement. This has been achieved by means of a grid of RC walls (Figure 7). They were designed to safely carry the vertical and horizontal forces induced at the tower base. The basement walls depicted yellow in Figure 7 are placed beneath the core walls.
The foundation is composed of a 3 m thick foundation slab and 64 “barrettes” (short sheet pile walls). The “barrettes” have a cross-section with the thickness of 1.2 m and the length of 7.2 m at the exterior of the foundation
Figure 6. Global effect of the façade bracing.
Figure 7. The RC walls within the basement (plan view).
slab and, respectively, 6 m at the interior (Figure 8). Their depth is 43 m.
4. Design Philosophy
The structural design of the minaret tower has been decisively influenced by the existing strong seismic risk and by the client requirement for a millennium lasting monument. The aseismic design of structures has been made in accordance with the performance criteria recommended by [3]. Accordingly, in order to ensure optimal energy dissipation during the design earthquake, one has to protect the “fragile” members by increasing their resistance whereas the “yielding” members have to possess a high ductility.
To ensure a best possible combination of load-bearing capacity and ductility the main structural elements, i.e. the cores, their façade bracing, the basement walls, the foundation slab and the “barrettes”, have been designed according to the following categories:
- Highly dissipative members (HDM), i.e. structural elements which will be the first to yield and hence have to possess a high ductility. They will dissipate the most part of the energy induced by a strong earthquake and will also act as “fuses” within the structure by topping the magnitude of the internal forces induced.
- Less dissipative members (LDM), i.e. structural elements which will suffer small, respectively moderate plastic deformation during the design earthquake.
- Elastic members (EM), i.e. structural elements which should remain elastic during the design earthquake. The load bearing capacity of these members should be so scaled, that the higher the risk of a fragile collapse the higher the existing resistance.
The necessary scaling of the members resistances is achieved by applying the method of “design capacity”. When two structural members are connected, the less ductile one should have a higher load-bearing capacity as the other.
Figure 8. Foundation system (plan view).
This design concept is outlined by the seismic structural response depicted in Figure 9. The magnitude of the internal forces induced by the design earthquake within the structure is dependent on the forces at which the highly ductile members start to yield. Each point on the curve corresponds to a member yielding. The lower the internal force at the first yielding is and the more HDmembers yield, the lower the total force induced by the seism. At the same time the earlier a HD-member enters his plastic range of behavior, the higher should be its ductility. The seismic performance of the structure depends crucially on the safety margin between the maximum horizontal displacement expected to be induced by the design earthquake (in Figure 9 denoted as “displacement demand”) and that which the structure is capable to undergo without attending a major disruption.
Figure 10 shows for which category the structural elements of the minaret have been designed. The chosen “fuses” for a strong earthquake are the façade bracings and the main floor coupling beams. In case of the façade bracings only the central parts of diagonals are designed to dissipate energy through plastic deformation. They are made of steel grade S235 which is very ductile and has less strength than the grade S355 used for the other steel profiles. In order to top the induced forces it has been required that the yield strength of S235 must not exceed 245 MPa. The dissipative parts of the façade bracing are bolted with the rest, so that they can be easily replaced after a very strong earthquake, if necessary.