The recent Turkey – Syria earthquake inspired us to write a blog post about the importance of seismic retrofit of existing buildings.

The current example – case study – employs a typical building built in Southern Europe in the late 1980s. The building has 4 floors of approximately 200 m2 each, and a soft storey at the ground level. The infills of the upper levels are relatively strong with good quality ceramic bricks and mortar of relatively high strength. The combination of the soft ground storey and strong infill panels in upper floors is the most important characteristic of the building and constitutes a serious structural problem related to its seismic behaviour.

The concrete grade is C20/25 the steel grade is S400, and in general there is adequate longitudinal reinforcement (for instance a typical rectangular or square column has 4Φ20+ 4Φ16 of rebars). The shear reinforcement however is just Φ8/20 for all the beams and Φ8/25 for the columns.

Figure 1: Front and back view of the building under consideration

Since the example is just for display purposes, the following simplifications have been made:

1. Checks will be carried out only in shear. This is a reasonable and realistic simplification, considering that in the vast majority of cases of such existing buildings, shear is the most critical check.
2. The checks will be done according to EC8 for a single Limit State, Significant Damage SD, which is probably the performance level mostly used for assessment purposes.
3. The building will be subjected to the two horizontal components of one of the strongest recordings from the recent Turkey-Syria earthquake of the 6th of February 2023 (Mw=7.8, record from station 3145). The record in the East-West direction has a peak ground acceleration of 0.69g and a maximum spectral acceleration of approximately 2.1g. In the North-South direction, the record has a peak ground acceleration of 0.59g and a maximum spectral acceleration of approximately 1.75g.

This record and hundreds of other recordings can be downloaded from the website of the Turkish Agency for Disaster and Emergency Management: https://tadas.afad.gov.tr/list-waveform

Figure 2: The acceleration, velocity and displacement time-histories in the East-West direction (screenshot from SeismoSignal v2023)

Figure 3: Response spectrum of the East-West component (screenshot from SeismoSignal v2023)

Figure 4: The acceleration, velocity and displacement time-histories in the North-South direction (screenshot from SeismoSignal v2023)

Figure 5: Response spectrum of the North-South component (screenshot from SeismoSignal v2023)

All the analyses and code-based checks have been carried out with SeismoStruct v2023 Release-1.

With eigenvalue analysis the fundamental periods in the X and Y directions are equal to 0.245 and 0.268 sec, respectively. The shape of the fundamental modes is characteristic of the weak storey, with very large deformations at the ground level and much smaller in upper floors (Figure 6).

Figure 6: Fundamental mode of the building in the Y direction (T=0.268 sec)

With dynamic analysis a maximum top displacement of 4.24 cm (t=7.02 sec) in the East-West direction is obtained, and the maximum top displacement in the North-South direction is equal to 3.10 cm (t=9.86 sec).

The top displacement indicates a total interstorey drift of 0.37% for the entire building. At first sight, this is a bit surprising, considering the significant weakness of the building at the ground level. Such low drift indicates light to very light damage (in general, significant structural damage is expected for drifts close to 1.0% or more). However, a closer look at the response on the local level gives a better insight. It can be observed that, whereas the total displacement at the top level (i.e. the total drift of all 4 storeys) is 4.2 cm, the deformation at the ground level only is 2.8 cm, meaning that 67% of the total deformation is concentrated at the soft ground storey. This leads to a drift of 0.93% at the ground floor, which indicates significant damage at the structural members.

Looking at the code-based checks (just in shear for the Significant Damage limit state, as mentioned before), the maximum demand to capacity ratio DCR is 1.70 for the columns and 2.42 for the beams. The value of the DCR is an indication of whether or not the member can sustain the imposed demand; with DCR>1 structural failure occurs, and with DCR<1 the member is safe. In the plots and the videos of this article the structural members (columns or beams) that have failed are depicted in blue. As expected, almost all the observed failures of the existing building are located at the ground level, confirming the fact that soft ground storeys are indeed an element of increased vulnerability. The time-history of the response and the structural failures of the building can be viewed in the following video .

Figure 7: Members failed in shear at the time step of maximum displacement in East-West direction

The retrofit scheme is applied in all the ground storey columns, which were strengthened with 10 cm wide shotcrete jackets. The jackets’ concrete grade is C25/30, and the steel grade is B500c. The longitudinal reinforcement of a typical rectangular jacket is 4Φ20+4Φ16, and the shear reinforcement of the members is significantly increased in all columns with stirrups of Φ10/10. Figure 8 shows indicative jacket layouts, as modelled in SeismoStruct. Some columns of the second floor are also strengthened with a single layer of FRP wrapping (read more about FRP Wrapping method).

Figure 8: Typical layout of the jacketed sections

Eigenvalue analysis indicates a significant increase in the stiffness of the building with a fundamental period of 0.206 sec, i.e. a considerable decrease of almost 25% with respect to the 0.268 sec of the initial building. What is more important is that the shape of the fundamental mode has changed with larger deformations at the upper stories, rather than at the ground floor (Figure 9).

Figure 9: Fundamental mode of the building strengthened with jackets (T=0.199 sec)

From the dynamic analysis (Figure 10 & Figure 11) we now observe smaller total deformations (a 37% decrease with respect to the initial building), significant increase of the stiffness at the ground level, larger deformations at the upper floors and overall an increased building capacity. Looking specifically at the shear code-based checks of Figure 11, very few component now fail in shear. What is more, in none of the columns is the demand larger than the capacity. The maximum DCR values in the columns is 0.90, and in the beams it is 1.75. The building has become considerably safer. The time-history of the response and the structural failures can be viewed in the following video.

Figure 10: Top displacement vs. time plot of the building strengthened with RC jackets

Figure 11: Shear checks at the time step of maximum displacement (building strengthened with jackets)

The retrofit is carried out with five new reinforced concrete walls at the perimeter of the building, which extend to the full height of the building. The reinforcement of the walls is similar to what would be found in shear walls (read more about new Reinforced Shear Walls method) in a new construction with pseudo-columns with a typical reinforcement of 8Φ20+16Φ16 with Φ10/10 stirrups. The walls are placed in symmetrical positions in the perimeter, in order not to introduce undesired torsional effects, and they are connected with the existing beams and the columns through a large number of dowels that are designed to transfer the seismic inertia forces from the weak building to the strong walls.

Figure 12: Front and back view of the building strengthened with new RC walls

Figure 13: Typical configuration of new RC walls

Eigenvalue analysis indicates that the strengthened building is significantly stiffer with respect to the initial building with a fundamental period of 0.142 sec (an impressive decrease of more than 45%). Furthermore, the eigenshape now has a cantilever-like shape, indicative of the presence of large shear walls in the construction.

Figure 14: Fundamental mode of the building strengthened with new RC walls in the Y-direction

Dynamic analysis shows an improved performance (Figure 15 and Figure 16). There is low amplitude and high frequency motion throughout the time-history, and the top maximum displacement has dropped to just 1.54 cm.

Figure 15: Top displacement vs. time plot of the building strengthened with new shear walls

Figure 16: Shear checks at the time step of maximum displacement (building strengthened with new RC walls)

Looking at the shear code-based checks of dynamic analysis of Figure 16, there are very few component failures, these are only beams and these have much smaller DCR values (maximum DCR = 1.11). The time-history of the response and the structural failures can be viewed in the following video .

Another interesting observation, which explains the good performance of the strengthened building, is that the walls at the ground level undertake over 75% of the base shear. This means that the existing members have to withstand less than 25% of the total seismic demand, and this is why they easily pass the required checks. This example shows us that, when large shear walls are added in an existing building, it is very common that no other intervention is required in the other structural members, provided of course that these are in relatively good condition, and possess some, non-negligible, existing reinforcement and lateral strength.

The retrofit is carried out mainly at the ground level: X-shaped steel braces are added at 6 different locations in the perimeter of the building and FRP wrapping is applied to all the ground storey columns. Moreover, some (not many) columns of the upper storeys that were proven problematic from the analysis are also upgraded with FRP fabrics. For the braces, 120x120x8 hollow rectangular sections are employed, which are pinned to steel plates that are fixed at the corners of each concrete bay with steel anchors and epoxy adhesives. For the FRP wrapping, 1 to 3 layers of a relatively strong FRP fabric are employed (dry fibre thickness approx. 0.33 mm, area density of the carbon fibres approx. 600 g/m2 and dry fibre tensile strength approx. 3,800 N/mm2). Similarly to the case of RC walls, the braces should be placed as symmetrically as possible to avoid unwanted torsional effects.

Figure 17: Front and back view of the building strengthened with steel braces

Eigenvalue analysis indicates a moderate increase in the stiffness of the building with a fundamental period of 0.234 sec, but now the shape of the fundamental mode has changed, and it has uniformly distributed deformations along the height, rather than large deformation concentrations at the ground level. It is noted that the increase in stiffness that can be achieved with steel braces is generally less than that with new RC walls. What is important however is that the stiffness of the braces can be easily calibrated (i.e. changing the cross-sections of the braces), in order to avoid irregularities in elevation, when the strengthening is not applied to the entire building height, as in this case.

Figure 18: Fundamental mode of the building strengthened with steel braces and FRP Wrapping (T=0.234 sec)

From the dynamic analysis we are able to see the following: increase of the stiffness at the ground level, increased building capacity and fewer failures in the structural elements (Figure 19 and Figure 20). Furthermore, the members that fail are only beams. The maximum DCR at the columns of the previously weak ground floor is less than 1.00 (0.94), indicating a much safer building with respect to the initial unstrengthened configuration. The time-history of the response and the structural failures can be viewed in the following video.

Figure 19: Top displacement vs. time plot of the building strengthened with steel braces and FRP wrapping

Figure 20: Shear checks at the time step of maximum displacement (building strengthened with steel braces and FRP wrapping)

In the current example 18 isolators are placed at the ground level, below the 18 columns of the ground floor, at the locations where the columns are connected to the footings (cutting the columns is required for this). In order not to allow large relative horizontal deformations between the isolators, two diaphragms are constructed, one at the foundation level (connecting all the individual footings) and one above the level of the isolators (connecting the 18 columns).

Figure 21: Location of the construction of the two diaphragms (building strengthened with seismic isolation)

With eigenvalue analysis a very large change in the dynamic characteristics of the structure is observed, as both the fundamental period (2.57 sec from 0.268 sec) and the first mode shape shown in Figure 21 are very different. Almost all the deformations are now concentrated at the isolators with the rest of the building remaining undeformed (Figure 22).

Figure 22: Fundamental mode of the building strengthened with base isolation (T=2.57 sec)

Moreover, the isolated building shows also a completely altered dynamic behaviour (Figure 23), which is not unreasonable given the completely different dynamic characteristics of the new building. Although the maximum top displacement has increased significantly with respect to the initial building, it is very impressive that more than 95% of the total deformations are now concentrated at the isolators, leaving the upper structure almost undeformed.

Figure 23: Top displacement vs. time plot of the building strengthened with base isolation

Due to the significant period elongation caused by the isolators, the shear demand at the ground storey columns is now less than 20% of the demand on the same columns of the original building! Consequently, the structural members are now able to effectively resist the seismic action with small DCR ratios; the maximum DCR for the vertical members being equal to 0.62, significantly smaller than unity. The time-history of the response and the structural failures can be viewed in the following video.

Figure 24: Shear checks at the time step of maximum displacement (building strengthened with base isolation)

The main technical observation made from the case study is that with all the methods the seismic performance of the building has significantly improved. The demand-to-capacity ratios (DCRs) decreased considerably, especially for the more critical vertical members, and the vulnerability of the structure has decreased considerably.

A rough estimation of the cost of the interventions is between 4%-8% of the total building value for all the methods, with the exception of base isolation, which is estimated at around 15% of the building’s value. Obviously these estimates can vary significantly depending on the seismic risk of the region, the labour and the material costs and the location of the building (average costs for materials and labour in Southern Europe have been considered).

These values have serious implications for our decision on whether or not to upgrade a structure. Essentially, it reminds us that, although a large percentage of the population of the planet lives in highly vulnerable buildings in high earthquake hazard zones, we are often reluctant to pay just a small fraction of the total building value to upgrade them, to make them safer and to save a significant amount of money in the case of a large seismic event that could lead to extensive damage or even to collapse (not to mention the possible loss of human lives, which, as indicated by the recent earthquakes in Turkey, can easily rise to dramatic levels).

NOTE:
The SeismoStruct input file for the analyses can be downloaded from this link . All the analyses and code-based checks have been carried out with SeismoStruct v2023 Release-1.

Keywords: seismic retrofit, structural strengthening, reinforced concrete jackets, FRP fabrics, FRP wrapping, Steel braces, base isolation, structural assessment, existing buildings