1. Introduction
Two major earthquakes that occurred independently of each other on 6 February 2023, caused great destruction in Türkiye, which is placed in a region of very high seismic hazard. The first earthquake was Mw = 7.7 in the Pazarcık (Kahramanmaraş) at 04:17 local time, while the other earthquake occurred at 13:24 local time in the Elbistan (Kahramanmaraş) with a magnitude of Mw = 7.6. There is a distance of approximately 90 km between the epicentres of two earthquake pairs that occurred nine hours apart. The successive occurrence of earthquakes caused an increase in the loss of both lives and property. More than 500,000 structures suffered structural damage at different levels while more than 50,000 people died. Both earthquakes significantly affected 11 different provinces such as Hatay, Kahramanmaraş, Adıyaman, Malatya, Adana, Osmaniye, Kilis, Şanlıurfa, Gaziantep, Osmaniye, and Diyarbakır. Many different types of structural systems were damaged at various levels such as heavy, moderate, and slightly, or collapsed. Many structures have survived despite the destructive nature of the first earthquake. However, after the second earthquake that happened on the same day, it achieved a higher level of damage and increased the extent of the destruction.
Every devastating earthquake experienced brings the consequence of lessons and precautions to be taken in terms of both earthquake and civil engineering. Detection of structural damage after an earthquake is one of the important phases of advanced post-disaster management. Regarding spatial planning and urban change, the information obtained from damage assessments is crucial [
1,
2,
3]. These and similar data can be used in earthquake zoning studies and the development of earthquake-resistant building design principles [
4,
5]. This study focused on how the Kahramanmaraş earthquake couple, which occurred on 6 February 2023, affected adobe structures.
In general, reinforced concrete structures are the dominant structural type of the existent building stock in urban regions in 11 different provinces affected by the earthquake. In rural areas, masonry structures built with local craftsmen and construction techniques using different types of materials are the dominant building stock. Although these structures in rural areas vary regionally, they are generally built using rubble stone, adobe, brick, briquette, or cut stone. By increasing the surface areas of the load-bearing walls, which form the structural system of masonry structures, the bearing capacity is increased. The reason for this is that the compressive strength of the wall material used is higher than the tensile strength. However, while these elements can respond to high compressive forces, they are not resistant to shear and bending effects [
6,
7,
8,
9,
10,
11]. The higher compressive forces that will occur depending on the magnitude of the earthquake cause damage at different levels. In addition, structural weaknesses, poor workmanship, and failure to apply earthquake-resistant building design rules directly affect the damage levels.
There are many studies on the damage and damage levels of adobe and other types of masonry structures after earthquakes. Valente [
12] investigated the earthquake behaviour of two historic masonry palaces with corner towers by making three-dimensional structural analyses. Numerical structural analyses revealed that the damage distributions and seismic response are highly dependent on the dynamic and geometric features of the structures. In the North-East of Italy, Valente and Milani [
13] attempted to predict the behaviour of eight medieval masonry towers using the finite element method using static and non-linear dynamic analysis methods. It has been determined that the geometrical properties of the structures affect the behaviour of the structures with the influence of seismic actions. After the 2003 Valle Scrivia earthquake, Ruggieri et al. [
14] examined the current state of 20 masonry churches in Piedmont. They analysed the behaviour of these churches under the influence of the earthquake. Statistical damage and vulnerability analyses were made using peak ground acceleration. They stated that it is possible to make priority lists for retrofit interventions through the methodologies they propose. Valente and Milani [
15], created three-dimensional models and examined them with the finite element method in order to reveal the earthquake-induced damage assessment and partial collapse mechanisms of an Italian medieval castle. With their numerical analysis results, they defined damage models for different seismic intensity levels and the most vulnerable parts of the structure. Uva et al. [
16] evaluated the structural vulnerabilities of masonry churches with the modern Internet of Things. They propose an application to address the damage measurement problem to evaluate the seismic sensitivity of historic masonry structures. Valente [
17], used the finite element method to evaluate earthquake response and damage models of two historic masonry churches with bell towers in Northern Lombardia (Italy). They checked whether the structure was sufficient for different degrees of seismic effects by finding the most critical sections in the structure. Işık et al. [
18] investigated the masonry damages in Adıyaman province after the 2023 Kahramanmaraş earthquakes. They stated that weak structural features are the primary reason for the structural damage. Milani and Valente [
19] investigated the failure mechanisms of masonry churches after the 2012 Emilia-Romagna (Italy) earthquake. They stated that it is useful to do various structural analyses to have a complete understanding of the parts of the buildings that are most likely to be partially destroyed. Milani [
20] investigated the limit analysis of three masonry churches in Italy after the 2012 Emilia earthquake. They compared the regulations in the country and made strengthening suggestions. Işık et al. [
21] also investigated the damages in minarets and mosques in Adıyaman after the 2023 Kahramanmaraş earthquakes. It was determined that the damages intensified especially at the transition points in the minarets that did not receive any engineering service. Kocaman [
22] investigated the damages in masonry mosques and minarets after the Kahramanmaraş earthquakes. Bilgin et al. [
23] investigated the masonry construction damages following the 2019-Durres (Albania) earthquake. Işık et al. [
24] made a case study on the damage rating of masonry structures using the EMS-98 after the 2020-Turkey (Elazig) earthquake. Dizhur et al. [
25] researched the damage to un-reinforced clay brick and stone masonry buildings during the 2010 New Zealand (Darfield) earthquake. Penna et al. [
26] investigated the seismic performance of masonry structures after the 2012 Italy (Emilia) earthquake. Bayraktar et al. [
27] investigated the damage mechanisms in masonry structures after the 2004 Türkiye (Ağrı) earthquake. Sorrentino et al. [
28] investigated the earthquake behaviour of ordinary masonry buildings after the 2016 Central Italy earthquakes. Hafner et al. [
29] made a post-earthquake evaluation of historical masonry structures during the 2020 Zagreb (Croatia) earthquake. Yoshimura and Kuroki [
30] specifically examined the damages in unreinforced adobe and brick buildings after the 2001-Elsalvador earthquake. Najafgholipour et al. [
31] investigated the structural damages in masonry structures after the 2017 Iran (Sarpole Zahab) earthquake. Vlachakis et al. [
32] conducted a study on the lessons that can be drawn from the damages in masonry structures during the 2017 Greece (Lesvos) earthquake. Miano et al. [
33] developed seismic vulnerability curves for existing masonry structures following the 2016 Amatrice earthquake based on Copernicus damage maps using the EMS-98 damage scale. Each of these studies is a case study, and the main results presented in the study provide valuable information for better understanding the seismic response and vulnerability of the structures and for use in the assessment and mitigation of seismic risk.
There are some special studies on the behaviour of adobe structures before and after the earthquake. Sumerente et al. [
34] conducted a study on the combined evaluation of in- and out-of-plane fragility functions for adobe buildings in the Peruvian Andes. Ahmad et al. [
35], investigated in detail the behaviour of confined adobe structures under earthquake effects. Greco and Lourenço [
36] used advanced numerical structural analysis to evaluate the overall behaviour of adobe buildings and to determine the failure mechanisms and damage distribution. Tarque et al. [
37], in light of recent earthquakes, investigated the seismic behaviour of adobe masonry structures under the influence of earthquakes. Ramirez et al. [
38] analysed the earthquake behaviour of the adobe buildings based on parametric and machine learning after the 2017 Mexico earthquakes. Varum et al. [
39] examined the structural behaviour and strengthening of the adobe structures. Sayın et al. [
40] investigated the structural damages in adobe and masonry structures during the 2011 Elazığ earthquake within the framework of engineering. Kiyono and Kalantari [
41] investigated the failure mechanisms of masonry and adobe structures after the 2003 Bam (Iran) earthquake. Webster and Tolles [
42] investigated the earthquake damage in historical adobe structures after the 1994 Northridge earthquake. Sayın et al. [
43] examined the damages in adobe and masonry structures after the 2011 Van (Türkiye) earthquakes within the framework of a cause-effect relationship.
Each of these and comparable studies can be thought of as case studies, and they can help with the design and strengthening of masonry structures, as well as the development of earthquake-resistant building design concepts for such structures. In addition, some studies can be seen as field studies for seismic risk assessments of such structures before a possible earthquake.
In this study, the effects of the 2023 Kahramanmaraş earthquake couple, which was the worst disaster of the century for Türkiye, on adobe structures were examined in detail. A significant part of the dominant masonry building type in rural regions consists of adobe structures. In the study, first of all, the studies in which the damages occurred after earthquakes in masonry structures and especially in adobe structures after different earthquakes were examined were examined. By giving information about the Kahramanmaraş earthquakes, the highest PGA values measured at the earthquake stations in 11 provinces most affected by earthquakes are given. It was converted to earthquake intensity for each earthquake station using these measured PGA values, with the help of the PGA-intensity relation, which is especially recommended for Türkiye. It has been tried to reveal the destructive features of the earthquakes with the structural damage data in 11 different provinces affected by the earthquakes. Observed structural damages at different levels in adobe buildings after earthquakes have been investigated within the framework of the cause-effect relationship in terms of civil and earthquake engineering. Moreover, damage levels were determined by using EMS-98 for 100 adobe structures as a result of these earthquakes. A comparison was made for adobe structures within the framework of the rules specified in the last two earthquake design codes used in Türkiye. This study aims to reveal the extent of structural damages caused by two independent earthquakes. This will be one of the first studies to look into in detail the damages caused to adobe structures by the Kahramanmaraş earthquakes, which caused great destruction. With this study, it can be used to improve earthquake design principles related to adobe structures. This, and comparable research, can be a useful tool for decision-makers in post-disaster studies, which are a part of modern disaster management.
2. 6 February 2023 Kahramanmaraş (Türkiye) Earthquakes
When the tectonic system in Türkiye is examined, it can be seen that the Arabian Plate moves to the north the Eurasian Plate compresses Eastern Anatolia, and the Anatolian Plate is affected by the North Anatolian Fault (NAF) and Eastern Anatolian Fault (EAF) formed after the resistance of the Eurasian Plate in the north. It is seen that it moves westward (
Figure 1). The EAF, with an average length of 580 km, is one of the most seismically active regions of Türkiye and many major earthquakes have occurred along this fault zone. The EAF is a left-lateral strike-slip fault that forms the southeast border of the Anatolian plate and intersects the NAF at Karlıova. Similarly, EAF merges with the Dead Sea Fault (DSF) around Antakya [
44,
45,
46,
47,
48]. The North and East Anatolian Faults (NAF and EAF) have caused thousands of people to die in the last 150 years, with earthquakes of more than 7 magnitudes they have produced. In short, the earthquake activity in and around Türkiye is directly related to the complex plate tectonics that continues between the Eurasian, Arabian, and African plates.
Firstly, an earthquake happened on 6 February 2023, at 04:17 local time, with a magnitude of M
w = 7.7 in Kahramanmaraş (Pazarcık). After 9 h, at 13:24, an earthquake happened with a magnitude of M
w = 7.6, where the epicentre was the Kahramanmaraş (Elbistan). Another earthquake with a magnitude of M
w = 6.4 happened in Hatay (Yayladağı) on 20 February 2023, at 20:04 local time. Approximately 14% of the country’s land area has seen severe destruction as a result of these earthquakes. In terms of severity and geographic coverage, these earthquakes are the worst recent disasters ever in Türkiye. More than 50,000 people died and more than 500,000 buildings were damaged as a result of these earthquakes. The communication and energy infrastructures were also damaged and created significant economic losses for the country. The second earthquake’s focal depth is 7 km, compared to the previous earthquake’s focal depth of 8.6 km. Since both of the earthquakes are very close to the surface and in terms of time interval, consecutive earthquakes have affected the damage levels very significantly. The earthquake region is shown in
Figure 2.
Spectral acceleration values were compared for Kahramanmaraş province, which is the epicentre of both earthquakes. The spectral accelerations that were recorded at strong ground motion stations in 17 different locations in the province are shown in
Figure 3 according to different local soil classes. In the Türkiye building earthquake code, six different local soil classes are expressed as ZA, ZB, ZC, ZD, ZE, and ZF. Soil properties weaken as one goes from ZA to ZF. Accordingly, for all soil types at seven locations, the projected acceleration values were surpassed. In ZC and ZD soil types, the number of stations exceeding the expected acceleration value was 10.
The second earthquake M
w = 7.6 magnitude, which happened 9 h after the first earthquake, was recorded in a smaller number of strong ground motion stations located in 11 different locations in Kahramanmaraş than the previous earthquake, probably due to its further distance. Spectral accelerations obtained from the acceleration values are shown in
Figure 4 according to different local soil classes. Accordingly, the expected acceleration values were exceeded for all local soil classes at one station. In ZC and ZD soil types, the number of stations whose acceleration value exceeded the expected acceleration value was 2.
The measured peak ground acceleration (PGA) at the stations located in 11 different provinces in the earthquake region was given shown in
Table 1.
The highest PGA was measured for Pazarcık (Kahramanmaraş) in the first earthquake, while in the second earthquake, the highest PGA was measured in Göksun (Kahramanmaraş).
In this study, PGA-intensity transformations were made by using the intensity—PGA relation recommended by Bayrak [
52] for Türkiye. This correlation is as follows;
where I denotes the intensity of the earthquake and PGA denotes the peak ground acceleration (cm/s
2).
According to the measured PGA values, the earthquake intensity in the provinces was obtained by using Equation (1), and the results of the obtained intensities are shown in
Table 2.
The highest intensity was obtained in Pazarcık (Kahramanmaraş) with XI, while the lowest intensity was obtained in Centre (Kilis) with VII for the first earthquake. The highest intensity was obtained in Göksun (Kahramanmaraş) with X, while the lowest intensity was obtained in Akçakale (Şanlıurfa) with VI for the second earthquake.
Table 3 lists the overall number of affected structures by the earthquakes across 11 provinces.
Damage assessment studies were completed for 1,712,182 buildings in 11 provinces affected during these earthquakes by the Türkiye Ministry of Environment, Urbanization, and Climate Change as of 6 March 2023. The classification of damages is given in
Table 4. The independent section can also be considered as the name given to each of the areas in the building that can be used alone and suitable for use. Apartments, business offices, stores, and shops belonging to the building are included in the independent parts.
According to the result of the damage assessment study carried out by the relevant Ministry, the total number of residences that fall into the categories of emergency demolition, demolition, or severe damage was determined as 518,009. Damage levels by province are shown in
Table 5.
Total housing damage calculated in terms of urgently demolished, heavily damaged, or demolished houses was 54.7 billion dollars [
53]. An aerial view of the destruction size of earthquakes in Hatay city centre is shown in
Figure 5.
3. Structural Damages in the Adobe Buildings in Earthquake-Prone Region
The dominant building stock type is masonry structures in the rural areas of 11 different provinces affected by earthquakes. Most of these masonry structures are in unconfined masonry type. In addition to the confined masonry structures, reinforced concrete structures are also encountered. Many of the buildings in rural areas were built by local labour and resources using traditional construction methods. Therefore, the earthquake vulnerability of such structures is high. The wall materials used are joined to each other with the help of a mortar. Cement-sand-lime mortar is the type of mortar that is frequently used in the area. When the cement was hard to come by and in places where it existed, a sand-lime mortar was employed. Most of the mortar used in adobe and extremely old masonry structures was earthen.
The masonry buildings located in the rural regions affected by the earthquake were built as one-storey or two-storeys using local craftsmen and materials. The lower floors of the two-story buildings are used as barns or storage areas. Generally, adobe, rubble, cut stone, brick, or briquette are used in buildings built in a masonry style. Structural walls are created using cement or earthen mortar. The strengths of both the wall material and the joining mortars remain quite low. Due to these properties, the structural wall thicknesses take large values. Therefore, in these load-bearing walls, which are planar, there are pressure forces transferred from the top to the supports in one direction. In such masonry systems, the structural members and the space dividing or covering elements overlap. These elements, whose compressive stress values are relatively higher than tensile stresses, are not resistant to bending and shear effects [
54,
55,
56,
57].
The materials used in masonry structures in rural areas where earthquakes are effective vary regionally. Some of these structures are adobe structures. In general, adobe buildings with flat earthen roofs were preferred because adobe structures were easy and cheap to build and allowed the production of grain. One of the earliest construction materials is adobe, which is made by combining soil, straw, and water, pouring it into molds, and then drying it in the sun and shade alternately. The hand-kneading and preparation of the clay is the first step in the creation of adobe. The process of shaping it in molds and letting it dry in the sun makes up the second stage [
58,
59,
60]. It is a material that is indispensable, especially for rural areas, has the least cost, does not require the establishment of a production facility, and has a high thermal insulation value [
61,
62]. The obtained adobe blocks are transported to the construction site and at the same time, these blocks must be built to form a wall. Sections of traditionally built adobe structures are shown in
Figure 6.
The fact that adobe buildings were built without taking into account the earthquake-resistant building design principles and the low strength properties of the adobe material caused the earthquake resistance of such structures to be quite weak. The ductility capacity of mudbrick structures is quite low. For this reason, the ability of such structures to consume the energy that will occur during an earthquake by displacement without collapse is weak. These types of structures, which are constructed without any engineering service, cannot meet the earthquake forces and reach the collapse mechanism. Examples of completely collapsed adobe structures are shown in
Figure 7.
In masonry structures, there may be cracking or partial collapse of the structural walls after an earthquake. The low tensile strength of the wall material used in adobe structures and the low shear strength of the mortar, heavy earthen roof, and partially inadequate element connections cause partial collapses. In this case, the structure becomes completely unusable. The examples of adobe structures with partial collapse are shown in
Figure 8.
In field observations, it was determined that adobe and a different wall material were used together in the construction of load-bearing walls in some adobe structures. During the construction of adobe structures, a compatibility problem between the materials was observed in the use of different types of wall materials. Insufficient interlocking and different material strength properties between different materials caused separation damage between layers or at the junction of two materials. The low shear strength of the mortar used in order to use two different materials together caused the separation between the layers to be greater. Such damages are shown in
Figure 9.
Roofs in adobe structures, as in other masonry structures, are usually built as a result of using heavy earthen roofs and wooden beams together. In addition, as a result of adding new soil layers to the soil roofs that are worn out due to climatic conditions over time, the old layers become compressed, and their weight increases. Due to the increased weights, significant structural damage occurs both on the soil roofs and on the load-bearing walls on which the roofs are supported [
63,
64,
65]. In addition, insufficient interlocking between the load-bearing wall and the heavy earthen roof interrupts the load transfer and causes such damage. As a result of the collapse of these roofs, the structures have become unusable. The completely collapsed heavy earthen roofs after the Kahramanmaraş earthquakes are shown in
Figure 10.
The bearing capacity of the wooden beams used in the roof has been exceeded due to heavy earthen roofs. Wooden beams collapsed completely because they did not have sufficient support with both the heavy earthen roof and load-bearing wall, while the ones with sufficient support were damaged according to additional loads and passed before collapse (
Figure 11).
Heavy earthen roofs, together with the vertical acceleration component of the earthquake, push the load-bearing walls out of the plane/inside under the effect of horizontal load. In this case, these walls, which do not have sufficient connection with the roof and whose rigidity is weak in the in-plane/out-of-plane direction, experience sudden power depletion and collapse on the walls together with the roof. Different types of damage to heavy earth roofs are shown in
Figure 12.
Another type of observed damage in adobe buildings is splitting at the corner points and partial collapse of the corner points. These damages are due to insufficient support of the load-bearing walls to the structural element to which they are attached in all three directions. In addition, the high weight of the soil roof causes a decrease in the strength at the corner points under horizontal loads. In addition, due to the weak workmanship at the corner points, the earthquake force coming in two directions at the corner points causes the walls to repel each other and change their directions [
66]. Separation damages at the corner points of adobe structures are shown in
Figure 13. Out-of-plane structural wall damages at corner points are shown in
Figure 14. The out-of-plane collapse damages on the outer load-bearing walls have increased significantly due to the lack of adequate jointing between the outer-inner load-bearing walls that make up the adobe structures, the lack of support detailing, and the effects of heavy earthen roofs. The observed out-of-plane wall damages in adobe structures are shown in
Figure 15.
As in other building types, the strength/rigidity differences between storeys in buildings with more than one storey affect the relative storey drifts negatively. Examples of damage at different levels due to strength/rigidity differences between floors are shown in
Figure 16.
Failure to provide load transfer on the structural walls of adobe structures, which were constructed using local materials and workmanship with traditional construction technologies, without paying attention to earthquake-resistant design rules, caused various levels of damage to the walls. In addition, poor masonry workmanship, and not using horizontal and vertical beams that should be used also negatively affected the level of damage. Poor quality and insufficient mortar, poor bonding between walls and layers directly affected damage levels. Different examples of damage caused by these reasons are shown in
Figure 17.
Window and door openings prevent the load-bearing walls from properly transferring shear and bending stresses [
67]. This type of damage has occurred in many structures because both the cavities and the hollow walls do not meet the support conditions specified in the earthquake regulations. Visuals of different types of damage occurring in the gap regions are given in
Figure 18.
The horizontal and vertical wooden beams to be used in adobe buildings have increased the earthquake resistance of the adobe structures.
Figure 19a shows the out-of-plane movement of the structural walls in the region where these lines are not used.
Figure 19b shows the crushing damage at the base of the building. Gable wall damage is similar to damage in cases where there is insufficient connection between the structural walls and roof. An example of roof gable wall damage is shown in
Figure 19c.
In adobe masonry structures, in-plane/out-of-plane movements of structural and non-bearing members cause possible loss of life due to different reasons within the structure. Examples of damage that may cause possible loss of life are shown in
Figure 20.
Although significant plaster spills were observed in the adobe structure surrounded by reinforced-concrete elements added later, the structure provided life safety performance comfortably. An example of a besieged mudbrick structure is shown in
Figure 21.
4. Damage Classification in Adobe Buildings
The first damage assessments to be made after the earthquake should be made practically and quickly as possible. Due to the need for rapid assessment of damage, the magnitude of the damage, and the shortage of personnel who are not sufficiently specialized in their field, this is not possible in earthquakes where large-scale damage has occurred. Difficult terrain conditions, climatic conditions, and insufficient public resources after the earthquake also negatively affect this process.
In addition, damage information on buildings is essential for search and rescue, humanitarian aid, and reconstruction operations in earthquake zones. The damages in buildings can be classified in the field using damage scales [
68]. In this study, the European Macro-Seismic Scale (EMS-98), which is widely used after earthquakes in different parts of the world, was used for damage classification of the adobe structures. The EMS-98 was developed by the European Commission of Seismology (ESC), taking into account the extensive damage levels [
69,
70,
71]. Damage classification of masonry structures in EMS-98 is shown in
Figure 22. Since all the structures examined are adobe structures, they were chosen as vulnerability class A. The selected adobe structures were chosen to reveal various types of damage.
Figure 23 shows the damage examples in which there is no structural damage in the adobe structures studied, only minor plaster spills and a small part of the loose mud bricks on the upper part of the building have fallen.
Structures with slight structural damage and moderate damage in non-structural elements were determined as Grade 2. Examples of adobe structures with large plaster spills observed at this level of damage are shown in
Figure 24.
In Grade 3, it states that there is significant heavy damage in the building. At this damage level, structural elements have moderate damage, while non-structural elements are heavily damaged. Examples of adobe structures with large and wide cracks on most of the walls are shown in
Figure 25.
The level at which heavy damage occurs in structural elements and very severe damage in non-structural elements is specified as Grade 4. At this level of damage, very large separations on the walls and partial breaks in the roof and floors are observed. Examples of adobe structures belonging to this damage level are shown in
Figure 26.
A partial or total collapse of any structure as a result of severe structural damage is defined as the 5th grade. Examples of adobe structures partially or completely collapsed as a result of earthquakes are shown in
Figure 27.
The classification of damage levels for the 100 adobe structures in rural areas of Hatay, Kahramanmaraş, and Adıyaman is listed in
Table 6.
Approximately 25% of the adobe buildings examined were partially or totally collapsed, 30% of them were very severely damaged and 19% of them were severely damaged. Therefore, 74% of them have become unusable. It can be said that 26% of them showed sufficient resistance against earthquakes. This is a clear indication of how low the earthquake resistance of adobe structures is. It also indicates that such structures are not built in accordance with earthquake-resistant building design rules. These results showed once again that the earthquake vulnerability of other masonry structures, especially adobe masonry structures, is very high. The main reason for this is that adobe structures do not obtain any engineering services. In this context, it will be beneficial to apply the building inspection system applied for the urban building stock in such buildings.
5. Evaluation of Investigated Adobe Structures within the Scope of Earthquake Codes
The last 5 earthquake codes used in Türkiye include design rules for masonry structures. In the last two earthquake codes (2007 and 2018) [
72,
73], there is a general section as “Masonry Structures”. However, in the previous 3 regulations (1968, 1975, and 1998), it was included with the title “Masonry Buildings”. Although these three regulations (1968, 1975, and 1998) each have sections for adobe buildings were not specifically mentioned in the last two regulations. While the masonry building type is not mentioned in the 1998 and 2007 regulations, four types of masonry buildings are mentioned in the 2018 regulation. These are specified as reinforced, unreinforced, reinforced panel systems, and encircled masonry buildings, respectively. In 1968, 1975, and 1998 regulations, simplified geometric rules were envisaged in case calculations could not be made. In the 2007 regulation, it was stated that only adobe masonry buildings can be constructed with only one storey at most in totally earthquake regions, without counting the basement. In the same regulation, it is stated that the height of a single story can be a maximum of 2.70 m in adobe masonry structures, and the height of the basement can be a maximum of 2.40 m if it is built. The regulations also specify the rules regarding horizontal and vertical lines that should be used. In addition, in the last two earthquake codes, strength rules such as controlling the normal and shear stresses that happen under the joint effect of horizontal and vertical loads on the structural walls in masonry buildings were introduced. Again, in these two regulations, keeping the window and door spaces limited, opening window and door spaces at least 1.0~1.5 m from the corners of the building, is extremely important for the earthquake behaviour of the masonry building. As can be seen from the visuals of the buildings examined, although the regulations require the use of reinforced concrete horizontal beams, these horizontal beams are not used. While there are no horizontal beams in some structures, in some structures these beams are made of wood. Masonry buildings should be constructed with reinforced concrete vertical beams at the corners, in the vertical cross-sections of the load-bearing walls, and on each side of the door and window spaces to strengthen their earthquake protection. In the regulation, floors of masonry buildings should be reinforced concrete slabs or toothed floors with dimensions and reinforcements designed according to the rules that are given in TS-500. These rules were not followed in most of the masonry structures that did not receive any engineering and were the subject of the study.
In the 2007 regulation, masonry structures in the first-degree earthquake zone can be built with a maximum of two storeys. The number of storeys in the considered adobe structures corresponds to the number of storeys given in the regulation. In the 2018 regulation, the number of storeys is not used, instead, the building height class (BYS) is expressed, and the total allowed building height is foreseen as 7–10.50 m. The adobe masonry structures examined provide this situation. It is stated that cement-lime mortar (cement/lime/sand volume ratio = 1/2/9) or cement mortar (cement/sand volume ratio = 1/4) should be used on load-bearing walls. Cement-added mortar was not used in any of the structures examined.
There are adobe structures that have not been damaged due to different reasons in the earthquake region. However, it is a known fact that the earthquake performance of adobe and other masonry structures in rural areas is quite low and insufficient. Therefore, it is a necessity to apply the earthquake-resistant design rules of masonry structures with precision. The application of the building inspection system applied in Türkiye for urban building structures to such buildings will solve this problem. With the help of this system, damage levels can be minimized with the necessary controls during the design and construction stages of adobe and other masonry structures.
6. Conclusions
Türkiye suffered great destruction on 6 February 2023, with the effect of two independent major earthquakes, the epicenters of which were Pazarcık (Kahramanmaraş; Mw = 7.7) and Elbistan (Kahramanmaraş; Mw = 7.6). In this study, the effects of the earthquake couple, which were effective in 11 different provinces, on the adobe structures were tried to be examined in detail. After giving information about both earthquakes, the spectral accelerations measured at earthquake recording stations and suggested for different soils were compared. It was determined that the spectral accelerations measured in both earthquakes exceeded the recommended values. In this study, earthquake intensity was obtained for 11 provinces by using the intensity-PGA conversion recommended for Türkiye. In both earthquakes, the highest intensity values were obtained for the province of Kahramanmaraş, which is the epicenter of the earthquakes. As a result of the observational evaluations made on the site for adobe structures, it has been revealed once again that the earthquake vulnerability of such structures is quite high. Damages of different types and levels in the investigated adobe structures were evaluated within the framework of cause-effect relationships in terms of earthquake and civil engineering. The damage caused by the adobe buildings examined can be sorted as follows:
The strength properties of the adobe material are very low,
Generally, loamy soil or lower-strength binders are used as joining mortar,
The construction of such structures using local materials, workmanship, and construction technologies without any engineering service,
Not using horizontal and vertical beams,
Insufficient support of load-bearing walls
Insufficient connection between heavy earth roof and structural walls
Usage of heavy earthen roof
Insufficient clamping of load-bearing walls in both directions
Usage of different types of structural wall material,
Insufficient connection between different structural walls layer
Poor workmanship
The fact that the earthquakes that occur are very close to the surface and large
In this study, damage classification was also carried out for 100 different adobe structures. All the results obtained showed that the earthquake resistance of adobe structures is very low.
As in other masonry structures, the earthquake vulnerability of adobe structures is quite high. The masonry building stock contains many structural uncertainties and emerges as a building type with high damage potential. For all these reasons, new buildings to be built in rural areas must be constructed and inspected in accordance with the regulations. In this context, it has become a necessity to examine the existing building stock in the regions affected by the earthquake in detail and to take the necessary precautions. It should be ensured that the rules in the earthquake regulations are applied in a controlled manner during the design and construction phase of all masonry structures to be built. Rules related to adobe structures that are not included in the last earthquake regulation should be determined and added to the earthquake code. It is considered that adobe material is friendly for environmental and sustainable construction. One of the most important causes of damage in adobe structures is the use of heavy earthen roofs. Although the use of heavy soil roofs is prohibited by regulation, they continue to be used. This should be prevented in new buildings by making necessary inspections.
This study can be used as a resource to improve the behavior of such structures in future earthquakes, especially in studies on adobe structures. In Türkiye, as in other countries, earthquake-resistant building design rules are renewed and updated over time. However, it is clear that these rules will be effective if they are fully implemented during the project and construction phase. The building control mechanism, which is actively used in urban areas in Türkiye, should also be implemented in rural areas. Rather than strengthening existing adobe structures that have been damaged at different levels, demolition seems to be a more realistic solution. Necessary structural analyses can be made for existing adobe structures that are not damaged, and strengthening or demolition decisions can be made about these structures. The adobe structures with low earthquake performance and no monumental value should be demolished first and new structures should be built according to the earthquake regulations. If there are adobe structures whose earthquake performance will reach a sufficient level after repair and strengthening, engineering applications that are foreseen in the earthquake regulations for the strengthening of masonry structures or that will not conflict with these methods can be used. Here, economical solutions should be selected by comparing the retrofitting with the reconstruction cost.