An Analysis Procedure for the Surviving Rural Architecture, Built with Raw Earth Mortars, in the Amatrice Area (Italy) as a Starting Point for the Development of Conservation Strategies

Abstract

The contribution of this article is that it provides indications for the conservation of the surviving architecture in the Amatrice area, which was severely affected by the long seismic sequence between 2016 and 2017. The traditional construction technique was carefully studied during the safety works that were carried out in the villages from 2018 to 2020. From the studies conducted, it emerged that most of the historical building fabrics date to a post-seismic reconstruction phase of the seventeenth–eighteenth century. The masonry construction technique found on the site is homogeneous throughout the area and is based on the use of local sandstone and raw earth mortars. The mineralogical nature of the materials used for building was identified by means of specific diagnostic analyses as well as the production processes of the materials. An interpretation of the use of different materials and processes was provided by cross-referencing the analytical results with the existing data on the seismic history and on the geographical and geological characteristics of the territory; additionally, an interpretation of the development of the local historical construction techniques was provided as well. Defining conservation strategies for buildings that are still recoverable is an important objective that is aimed at safeguarding the material evidence of the local construction tradition. The same conservation strategies could be pre-emptively adopted in similar rural contexts.

1. Introduction

Between 2016 and 2017, an extreme series of earthquakes hit central Italy. The first mainshock occurred on August the twenty fourth, 2016; the epicentre was localized nearby Amatrice, a small historical city in northern Lazio. The aftershocks came in succession up to January 2017, and the most intense aftershock occurred nearby Norcia (Umbria) on October the thirtieth 2016, considerably increasing the damage [1]. The entire earthquake sequence involved a wide territory, including four different regions: Lazio, Umbria, Abruzzo, and Marche. Among the more than one hundred historical city centres severely damaged, Amatrice, hit since the first mainshock, had the worst balance; the current study has focused on this territory. The need to urgently study the local building tradition to increase knowledge and to provide a documentation of the conservative conditions of the historical fabrics immediately became clear. The historical urban fabric was not well-known until the earthquake, except for architectural excellences such as churches and palaces. It was therefore considered important to develop an analysis procedure as it is necessary for the study of historic buildings. In fact, in recent years, scientific research has shown that in the many cases where the historical evolution of architecture is not known, it is important to establish a precise and repeatable analysis procedure for an in-depth knowledge of the built heritage [2]. The focus of this research has been oriented on ordinary historical residential and rural buildings. To conduct this research, it was necessary to project and carry out a program of emergency studies to know and document local historical building traditions. The conditions of collapse or severe damage have been converted into opportunities of a fast examination of the structures and building techniques (Figure 1, Figure 2 and Figure 3).

It emerges from the research that the local building tradition is totally connected to the available materials in the territory. Over time, a masonry technique based on the exclusive use of local stone and raw earth mortars has been developed. The use of raw earth as a building material, often in rural settings, is well-documented by the scientific literature [3,4,5,6]. The aspects that are much less known are the mixed construction techniques, such as the one detected, which are based on the use of stone and raw earth mortars [7].The knowledge base currently available for the materials involved in the local construction technique makes it possible to conduct further studies on their response to seismic stresses, as has been performed in other cases [7,8], so as to be able to evaluate their vulnerability and consolidate them [9]. The research results could be very useful for future conservation and reconstructive actions in the Amatrice territory and could also suggest the development of investigation and conservation strategies for other similar contexts.

2. Materials and Methods

The research was conducted in an emergency context. The investigations were conducted in Amatrice and in ten selected hamlets1. At the time of the on-site investigation, the historic centres had been evacuated and bounded. The only ones who could access the vicinity were firefighters, civil protection, and teams of experts for the survey of seismic damage2. In order to conduct the studies on the local building tradition in safety, it was necessary to project a method that would allow for data to be collected effectively and quickly without hindering the routine operations of the survey of damages and safety interventions that were being conducted on the structures. We proceeded by developing the work in two distinct phases. The first phase was an operational phase based on exploratory inspections, data collection, and sampling of material to be analytically studied. The second phase was carried out in several university laboratories, where the analyses of sampled materials were performed: characterization in optical microscopy, microanalytical tests for the detection of CaCO₃ in mortars, petrographic characterization of thin sections of mortars, and analysis of materials in XRD and SEM + EDS3.

2.1. Phase I: Inspection, Data Collection and Sampling

The organization of the inspections was based on prior knowledge of the urban structure and the pre- and post-earthquake conditions of the historic centres. A historical and urban planning study of Amatrice and its hamlets was therefore carried out in advance, which was also performed through the use of graphic and photographic documentation dating back to shortly before the earthquake. We then proceeded by projecting forms for the on-site data collection. The data collection forms have been conceived as an attachment of the already existing schedographic tools used for the survey of post-seismic damage according to the Italian legislation in force (AeDES)4. This on-site editable document has been called the “Materic-Diagnostic Attachement” (AMD) to AeDES form. Creating a document that was directly connected to those produced in standard practice was useful for the speed of data collection, integration with other information on the building structures normally examined following an earthquake, and sharing of knowledge. The AMD is dedicated to the in-depth knowledge of materials, local construction techniques, and the state of conservation of the ordinary historical building heritage. The compilation of the AMD and the damage detection activities conducted by the civil defence is simultaneously conducted. In this way, it is possible to provide an overall assessment of the extent of the seismic damage that is also in relation to the previous degradation, which in turn is dependent or not dependent on the traditional construction techniques. The AMD also allows users to compile a part relating to preliminary general indications on any diagnostic investigations, with a vision oriented towards the conservation of ordinary historical buildings affected by seismic events. The AMD forms were extensively compiled on all ordinary buildings in the accessible areas of the historical centres chosen for this research. In this way, it was finally possible to gather enough information to document the material and technical characteristics of the historical building fabric (Appendix A,

Table A1. Masonry: “irregular and of bad quality” texture ▣.

	Lithic Elements	① Non Processed Stone Elements	② Pebbles		③ Bricks		④ Other Elements
Mortars
⑤ Traditional ◻ In a good state of conservation ▣ In a bad state of conservation		▣	▣		▣		▣
⑥ Recent grouting ◻ In a good state of conservation ▣ In a bad state of conservation		▣	▣		▣		▣
	Decay	Joints Erosion		⑦ Lithic Element’s Decay		Lack of Lithic Elements
Extension E
E > 75%		▣		◻		◻
NOTES:
① Local sandstone
② Sandstone pebbles used to fill gaps between other irregular lithic elements.
③ Brick fragments used to fill gaps between other irregular lithic elements.
④ Flysch laminar fragments used to create horizontal support surfaces for the sandstone blocks.
⑤ Raw earth binding mortars.
⑥ Cementitious grouting made to fill the eroded joints.

,

Table A2. Wooden structures ▣.

	Decay	① Degraded Elements		Missing Elements
Structural Typology
Main beams		◻		◻
Secondary beams		▣		▣
Extent		▣ 10–50%	◻ 50–75%		◻>75%
NOTES:
① Biodeteriogen attacks, fungal attacks

and

Table A3. Notable elements ▣.

▣Corner structures	Made with carved blocks of local sandstone
▣Frames	Windows frames: simple frames in local sandstone (partially collapsed).
▣Portals	Very simple portal in local sandstone with a segmental arch.
▣Decorative elements/Other type of elements	Flat arches made with bricks over the portal and the windows.

). The study of the building techniques was based on a direct observation of the walls in the elevation. The sections of walls that had collapsed were also directly studied. The identification of the lithotypes and the study of the morphology and processing of the stone wall elements were related to the research in the territory of the supply sites of raw materials: quarries and natural deposits. Stone samples from the masonry and quarry sites were taken, compared, and analysed in the laboratory (Appendix B and Appendix C). An analytical characterization campaign was also conducted for the bedding mortars of the masonry and for the residues of traditional plasters that were still present on many buildings.

2.2. Phase II: Analytical Characterization of Materials

During the on-site survey campaign, the resistance of the bedding mortars was estimated by means of a campaign of on-site penetrometric measurements5 (

Table 1. Resistance of the mortar to penetration measurements (RPM).

ID Test	Penetration Depth mm/10 Strokes				RPM
Building 01	Test a	Test b	Test c	Average
	20 mm	20 mm	12 mm	17.33 mm	0.77 MPa
Building 02	Test a	Test b	Test c	Average
	9 mm	6 mm	7 mm	7.33 mm	2.3 MPa
Building 03	Test a	Test b	Test c	Average
	5 mm	8 mm	12 mm	8.33 mm	2.02 MPa
Building 04	Test a	Test b	Test c	Average
	10 mm	12 mm	9 mm	10.33 mm	1.65 MPa
Building 05	Test a	Test b	Test c	Average
	15 mm	10 mm	12 mm	12.33 mm	1.35 MPa
Building 06	Test a	Test b	Test c	Average
	16 mm	9 mm	12 mm	12.33 mm	1.35 MPa
Building 07	Test a	Test b	Test c	Average
	6 mm	18 mm	9 mm	8 mm	2.15 MPa
Building 08	Test a	Test b	Test c	Average
	10 mm	15 mm	12 mm	12.33 mm	1.35 MPa
Sampling Map for the penetrometric measurements in Cornillo Nuovo (Amatrice)

). Samples of bedding mortars and traditional plaster were collected (Appendix D and Appendix E). Samples of lithic materials from the masonry of the buildings and samples from quarries and natural deposits were also collected (Appendix B and Appendix C). In fact, it seemed appropriate to compare the characteristics of the lithic materials used in the masonry with those of the lithotypes available in the area to investigate how the possible selection of materials took place in the historical constructive tradition. In total, 16 samples were selected for diagnostics, 8 mortars and 8 lithic materials (

Table 2. List of mortar samples (actually used for the diagnostic investigations)6.

Sample Code	Location	Common	Material
C.I.01	Capricchia	Amatrice	Binding mortar
C.I.02	Capricchia	Amatrice	Finishing mortar
C.N.E	Cornillo Nuovo	Amatrice	Binding mortar
C.N.I	Cornillo Nuovo	Amatrice	Binding mortar
C.N.P	Cornillo Nuovo	Amatrice	Finishing mortar
P.B.	Preta	Amatrice	Plaster
R.L.01	Retrosi	Amatrice	Binding mortar
R.L.02	Retrosi	Amatrice	Binding mortar

and

Table 3. List of lithic samples.

Sample Code	Locationat	Common	Material
P.C.O.01	Preta-Ortanza	Amatrice	Sandstone
P.C.O.02	Preta-Ortanza	Amatrice	Sandstone
P.C.O.03	Preta-Ortanza	Amatrice	Sandstone
P.C.O.04	Preta-Ortanza	Amatrice	Quartz
M.G.05	Mount Gorzano	Amatrice	Marl
M.G.06	Mount Gorzano	Amatrice	Compact limestone
M.G.07	Mount Gorzano	Amatrice	Clay
C.N.A7	Cornillo Nuovo	Amatrice	Sandstone
Localization of Sampled Building and Locations
C.I.01	C.I.01	C.N.E	C.N.I
Capricchia Building I	Capricchia Building I	Cornillo Nuovo Building E	Cornillo Nuovo Building I
GPS coordinates:	GPS coordinates:	GPS coordinates:	GPS coordinates: 42°61′43.4″ N
42°62′25.7″ N 13°31′10″ E	42°62′25.7″ N 13°31′10″ E	42°61′45.6″ N 13°33′16.6″ E	13°33′19.3″ E
C.N.P	P.B	R.L.01	R.L.02
Cornillo Nuovo Building P	Preta Building B	Retrosi Building L	Retrosi Building L
GPS coordinates: 42°61′41.7″ N	GPS coordinates: 42°61′75″ N	GPS coordinates:	GPS coordinates: 42°62′30.5″ N
13°33′21.4″ E	13°34′47.4″ E	42°62′30.5″ N 13°31′78.9″ E	13°31′78.9″ E
C.N.A.	P.C.O.01	P.C.O.02	P.C.O.03
Cornillo Nuovo Building A	Preta Ortanza Cascade	Preta Ortanza Cascade	Preta Ortanza Cascade
GPS coordinates: 42°61′45.3″ N	GPS coordinates: 42°60′84.2″ N	GPS coordinates:	GPS coordinates: 42°60′84.2″ N
13°33′12.6″ E	13°35′00.6″ E	42°60′84.2″ N 13°35′00.6″ E	13°35′00.6″ E
P.C.O.04	M.G.05	M.G.06	M.G.06
Preta Ortanza Cascade	Monte Gorzano	Monte Gorzano	Monte Gorzano
GPS coordinates: 42°60′84.2″ N	GPS coordinates: 42°62′59″ N	GPS coordinates:	GPS coordinates: 42°62′59″ N
13°35′00.6″ E	13°35′83.7″ E	42°62′98.33″ N 13°35′98.3″ E	13°35′83.7″ E

). The code assigned to each sample was derived from the mapping of the buildings, which was carried out during the on-site inspections. Any location that was inspected was coded using the capital initial letter (Capricchia = C, Cornillo Nuovo = C.N., etc.) and mapped, where we assigned each building a capital letter of the alphabet, as can be seen in the figure below (Figure 4).

The materials were analysed using optic microscopy (optic stereo-microscopy Euromex 10-70X, cross-section; Zeiss Axioskop 40 polarizing microscope, thin section). The lime content in the two types of mortar was verified by means of micro-analytical assays. The mineralogical components of the building materials were identified by an XRD analysis (Bruker D5000 Diffractometer, CuKα radiation, at 40 kV and 40 mA, step size 0.02° for 2 s, powder sample). Other characteristics of the traditional plasters were investigated by SEM imaging (Zeiss Gemini 500 SEM).

3. Results

3.1. The Traditional Building Technique

Most of the masonry analysed were determined to be made up of lithic elements of local sandstone bonded with raw earth. Based on the masonry cataloguing criteria included in the AeDES forms, the wall textures fall within the case study defined as “irregular and of bad quality”. This does not mean that the walls are completely chaotic, but they are not made up of regular elements arranged in perfectly horizontal rows. In fact, in almost all of the buildings with exposed masonry, the stone elements have little to no work performed at all, presenting uneven dimensions. However, there is a good attention to the arrangement of the stone elements in order to respect some horizontal line of the rows. This construction technique has certainly developed as a result of making the most of the territorial resources and natural conditions of the available materials [10]. In fact, the fracture planes of the local sandstone, which emerge in benches throughout the territory, indicate that this type of stone tends to naturally detach in blocks of rather regular shape (Figure 5 and Figure 6).

The use of the local lithotype has therefore made it possible to build walls which, although not made of square blocks, respect a fair regularity of the textures (Figure 7).

The alignments of the rows were also managed using another material available in the area, namely flysch. Flysch is a deposit of marly sandstone that detaches very easily from the substrate in the form of very regular plates with parallel and orthogonal breaking planes according to their typical laminar structure (Figure 8). This lithotype was used in the masonry to regularize the wall equipment by creating horizontal support surfaces for the sandstone blocks. However, there is no lack of more irregular stone elements with a rounded shape. The territory is in fact characterized by a complex network of water resources, mostly in a torrential regime, whose action transforms the arenaceous lithotype into a pebble formation (Figure 9). The use of pebbles was found in the upper parts of the walls and inside the wall cores as filling. The latter data were recorded by analysing the wall sections that were made inspectable due to the numerous collapses; the walls are almost always made up of two similar vestments and a core filled with flakes of stone, pebbles and raw earth. In all of the cases analysed, the presence of connecting headers between the wall leaves was recorded (Figure 10). The walls were covered with protective layers of plaster. In fact, in many of the buildings with exposed or partially exposed masonry, remains of thin plaster were found. (Figure 11). From the visual analysis, the plaster appeared to be created from lime. The diagnostic investigations have thus confirmed the use of lime as a binder.

3.2. Results of Diagnostic Investigations on Masonry Binding Mortars

The resistance of the bedding mortars was estimated by means of penetrometric tests carried out in situ on eight historic buildings in Cornillo Nuovo, a hamlet of Amatrice8. Resistance values between 0.8 and 2.3 MPa were recorded, which are well below the minimum quality requirements normally accepted9 (Table 1). From the results of the diagnostic campaign, the historical binding mortars are composed of pressed raw earth. Optical microscopy investigations revealed a composition of a clayey binder and sandy aggregate, which was also mineralogically characterized by means of XRD analysis. The very fine grain size of the aggregates and the apparent absence of sieving suggests a direct use of the land available on-site without special processing. The results of the XRD analyses also highlighted the close connection between the mineralogical components of the arenaceous lithotypes emerging in the area and those of the binding mortars. The minerals identified are the same except for two exceptions: in the raw earth mortars, there is the presence of sanidine, which is not significantly present in the sandstones, and the absence of calcite and dolomite, which are instead minority components of the local sandstones. These data confirm the hypothesis of the use of raw earth taken from the soil. The soil, in fact, arises from the degradation of the mother rock, in this case from the claying of the sandstone. This phenomenon is very slow and is caused by mechanical, physical, and chemical processes. Among the chemical transformation processes, there is the dissolution of calcite and dolomite, which are generally decomposed before other silicate minerals. At the same time, the chemical alteration of some of the albite components of the sandstones can produce newly formed minerals such as sanidine [11] (

Table 4. Comparative analysis of the minerals detected by XRD between lithic samples and mortars.

Albite
Calcite
Clorite
Dolomite
Lizardite
Muscovite
Quartz
Sanidine
Samples	C.N.A.	P.C.O.01	M.G.06	C.I.02	R.L.01	R.L.02	C.N.I.
Typology	LITOTYPES			PLASTERS		BINDING MORTARS

The first column on the left lists the minerals detected by XRD analysis in the samples studied. The penultimate line below lists the codes of the samples that have been studied by XRD analysis. In the last row below, the three types of samples: lithotypes in light blue, plasters in violet, and binding mortars in pink. The boxes with the most saturated colour indicate the presence of the mineral in the sample for the column to which it belongs. It can be immediately noticed that the samples of binding mortar contain neither calcite nor dolomite, whereas there is a fair overlap with the other minerals that are also contained in the lithotypes and in addition to the sanidine. The plasters instead contain all of the minerals present in the lithotypes in addition to sanidine and lizardite.

).

3.3. Results of Diagnostic Investigations on Plasters

The plasters were made up of a binder based on air lime and sandy aggregate. The binder showed cooking defects, which manifested themselves through the presence of large and numerous calcinelli10. The observation using optical microscopy showed a good sieving of the aggregates and the intentional and systematic addition of vegetable fibres. This finding attests to a good tradition acquired in the processing of mortars for plasters. The XRD investigations on the mineralogical species confirmed the local and natural origin of the sands used as aggregates. The SEM investigations provided information on the good adhesion capacity of the binder to the vegetable fibres, which confirms the intentional choice of the addition, evidently aimed at avoiding shrinkage cracks and ensuring good adhesion of the plaster to the substrate. The substantial difference between the composition and processing of mortars for plaster and structural mortars for masonry is surprising, first of all for the use of air lime as a binder, and then for the well-structured use of vegetable fibres.

4. Discussion

The presence of air lime in the plasters and, at the same time, the total absence of this binder in the mortars of the masonry, can raise some questions about the reasons or the contemporaneity of the work. The mortar samples analysed were taken from historic buildings, which is presumably attributable to the post-seismic reconstruction phase following the devastating earthquakes that occurred from 1639 to 1703 [12]. The reuse of stone materials to rebuild the walls is quite plausible, whereas the exclusive use of raw earth in the walls is reasonably due to the scarce presence of limestone suitable for producing lime in the area. This post-seismic reconstruction also took place in a particularly depressed historical period from an economic point of view, in which the use of raw earth was also encouraged by building manuals in both Italy and in Europe [13,14]. If we consider, as can be assumed, that the mortars for the plasters are contemporary with the structural ones, it can be assumed that the builders were aware of the intrinsic fragility of masonry being bound with only raw earth. The differences found would derive from this awareness: the presence of lime, the careful sieving of the aggregates, and the conscious addition of fibrous vegetable materials. In fact, these walls, if well-protected by a layer of good plaster, could still ensure the stability and durability of the building as long as periodic maintenance of the structures was guaranteed, even with the possible replacement of damaged parts. Therefore, the lime produced with the few suitable raw materials available in the area was consciously reserved to produce plaster mortars, which acted as a protective material for the underlying walls.

5. Conclusions

By re-elaborating the analytical results in light of the existing data on the seismic history and on the geographical and geological characteristics of the territory, an interpretation was provided on the use of the different materials available and on the resulting processing in the different types of mortars; furthermore, a general interpretation of the development of traditional building techniques in the Amatrice area was provided. What emerges is the extraordinary ability of local communities to make the most of available resources. The lithological homogeneity (with very little presence of limestone suitable for the production of lime), the geographical isolation, the rural character of the villages, and the agricultural and pastoral vocation of the territory are factors that jointly contributed to the development of the peculiar and rather atypical building traditions at the national level. The traditional and rural architectural heritage that survived in the villages in the aftermath of the extensive destruction caused by the earthquake is configured as the bearer of deeply identifying historical values of the places affected. It is therefore necessary to propose appropriate conservation strategies in view of the next phase of the restoration and reconstruction of the villages and for the future of this territory. The first strategic action that should be introduced is that of scheduled maintenance. The vulnerability of earthen architecture is well-known, as are the techniques for its consolidation. However, the historical architecture of the Amatrice area does not directly fall into this series, as the buildings are built using a sort of mixed technique that uses stone and raw earth. Therefore, we hope for future insights and research with the aim of conducting conservative actions that could also be only based on the scheduled maintenance and reproduction of traditional building techniques, which is an approach that has proven to be a winning approach for earthen architecture [15]. Such conservation actions should systematically include the goal of the seismic improvement of buildings [7,9]. It should not be forgotten that the studied realities insist on a territory with a very high seismic risk, and these data cannot be neglected in the conservation strategies. Numerous factors contribute to the stability of the buildings; among these factors, one of the most important is the masonry quality. The direct observations of the buildings have shown that despite almost all of the walls having been built using little to no worked lithic material, those that have shown greater resistance to the earthquake show a certain care in the juxtaposition of the elements, the particular attention that has been paid to the realization of horizontal elements and regularizations in the facing, the technical measures that have been conducted in the corner solutions and insertion of connecting headers between external and internal facing, and the use of traditional anti-seismic devices, such as wooden radiciamenti11. These traditional constructive characteristics express a probable awareness of the seismic risk on the part of the populations that have historically settled in the territory and must be respected and understood for the purposes of creating adequate and compatible seismic improvement planning for the structures. The phase of restoring and reconstructing the affected territories must develop in parallel with incentive strategies for the repopulation of the villages. The repopulation, however, must be accompanied by campaigns to raise awareness of the conservation of historical building traditions. In fact, it should be considered that many of the small towns affected by the earthquake had already suffered a considerable depopulation and, consequently, a lack of maintenance for many of the buildings [16]. On the other hand, the buildings still inhabited before the earthquake were in many cases renovated but in an improper way, for example, with the reconstruction of floors and reinforced concrete roofs. The latter proved to be too rigid and heavy and were unable to ensure the stability of the buildings. The sum of all of these intrinsic frailties has led to a considerable level of damage to the centres affected by earthquakes, to which has been added the impossibility of preserving much of the urban fabric from demolition. It is therefore necessary to reflect on the advisability of intervening in advance in other centres where similar characteristics are found in the building tradition; in this sense, the areas for consideration in Italy are, for example, those of Campidano in Sardinia, where there are numerous historical testimonies of rural architecture in raw earth. However, above all, other areas where the seismic risk is greater must also be considered. In the Modena area, for example, though the seismic risk was not among the highest, the effects of the earthquake that struck Emilia-Romagna in 2012 were felt; in examples of rural architecture that were studied, and even in some cases of the nobler architecture, the construction methods were found to be based on the use of bricks and raw earth [14]. In southern Italy, the use of raw earth as a binder for the lithic elements in the walls of historic buildings has been studied in some areas of Basilicata and in the historic centre of Lamezia (Calabria) [17]. Those regions present a very high seismic risk. To prevent what happened in Amatrice, detailed studies on building techniques and historical materials should be immediately launched and maintenance plans should be implemented that mainly provide for the protection of the masonry built using raw earth while excluding the overlapping of reinforced concrete elements. These criteria should be applied on a global scale in all cases where there are historical testimonies of raw earth architecture or mixed techniques such as the one represented by this study where there is also a condition of high seismic risk.