fibers

Article with Straw Fibers and Earth Mortar: Mix Design and Mechanical Characteristics Determination

Maria Francesca Sabbà , Mariateresa Tesoro, Cecilia Falcicchio and Dora Foti *

Department of Civil Engineering Sciences and , Polytechnic University of Bari, Via E. Orabona 4, 70125 Bari, Italy; [email protected] (M.F.S.); [email protected] (M.T.); [email protected] (C.F.) * Correspondence: [email protected]

Abstract: Raw earth is one of the oldest building materials, which is suitable for various uses: from the of load-bearing to use for plasters and finishes. The presence of straw fibers can give different behavior to this material. The present paper illustrates preliminary sensory and qualitative analyses, and subsequent laboratory tests that allow the characterization of the raw earth material with straw fibers for rammed earth through mechanized compaction and the identification of a compatible earth mortar. The raw material considered in this study is mainly clayey; for this reason, a mix design usable with the pisé (or ) technique has been developed. Cylindrical samples have been made through a press and subject to unconfined compression and indirect tensile tests. The results of the tests showed consistent tensile and compressive strength values in the context of earth materials. At the same time, a study for the realization of a mortar with the same base

 was carried out considering four mixtures, in order to investigate the best compromise between  workability, shrinkage and compressive strengths. The purpose of the study was to investigate the

Citation: Sabbà, M.F.; Tesoro, M.; mechanical characteristics of the local material through preliminary and laboratory tests, to classify it Falcicchio, C.; Foti, D. Rammed Earth according to the Unified Soil Classification System (USCS) and to verify its suitability for a possible with Straw Fibers and Earth Mortar: use in the construction field. Mix Design and Mechanical Characteristics Determination. Fibers Keywords: raw earth; rammed earth; earth mortars; straw fibers; characterization tests; pisé technique 2021, 9, 30. https://doi.org/ 10.3390/fib9050030

Academic Editor: Marco 1. Introduction Francesco Funari Raw earth is one of the most utilized materials in the world, as well as having been used by most ancient civilizations. Raw earth buildings were among the first capable of Received: 16 February 2021 guaranteeing the building’s protection from the outside and durability, thus constituting Accepted: 20 April 2021 Published: 4 May 2021 the first forms of urban settlement. By means of some archaeological excavations carried out in Italy and in the rest of the world, today it is possible to argue that the origin of the

Publisher’s Note: MDPI stays neutral constructions made of raw earth dates back to about 6000 B.C. The most ancient examples with regard to jurisdictional claims in are the mosque of Djienne in the Republic of Mali, the granaries of the temple of Ramesses published maps and institutional affil- in Egypt, the monastery of Tabo in India [1]. In addition to individual buildings, entire iations. settlements were developed, such as the city of Taos in New Mexico, the city of Shibam or Manhattan of the , the city of Sanaa in Yemen and the citadel of Bam in Iran [1]. In Italy, the use of raw earth as a has been limited to only three categories: fortification, temple and housing [1]. The most common techniques to make raw earth a building material are , Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. and clay or pisé [2]. Adobe consists of making raw earth prepared manually or This article is an open access article through a mechanized process, left to dry in the sun. It is suitable for the construction of distributed under the terms and load-bearing and infill walls [3]. Cob provides a dense and mixture, made with earth conditions of the Creative Commons and straw, modeled by hand or with automatic mixers, and stacked layer by layer, to erect Attribution (CC BY) license (https:// walls without the aid of formwork [4]. The pisé technique consists of the compaction of creativecommons.org/licenses/by/ damp earth inside a wooden formwork [5] to produce mainly vertical structures. The earth 4.0/). usually used is that of the excavations, worked on site and compacted in layers of about

Fibers 2021, 9, 30. https://doi.org/10.3390/fib9050030 https://www.mdpi.com/journal/fibers Fibers 2021, 9, x FOR PEER REVIEW 2 of 20

Fibers 2021, 9, 30 2 of 20 compaction of damp earth inside a wooden formwork [5] to produce mainly vertical structures. The earth usually used is that of the excavations, worked on site and compacted15 cm (Figure in1 layers). For pisof abouté, lean 15 or cm medium-fat (Figure 1). earth, For oftenpisé, stonylean or [6 medium-fat], is used. Compaction earth, often is stonyperformed [6], is using used. water Compaction content, whichis performed is considered using optimalwater content, as it provides which theis considered maximum optimaldry density as it forprovides constant the compaction maximum energydry density [7]. for constant compaction energy [7].

Figure 1. TheThe production production cycle of clay or pisé pisé with reinforcement [[8].8].

The mechanical behaviorbehavior ofof rammedrammed earth earth depends depends on on a a number number of of factors factors [9 ],[9], first first of ofall all the the nature nature of of the the earth, earth, the the quantity quantity and and quality quality of of clay,clay, thethe fibersfibers added,added, thethe particle sizesize distribution of of the earth and the quantityquantity of additions,additions, the constructionconstruction technique (Adobe, Cob andand pispisé)é) andand nono less less influential, influential, the the method method of of compaction, compaction, either either manual manual or ormechanized. mechanized. These These factors factors determine determine the apparent the apparent density anddensity porosity. and Theporosity. mechanical The mechanicalproperties ofproperties rammed of earth rammed provided earth in provid the literatureed in the show literature very differentshow very values different (see valuesTable1 ).(see It isTable possible 1). It to is notice possible that to the notice bulk that density the valuesbulk density of un-stabilized values of rammed un-stabilized earth 3 rammedare in the earth range are (1700–2400 in the kg/mrange )[(1700–240010]; the corresponding kg/m3) [10]; axialthe compressivecorresponding strength axial 2 compressivevalues are between strength 0.60 values and 4.00are between N/mm .0.60 and 4.00 N/mm2.

Table 1. Material properties for rammed earth from thethe literature.literature.

BulkBulk CompressiveCompressive Young’sYoung’s ReferenceReference 3 DensityDensity (kg/m (kg/m3) ) StrengthStrength (MPa) (MPa) ModulusModulus (MPa) NZS 4297- NZS 4298 (1998) NZS 4297- NZS 4298 (1998) nd >1.30 E = 300 f nd >1.30 E = 300 fcc [11,12][11,12 ] P. Jaquin et al. (2006–2007) P. Jaquin et al. (2006–2007) nd 0.60–0.70 60 [13,14] nd 0.60–0.70 60 U. Röhlen[13,14] et al. (2010) 1700–2400 1.50–4.00 750 U. Röhlen et[10 al.] (2010) K. Dierks et al. (2000) 1700–2400 1.50–4.00 750 [10] 2100–2300 2.40–3.00 650 [15] K. DierksQ. Bui et et al. al. (2000) (2014) 2100–23001800 2.40–3.00 1.00 90–105 650 [15][16 ] Q. M.Bui Hall et al. et al.(2014) (2014) 18002020–2160 0.75–1.461.00 90–105 nd [16][17 ] D.M. Lilley et al. (1995) M. Hall et al. (2014) 1870–2170 1.80–2.00 nd [18] 2020–2160 0.75–1.46 nd V. Maniatidis[17] et al. (2007) 1850 3.88 205 D.M. Lilley et[19 al.] (1995) V. Maniatidis et al. (2008) 1870–2170 1.80–2.00 nd [18] 1850 2.46 160 V. Maniatidis[ 20et] al. (2007) L. Miccoli et al. (2012–2014) 1850 3.88 205 2190 3.73 4143 [19][21,22 ] V. ManiatidisA. Romanazzi et etal. al. (2008) (2019) 18502080 2.46 1.50 160 471 [20][23 ] L. Lacouture et al. (2007) L. Miccoli et al. (2012–2014) 1930 0.60 67 [24] 2190 3.73 4143 [21,22] A. Romanazzi et al. (2019) 2080 1.50 471 The scatter[23]of mechanical property values in the literature is large, especially in the case of Young’s modulus. It can depend on many factors such as the characteristics of the local soil, the different mix designs, the workmanship, the weathering and the different Fibers 2021, 9, 30 3 of 20

procedures for the mechanical characterization tests of the mix design. In particular, the studies in Table1 report both manual [21–23] and mechanical [13,19,20] constipation. Other uses of raw earth concern mortars and plasters reproduced by mixing earth with . Although these materials do not increase the mechanical strength, they are necessary for the installation of strengthening components as natural (e.g., straw fibers) or plastic nets on bearing walls and panels. The scientific literature [23,25–27], in fact, investigates the use of compatible textile-reinforced mortars (TRM) as reinforcement techniques for rammed earth buildings: the first studies [28] showed an improvement in the overall seismic capacity, similar to that obtained in the Adobe case [29]. To ensure the effectiveness and durability of the reinforcement technique, the charac- terization of each component is essential. The main control parameters for the mortar are workability, linear shrinkage and compressive and tensile strength [30]. Table2 summarizes the data relating to the compressive strength of mortars available in the literature and the minimum values reported in the NZS 4297 and NZS 4298 standards [11,12]. In particular, both Young’s modulus (160–1034 N/mm2) and compressive strength (0.75–2.88 N/mm2) present a high scatter of values, due to the reasons mentioned above for the mix design of the mixtures of the structural elements.

Table 2. Material properties for earth-based mortar from the literature.

Bulk Density Compressive Young’s Reference (kg/m3) Strength (MPa) Modulus (MPa) NZS4297–NZS4298 (1998) nd >1.30 E = 300 f [11,12] c M. Van Gorp (2017–2018) 2140 1.40 453 [31] Silva et al. (2014) 1830 1.26 1034 [32] A. Romanazzi et al. (2019) 2130 1.50 471 [23] M. Hall et al. (2014) 2020–2160 0.75–1.46 nd [17] D.M. Lilley (1995) 1870–2170 1.80–2.00 nd [18] V. Maniatidis et al. (2007) 1850–2100 2.88 205 [19] V. Maniatidis (2008) 1850–2100 2.46 160 [20] Domínguez Martínez (2015) 2210 1.50 536 [33]

The present study performed the characterization of the mix design of raw earth with straw fibers for the use of the pisé technique starting from local soil (Altamura, Italy) which was corrected and mechanically compacted in cylinders with a diameter and height of 150 mm. The goal of the study was to define the mechanical properties of the local soil (Apulia Region), through preliminary and laboratory tests, and to verify its suitability for possible use in construction, trying to standardize both the procedure for its characterization and the methodology for the realization of specimens that have a material as homogeneous as possible and a degree of compaction comparable with that of upcoming panels. Today in Italy there is no legislation that regulates the procedures for the mechani- cal characterization of raw earth, both from the point of view of the material and of the structural elements, always considering that there are so many variables that it is almost impossible to establish universally valid reference values. Therefore, the characterization procedure described combines regulations and manuals from different countries with indi- cations deriving from scientific literature studies. In addition, it is important to highlight the use of straw fibers for the pisé technique, which generally does not include any, as Fibers 2021, 9, 30 4 of 20

opposed to the adobe and cob techniques; this choice was made because the addition of straw allows containing the shrinkage phenomenon and increasing the ductility of the structural element. To define a procedure, easily repeatable, avoiding imperfections due to the operator’s manual skills, it was decided to opt for a mechanical compaction method, preferring it to a traditional one. This distinguishes the study from most of the existing scientific literature about rammed earth that uses manual compaction, e.g., among the case studies reported in Table1, only Jaquin [13] and Maniatidis [19,20] utilized mechanical constipation. Furthermore, the criteria for identifying a compatible earth mortar were identified through the study of four earth and sand mixtures and the results relating to their mechan- ical properties. The tests were conducted at “Madeinterra” Cooperative in San Vito dei Normanni (Brindisi, Italy), the Geotechnical Laboratory of the Polytechnic University of Bari, “Tecno- Lab” laboratory in Altamura (Bari, Italy) and “M. Salvati” testing materials laboratory of the Polytechnic University of Bari (Italy).

2. Materials and Methods In the raw earth field in Italy, there is no legislation that regulates the procedures for the mechanical characterization of the material and its structural elements. This depends on two reasons: the first one, the reduced presence of buildings, if not for some examples in restricted areas of the country, and the second one, the existence of many variables that characterize the material, so that a universal identification of valid reference values is difficult. First of all, the mechanical behavior of the material is strongly influenced by the nature of the earth and above all, by the quantity and quality of the clay it contains. In the case of the pisé technique, the method of compaction, whether manual or mechanized, is also highly influential. In order to use the earth as a building material, it is necessary to know its composition and characteristics. The tests that allow the characterization of each earth are of two types: in situ tests and laboratory tests [34].

2.1. Preliminary Knowledge Analysis In Situ Preliminary cognitive analyses were conducted at “Madeinterra” Cooperative in San Vito dei Normanni. A soil sample was taken from an area in the municipality of Altamura (San Giuliano Park area) at a depth of about 50 cm, in order to exclude the presence of molds and other living organisms. First, the density of the soil was calculated through the use of a cubic silicone mold with internal dimensions equal to (50 × 50 × 50) mm. The mold was then filled with pressed and well compacted earth to reach a humid density of 1.88 g/cm3, which is a non-optimal density for a mixture. The latter data will be useful in the sampling phase of the earth and subsequent preparation of the samples for estimating the quantity of earth needed for the study. Subsequently, preliminary sensory and qualitative analyses were performed [2,34], which provide an initial assessment of the suitability of a given earth for use in construction. It is easily possible to detect the presence of organic matter, or fine elements, as well as to distinguish silts from clays. Visual, olfactory, tactile, sedimentation, cohesion, cigar, adherence, resistance to dryness, collapse and washing led to the same outcome: the earth taken was mainly characterized by a clay component (Figure2). In particular, the sedimentation test (Figure2a) was decisive for obtaining the first percentages regarding the composition of the soil sample: 53% clay, 29% sand and 18% silt. To evaluate the plasticity of the earth, the slump test was performed (Figure2b); in the present case, the material of the specimen resulted as plastic and clayey, where in fact the sedimentation was not significant compared to the decrease in the sample and mold. The dry resistance was verified through the realization and the failure of “medallions” of about 7 cm in diameter (Figure2c), which were left to dry for 10 days; the material resulted as almost pure clay, Fibers 2021, 9, x FOR PEER REVIEW 5 of 20

Subsequently, preliminary sensory and qualitative analyses were performed [2,34], which provide an initial assessment of the suitability of a given earth for use in construc- tion. It is easily possible to detect the presence of organic matter, gravel or fine elements, as well as to distinguish silts from clays. Visual, olfactory, tactile, sedimentation, cohesion, cigar, adherence, resistance to dryness, collapse and washing led to the same outcome: the earth taken was mainly characterized by a clay component (Figure 2). In particular, the sedimentation test (Figure 2a) was decisive for obtaining the first percentages regarding the composition of the soil sample: 53% clay, 29% sand and 18% silt. To evaluate the plas- ticity of the earth, the slump test was performed (Figure 2b); in the present case, the ma- terial of the specimen resulted as plastic and clayey, where in fact the sedimentation was Fibers 2021, 9, 30 5 of 20 not significant compared to the decrease in the sample and mold. The dry resistance was verified through the realization and the failure of “medallions” of about 7 cm in diameter (Figure 2c), which were left to dry for 10 days; the material resulted as almost pure clay, and in fact the specimens broke with difficul difficultyty and did not create dust. Additionally, the “cigar“cigar test”test” (Figure2 2d)d) validated validated the the clayey clayey nature nature of of the theearth earthanalyzed analyzed identifying identifying a awell well cohesive mixture; in fact, the cylindrical or cigar-shaped element with a length of 10 cm and a diameter of 2 cm, did not fail by making it slide by 5 cm along the edgeedge ofof aa plan.plan.

Figure 2. Preliminary cognitive tests: ( a) simplified simplified sedimentation test;test; ((b)) collapse test;test; ((c)) adhesion test;test; andand (d)) cigar test.

2.2. Laboratory Analyses 2.2.1. Particle Size Analysis Laboratory tests tests on on the the soils allowed allowed to ex toperimentally experimentally define define their their composition. composition. The Theparticle particle size analysis size analysis of the of soil the was soil carried was carried out at the out Geotechnical at the Geotechnical Laboratory Laboratory of the Pol- of ytechnicthe Polytechnic University University of Bari (Figure of Bari 3) (Figure accordin3) accordingg to the ASTM to the D422 ASTM standard D422 standard[35] by means [ 35] ofby two means techniques, of two techniques, sieving for sievingthe coarse for thefrac coarsetion (grain fraction diameter (grain greater diameter than greater 0.075 mm) than and0.075 sedimentation mm) and sedimentation for the fine fraction for the fine(grain fraction diameter (grain less diameterthan 0.075 less mm). than The 0.075 reference mm). Theclassification reference system classification was USCS system [36], was developed USCS [36 by], Casagrande developed by[37] Casagrande. In this system, [37]. coarse- In this grainedsystem, soils coarse-grained were classified soils on were the classifiedbasis of gr onain the size, basis while of grainfine-grained size, while ones fine-grained on the basis ofones plasticity on the characteristics basis of plasticity (Atterberg characteristics limits), since (Atterberg the mineralogical limits), since composition the mineralogical has an importantcomposition influence has an importanton the mechanical influence response on the mechanical of the material. response of the material. The particle size analysis of the soil led to obtain the weight fractions relating to each class: 49% clay, 28% sand, 19% silt or lime, 4% gravel. It is a fine-grained soil (USCS) as the ASTM 200 sieve pass-through of 77% is greater than 50%. Figure3 is representative of the most frequent obtained results.

2.2.2. Atterberg Limits Atterberg limits [38] were calculated at the Geotechnical Laboratory of the Polytechnic University of Bari both on natural soil and on soil dried at 105 ◦C for 24 h. The measured values of the liquid limit wL are 40.4% for the soil in the natural state and 40.6% for the soil subject to the pre-drying treatment; those of the plastic limit wP are, respectively, 19.9 and 19.8%. The IP plasticity index is equal to 20.5% for the soil in the natural state and 20.8% for the pre-dried soil. They are comparable values; therefore, it is clear that the organic component is almost entirely non-existent, or in any case not very influential. Consequently, no pre-treatment is necessary. The soil activity index, with an ASTM 200 sieve, is equal Fibers 2021, 9, x FOR PEER REVIEW 6 of 20

Fibers 2021, 9, 30 6 of 20

The particle size analysis of the soil led to obtain the weight fractions relating to each class:to 0.42.49% Theclay, material 28% sand, under 19% examination, silt or lime, therefore, 4% gravel. is an inorganicIt is a fine-grained and inactive soil clay (USCS) with as the mediumASTM 200 plasticity sieve pass-through (CL classification of USCS).77% is greater than 50%.

FigureFigure 3. Grain–size 3. Grain–size curve curve of the of theearth earth under under exam exam S1; S1; Particle sizesize analysis analysis was was carried carried out onout different on different samples samples relative relative to to the thelocal local soil soil with with star startingting mass mass of of about about 1.71.7 Kg.Kg.

2.3.Figure Rammed 3 is Earth representative Mix Design Characterizationof the most frequent obtained results. 2.3.1. Mix Design 2.2.2. AtterbergRammed Limits earth constructions consist of sand, silt, clay and gravel; straw fibers are also addedAtterberg in the limits mix design [38] were in a second calculated step. Forat th thee Geotechnical pisé technique, Laboratory it is important of the that Polytech- each nic ofUniversity these classes of Bari is present both inon the natural soil in correctsoil and proportions. on soil dried In the at raw105 earth°C for constructions, 24 h. The meas- regardless of the technique, the presence of clay is decisive [39] as the latter has binding ured values of the liquid limit wL are 40.4% for the soil in the natural state and 40.6% for properties, which are essential for the performance of the final material, but at the same the timesoil subject entails ato high the shrinkagepre-drying [40 trea]. Fromtment; a comparison those of the with plastic the literaturelimit wP are, on raw respectively, earth 19.9constructions and 19.8%. The [22,41 IP– plasticity43], the soil index in question is equal “soil to S1”20.5% was for too the clayey; soil in so the it was natural degreased state and 20.8%with for river the sand pre-dried to improve soil. resistance They are to comparable atmospheric agents,values; fine therefore, calcareous it gravelis clear (4–8 that mm) the or- ganicand component coarse calcareous is almost gravel entirely (16–32 non-existent mm) were utilized, or in to any improve case not structural very influential. stability. Con- sequently,In particular,no pre-treatment the first correction is necessary. S2 was The created soil activity by mixing index, 45% ofwith soil an and ASTM 55% of 200 river sieve, is equalsand. to It is0.42. observed The material that the contributionunder examination, of the clay therefore, fraction has is beenan inorganic reduced byand about inactive clayhalf, with from medium 49 to 24%. plasticity (CL classification USCS). A second correction S3 was performed following a more scientific approach to obtain a particle size distribution closer to the Füller curve, allowing to optimize the density 2.3.of Rammed the rammed Earth earthMix Design and therefore Characterization the relative mechanical performances. In the Füller 2.3.1.grain–size Mix Design curve [40], the non-perfect spherical shape of the particles has been taken into considerationRammed earth [31]. constructions consist of sand, silt, clay and gravel; straw fibers are The corrected soil S3 consists of 30% of original soil, 30% river sand, 30% of fine gravel also added in the mix design in a second step. For the pisé technique, it is important that and 10% of coarse gravel (by weight). The new curve obtained was compared with the eachgrain of these particle classes size distributionis present in (PSD) the soil curves in correct recommended proportions. by Houben In the and raw Guillaud earth construc- [44] tions,and regardless by the Portuguese of the technique, National Laboratory the presence of Civil of clay Engineering is decisive LNEC [39] [45 as] (Figurethe latter4). has bindingTherefore, properties, the final which S3 is composedare essential of 42%for the sand, performance 35% gravel, of 16% the clay final and material, 7% lime; but it is at the samefound time within entails the a two high PSD shrinkage envelopes [40]. considered From anda comparison is similar to with the compositions the literature of theon raw earthsoils constructions used for the pis[22,41–43],é technique, the indicated soil in ques by othertion studies“soil S1” in thewas literature too clayey; (Table so3 ).it was de- greased with river sand to improve resistance to atmospheric agents, fine calcareous gravel (4–8 mm) and coarse calcareous gravel (16–32 mm) were utilized to improve struc- tural stability. In particular, the first correction S2 was created by mixing 45% of soil and 55% of river sand. It is observed that the contribution of the clay fraction has been reduced by about half, from 49 to 24%. A second correction S3 was performed following a more scientific approach to obtain a particle size distribution closer to the Füller curve, allowing to optimize the density of the rammed earth and therefore the relative mechanical performances. In the Füller grain– Fibers 2021, 9, x FOR PEER REVIEW 7 of 20

size curve [40], the non-perfect spherical shape of the particles has been taken into consid- eration [31]. The corrected soil S3 consists of 30% of original soil, 30% river sand, 30% of fine gravel and 10% of coarse gravel (by weight). The new curve obtained was compared with the grain particle size distribution (PSD) curves recommended by Houben and Guillaud [44] and by the Portuguese National Laboratory of Civil Engineering LNEC [45] (Figure 4). Therefore, the final S3 is composed of 42% sand, 35% gravel, 16% clay and 7% lime; it Fibers 2021, 9, 30 7 of 20 is found within the two PSD envelopes considered and is similar to the compositions of the soils used for the pisé technique, indicated by other studies in the literature (Table 3).

Figure 4. ComparisonComparison between the PSD curves of the soils soils and and th thee envelopes envelopes for for rammed rammed earth earth construction construction recommended by: ( a) Houben and Guillaud [[44];44]; and (b) the Portuguese NationalNational LaboratoryLaboratory ofof CivilCivil EngineeringEngineering [[45]45]..

Table 3. Theoretical andTable experimental 3. Theoretical values and experimental of the components values in of the the earth components mixture forin the pis éearthconstructions. mixture for pisé con- structions. Reference Clay Lime Sand Gravel Sand + Gravel Sand + Füller (1997) [40] 10%Reference 14%Clay 33%Lime 43%Sand Gravel 76% Van Gorp (2018) [31] 12% 13% 34% 41% 75% Gravel Houben & Guillaud (2004) [44] 8–16%Füller (1997) [40] 13–21% 10% -14% 33% - 43% - 76% Architecturas de terra en Van20–30% Gorp (2018) [31] - 12% -13% 34% - 41% 60–80% 75% Iberoamérica (2003) [41] P. Walker (2005) [46]Houben 5–20% & Guillaud (2004) 10–30% [44] 8–16% -13–21% -- - 45–80% - Lehmbau Regeln (2009) [47]Architecturas 15% de terra en Iberoamé 30% 15% 40% 55% MOPT (1992) [48] 5–16% -20–30% ------60–80% IETcc (1971) [49] 10–40%rica (2003) [41] 20–40% 10–40% 10–20% - Code of Practice SADCSTAN [42] P.5–15% Walker (2005) [46] 15–30% 5–20% -10–30% -- - 50–70%45–80% Standards Australia (2002) [50] 5–20% 10–30% - - 45–75% Miccoli et al. (2014) [22]Lehmbau 11% Regeln (2009) 25%[47] 15% -30% 15% - 40% 64% 55% Doat et al. (1991) [43] 15–18%MOPT (1992) [48] 10–28% 5–16% -- -- - 55–75% -

S3 IETcc16% (1971) [49] 7%10–40% 42% 20–40% 10–40% 35% 10–20% - -

Code of Practice SADCSTAN [42] 5–15% 15–30% - - 50–70%

2.3.2.Standards Laboratory Australia Tests: (2002) Proctor [50] Test 5–20% 10–30% - - 45–75%

TheMiccoli compaction et al. (2014) capacity [22] is a fundamental11% property25% of clay.- To achieve- the maximum64% compaction, earth must have a specific water content, the so-called “optimal water content” Doat et al. (1991) [43] 15–18% 10–28% - - 55–75% (OWC), which allows the particles to be brought into a denser configuration without too much friction. TheS3 Proctor test [51] was16% carried out7% at the “TecnoLab”42% laboratory35% in Alta-- mura (Italy) and was performed on S1, S2 and S3. For each specimen, a different number of tests was carried out involving the realization of samples with different quantities of water, keeping the compaction energy constant and starting from about 6 kg of material. As defined in [31], soil samples with a dry density between 1.76 g/cm3 and 2.10 g/cm3 have satisfactory performance, and those between 2.10 g/cm3 and 2.20 g/cm3 are excellent. The test was first performed directly on S1, finding a very high-water content of 20.5% and a very low maximum dry density of 1.67 g/cm3. These values confirm the unsuitability of the original soil to be used in construction: the water content would cause a considerable shrinkage and a wall made with this material would have insufficient performance, due to its low-density value. The obtained OWC for S2 is equal to 13.8% with a maximum dry density equal to 1.77 g/cm3, and indeed for S3, the maximum dry density value resulted as equal to Fibers 2021, 9, x FOR PEER REVIEW 8 of 20

2.3.2. Laboratory Tests: Proctor Test The compaction capacity is a fundamental property of clay. To achieve the maximum compaction, earth must have a specific water content, the so-called “optimal water con- tent” (OWC), which allows the particles to be brought into a denser configuration without too much friction. The Proctor test [51] was carried out at the “TecnoLab” laboratory in Altamura (Italy) and was performed on S1, S2 and S3. For each specimen, a different num- ber of tests was carried out involving the realization of samples with different quantities of water, keeping the compaction energy constant and starting from about 6 kg of mate- rial. As defined in [31], soil samples with a dry density between 1.76 g/cm3 and 2.10 g/cm3 have satisfactory performance, and those between 2.10 g/cm3 and 2.20 g/cm3 are excellent. The test was first performed directly on S1, finding a very high-water content of 20.5% and a very low maximum dry density of 1.67 g/cm3. These values confirm the un- suitability of the original soil to be used in construction: the water content would cause a considerable shrinkage and a wall made with this material would have insufficient per- formance, due to its low-density value. Fibers 2021, 9, 30 8 of 20 The obtained OWC for S2 is equal to 13.8% with a maximum dry density equal to 1.77 g/cm3, and indeed for S3, the maximum dry density value resulted as equal to 2.05 g/cm3 and the optimal water content equal to 8.9%. Figure 5 represents the graph of the dry2.05 density g/cm3 andvariation the optimal as a functi wateron contentof the water equal content to 8.9%. of Figure S3. 5 represents the graph of the dryThese density results variation confirm as that a function the selected of the mix water has content the qualities of S3. necessary to be used in the clayThese technique. results confirm The obtained that the water selected conten mixt hasguarantees the qualities a low necessary probability to be ofused shrinkage in the cracking.clay technique. The obtained water content guarantees a low probability of shrinkage cracking.

Figure 5. 5. ProctorProctor test test results results of ofthe the corrected corrected soil soil S3: S3:(a) results (a) results with withthe saturation the saturation curves; curves; (b) standard (b) standard Proctor Proctor test equip- test ment—mold.equipment—mold.

2.3.3.2.3.3. Laboratory Laboratory Tests: Tests: Realization Realization of Specimens EightEight cylindrical cylindrical specimens specimens were made for S3 mix, six to determine the compressive strengthstrength and and two two for for the the tensile tensile strength. strength. The The cylinders cylinders (with (with dimensions dimensions d d= =15 15 cm cm and and h =h 15= 15cm) cm) were were compacted compacted in in two two layers, layers, inside inside a asteel steel mold mold and and with with the the aid aid of of a a gyratory gyratory press, “Matest “Matest Gyrotronic”, Gyrotronic”, at at the the “TecnoLab” “TecnoLab” laboratory laboratory in in Altamura. Altamura. Furthermore, Furthermore, it is it importantis important to highlight to highlight that that the theProctor Proctor test allows test allows to archive to archive the sample’s the sample’s optimal optimal com- pactioncompaction conditions. conditions. The Thechoice choice of making of making the thespecimen specimen with with two two layers layers was was intended intended to reproduceto reproduce the the stratigraphy stratigraphy that that characte characterizesrizes the the pisé pis techniqueé technique on onthe the sample. sample. ScientificScientific literature, manuals manuals and and legislation [12] [12] often recommend adding fibrous fibrous elementselements such such as as straw straw to to the the raw raw earth earth mixtur mixture.e. One One of of the the main main reasons reasons for for their their use use is certainlyis certainly the theincrease increase in the in compound’s the compound’s tensile tensile strength, strength, but also but the also improvement the improvement of the of the material ductility. Generally, the amount of straw is closely related to the added material ductility. Generally, the amount of straw is closely related to the added quantity quantity of water in a raw earth mixture, as it helps to reduce shrinkage and cracking during the drying process. The adobe and cob techniques involve the use of mixtures in which a greater amount of water is added than the pisé technique. For these two techniques, the studies report a quantity of straw in regard to the total weight of mixture, respectively, for the adobe smaller than 3% in [34] and between 0.5 and 8% in [52] (although it is specified that in the traditional Peruvian technique the percentage is smaller than 0.5) and instead in a range between 1 and 3% [53] for cob. Therefore, for pisé, having no other references in the literature and using a soil with its natural humidity, a lower value was chosen than for the other two techniques. Then, the 0.25% percentage by weight of dry straw fiber was added, chopped into small elements with a maximum length of 3 cm, clearly distinguishable in the mix of the rammed earth specimens. As recommended by the literature that reports adobe and cob cases, straw was cut into strips with a length not exceeding half the thickness of the finished wall [12]. Wheat straw was chosen because it has already been used in some rural in Italy realized with the adobe technique [34]. In addition, the wheat straw chosen comes from the local cultivation of the “Murgia”, an area near the Municipality of Altamura. From laboratory tests, the mechanical properties of the wheat straw was obtained; in particular the Young’s modulus, which is in the range of 5–6 GPa, the average tensile strength which is in the range between 22 and 79 MPa and Fibers 2021, 9, x FOR PEER REVIEW 9 of 20

of water in a raw earth mixture, as it helps to reduce shrinkage and cracking during the drying process. The adobe and cob techniques involve the use of mixtures in which a greater amount of water is added than the pisé technique. For these two techniques, the studies report a quantity of straw in regard to the total weight of mixture, respectively, for the adobe smaller than 3% in [34] and between 0.5 and 8% in [52] (although it is spec- ified that in the traditional Peruvian technique the percentage is smaller than 0.5) and instead in a range between 1 and 3% [53] for cob. Therefore, for pisé, having no other references in the literature and using a soil with its natural humidity, a lower value was chosen than for the other two techniques. Then, the 0.25% percentage by weight of dry straw fiber was added, chopped into small elements with a maximum length of 3 cm, clearly distinguishable in the mix of the rammed earth specimens. As recommended by the literature that reports adobe and cob cases, straw was cut into strips with a length not exceeding half the thickness of the finished wall [12]. Wheat straw was chosen because it has already been used in some rural houses in Italy realized with the adobe technique [34]. In addition, the wheat straw chosen comes from the local cultivation of the “Murgia”, an area near the Municipality of Altamura. From laboratory tests, the mechanical properties Fibers 2021, 9, 30 of the wheat straw was obtained; in particular the Young’s modulus, which9 of 20 is in the range of 5–6 GPa, the average tensile strength which is in the range between 22 and 79 MPa and the average water absorption, which is in the range between 280% and 300%. In this case, strawthe average was waterused absorption,to limit the which sensitivity is in the rangeto wa betweenter, help 280% the and setting 300%. phenomenon In this case, and reduce shrinkagestraw was used cracks to limit [39]. the After sensitivity the compaction to water, help theprocess setting (by phenomenon means of and 5 tons reduce load), the cylin- shrinkage cracks [39]. After the compaction process (by means of 5 tons load), the cylinders ders were immediately demolded with the aid of compressed air to facilitate their extrac- were immediately demolded with the aid of compressed air to facilitate their extraction tion(Figure (Figure6). 6).

FigureFigure 6. 6.Cylindrical Cylindrical specimen specimen in beaten in beaten earth: ( aearth:) top view; (a) top and view; (b) lateral and view. (b) lateral view. Each S3 cylinder has a wet weight of approximately 6.0 kg and a wet density of approximatelyEach S3 2.28cylinder g/cm 3has. The a average wet weight value ofof the approximately specimen’s maximum 6.0 kg dry and densities a wet density of ap- proximatelywas 2.07 g/cm 32.28, which g/cm demonstrates3. The average the reliability value of thethe Proctor specimen’s test considering maximum that dry the densities was 3 2.07maximum g/cm value3, which was demonstrates equal to 2.05 g/cm the. reliability of the Proctor test considering that the max- The specimens were placed in a climatic chamber for at least 28 days. The tests were imum value was equal to 2.05 g/cm3. performed after the samples reached the water content in equilibrium (at 22 ± 2 ◦C and at a relativeThe humidityspecimens of 50%), were which placed was in found a climatic on average chamber at 2.0%. for at least 28 days. The tests were performed after the samples reached the water content in equilibrium (at 22 ± 2 °C and at a2.3.4. relative Laboratory humidity Tests: of Simple 50%), Compressive which was Test found on average at 2.0%. To determine the compressive strength and Young’s modulus, six cylindrical speci- mens were subject to non-confined compressive tests at the “M. Salvati” laboratory of the 2.3.4. Laboratory Tests: Simple Compressive Test Polytechnic University of Bari. The specimen type used for the test is shown in Figure6. TheTo INSTRONdetermine 5869 the press compressive was used and strength programmed and in Young’s such a way modulus, to ensure the six grad- cylindrical speci- mensual and were controlled subject application to non-confined of displacement compressive knowing the tests initial at height the “M. of the Salvati” specimens; laboratory of the Polytechnicand a layer of aboutUniversity 1.00 mm of of Bari. damp The earth specimen was interposed type onused the upperfor the and test lower is shown faces in Figure 6. to reduce the irregularities. Therefore, the test was displacement-driven and the specimen failure was carried out only after running three complete cycles (Figures7 and8 ); they showed that the material assumes an elastic behavior. For each cycle, the test equipment detected the longitudinal displacements reporting a typical stress–strain curve through the data processing software (Figure8). Fibers 2021, 9, x FOR PEER REVIEW 10 of 20

The INSTRON 5869 press was used and programmed in such a way to ensure the Fibers 2021, 9, x FOR PEER REVIEW 10 of 20 gradual and controlled application of displacement knowing the initial height of the spec- imens; and a layer of about 1.00 mm of damp earth was interposed on the upper and lower facesThe INSTRON to reduce 5869 the press irregularities. was used and Therefor programmede, the in testsuch wasa way displacement-driven to ensure the and the gradualspecimen and controlled failure application was carried of displaceme out only ntafter knowing running the initial three height complete of the spec-cycles (Figures 7 and imens;8); andthey a layer showed of about that 1.00 the mm material of damp earthassumes was interposed an elastic on thebehavior. upper and For lower each cycle, the test Fibers 2021, 9, 30 facesequipment to reduce the detected irregularities. the longitudinal Therefore, the displacements test was displacement-driven reporting a10 typical of and 20 the stress–strain curve specimenthrough failure the was data carried processing out only software after running (Figure three 8).complete cycles (Figures 7 and 8); they showed that the material assumes an elastic behavior. For each cycle, the test equipment detected the longitudinal displacements reporting a typical stress–strain curve through the data processing software (Figure 8).

Figure 7. Failure of the specimens of compression test. FigureFigure 7. Failure 7. Failure of the specimensof the specimens of compression of compression test. test.

3 3 2.5 2.5 S3_1 2 S3_2 S3_1 1.5 2 S3_3 S3_2 MPa 1 1.5 S3_4 S3_3

0.5 MPa S3_5 1 S3_4 0 S3_6 00.5 0.01 0.02 0.03 0.04 0.05 0.06 S3_5 mm/mm 0 S3_6

Figure 8. Stress–deformation0 0.01 plot relating 0.02 to S3_1–S3_6 0.03 cylindrical samples 0.04 subject 0.05 to an unconfined 0.06 Figurecompression 8. Stress–deformation test. plot relating to S3_1–S3_6 cylindrical samples subject to an uncon- fined compression test. mm/mm The compressive strength values obtained show a mean value equal to 2.58 MPa, a standard deviation (SD) equal to 0.08 and a relative standard deviation (RSD) equal to The compressive strength values obtained show a mean value equal to 2.58 MPa, a 3.10%.Figure This 8. lowStress–deformation SD value can be explained plot relating by the use to ofS3_1–S3_6 the gyratory cylindrical press which samples allowed subject to an uncon- standardto obtain deviation similar specimens (SD) equal as made to 0.08 with and a standard a relative production standard method. deviation (RSD) equal to 3.10%.fined ThisThe compression longitudinal low SD value elastic test. can modulusbe explained (Young’s by the modulus use of E) the is calculatedgyratory press through which a linear allowed to obtainregression similar considering specimens the peak as made stress with in the a rangestandard between production 5 and 30% method. [36]. TheThe longitudinalThe results compressive are reported elastic modulus in strength Table4 and (Young’s values Figure 9modulus. obtained E) isshow calculated a mean through value a linear equal to 2.58 MPa, a regressionstandard considering deviation the peak(SD) stress equal in theto 0.08range and between a relative 5 and 30% standard [36]. deviation (RSD) equal to 3.10%.The results This are low reported SD value in Table can 4 be and explained Figure 9. by the use of the gyratory press which allowed Tableto 4. obtain The compressive similar strengthspecimens values as and made the Young’s with modulusa standard E of S3 production specimens. method. The longitudinal elastic modulus (Young’s modulus E) is calculated through a linear N° Specimens Compressive Strength Young’s Modulus E regression considering the peak stress in the range between 5 and 30% [36]. The results are reported in Table 4 and Figure 9.

Table 4. The compressive strength values and the Young’s modulus E of S3 specimens.

N° Specimens Compressive Strength Young’s Modulus E Fibers 2021, 9, 30 11 of 20

Fibers 2021, 9, x FOR PEER REVIEW 11 of 20

Table 4. The compressive strength values and the Young’s modulus E of S3 specimens.

Compressive Strength Young’s Modulus E N◦ Specimens (MPa)(MPa) (MPa)(MPa) S3_1S3_1 2.482.48 65 65 S3_2S3_2 2.672.67 67 67 S3_3S3_3 2.592.59 78 78 S3_4 2.47 69 S3_4 2.47 69 S3_5 2.58 75 S3_5S3_6 2.672.58 75 70 S3_6 2.67 70 Mean Value 2.58 71 Mean Value 2.58 71 Standard Deviation (–) 0.08 4.49 Standard Deviation (–) 0.08 4.49 RelativeRelative Standard Standard Deviation Deviation (%)(%) 3.10% 3.10% 6.32% 6.32%

3 σ 2.5 2 Mix Design S3 1.5 Average Mix design S3 1 (MPa) 0.5 0

Compressive strength 60 65 70 75 80 Young’s modulus E (MPa)

Figure 9. Compressive strength of S3_1–S3_6 specimens vs. the Young’s modulus. Figure 9. Compressive strength of S3_1–S3_6 specimens vs. the Young’s modulus. By relating the Young’s modulus to the corresponding experimentally measured By relating the Young’s modulus to the corresponding experimentally measured compressive strength value, the obtained coefficient ratio is approximately 30 for S3 mix. compressive strength value, the obtained coefficient ratio is approximately 30 for S3 mix. For further confirmation, the specimens of the mix design S2 were tested anyway. As For further confirmation, the specimens of the mix design S2 were tested anyway. As expected, the compressive strength values are lower than S3; in particular, the S2 average expected, the compressive strength values are lower than S3; in particular, the S2 average value is equal to 1.59 MPa. In fact, it is important to consider the density factor which valuepositively is equal affects to 1.59 to MPa. the compressive In fact, it is strength important of the to specimens:consider the S3 density density factor is greater which than positivelyS2 density. affects to the compressive strength of the specimens: S3 density is greater than S2 density. 2.3.5. Laboratory Tests: Indirect Tensile (Brazilian) Test 2.3.5. LaboratoryThe test was Tests: performed Indirect at Tensile the “TecnoLab” (Brazilian) laboratory Test using a CONTROLS 65-L1200 au- tomaticThe test press was (Figure performed 10) according at the “TecnoLab” to the code laboratory [54]. Two cylindricalusing a CONTROLS specimens 65-L1200 were tested automaticwith a diameter press (Figure and height 10) according equal to 150 to mm,the code so with [54]. a ratioTwo equalcylindrical to 1.The specimens values obtained were testedfor thewith tensile a diameter strength andfct heightare reported equal to in 15 Table0 mm,5. so with a ratio equal to 1. The values obtained for the tensile strength fct are reported in Table 5. Table 5. Indirect tensile strength values of S3 specimens. Table 5. Indirect tensile strength values of S3 specimens. Tensile Strength f N◦ Specimens ct Tensile Strength(MPa) fct N° Specimens S3_7(MPa) 0.212 S3_7S3_8 0.212 0.269 MeanS3_8 Value0.269 0.240 StandardMean Value Deviation (–) 0.240 0.028 Standard Deviation (–) 0.028 Relative Standard Deviation (%) 11.67% Relative Standard Deviation (%) 11.67%

The failure of S3 specimens showed pseudo-vertical cracks, slightly curved in the central part due to the presence of coarse limestone gravel particles along the failure plane (Figure 10a). It has been possible to determine the cohesion (c) and friction angle () of the soil S3 through the following relations: Fibers 2021, 9, x FOR PEER REVIEW 12 of 20

− =sin = 56° (1) +

Fibers 2021, 9, 30 1 12 of 20 c= σ ·σ = 0.39 MPa (2) 2

Figure 10. Failure of specimens during the indirect tensile test: top view (a); and transversal view (b). Figure 10. Failure of specimens during the indirect tensile test: top view (a); and transversal view (b). The failure of S3 specimens showed pseudo-vertical cracks, slightly curved in the 2.3.6. Results Discussion central part due to the presence of coarse limestone gravel particles along the failure plane (FigureTo compare 10a). It hasthe beenobtained possible values to determine from the co thempressive cohesion (testc) and with friction those angleof the ( ϕlitera-) of the ture,soil reported S3 through in Table the following 1, for the relations: Young’s modulus it was necessary to consider the value of the maximum peak between 5 and 30%, through linear regression, since the literature σ − σ data were calculated under such ϕconditions.= sin−1 Mostc tof= the56 studies◦ in Table 1 were carried(1) out on cylindrical specimens with an h/d ratioσc equal+ σt to 2 and a cross-section equal to 100 mm, with the exception of [17,18], who1 use√ cubic specimens. Unfortunately, in the present study, h/d = 2 could not be utilized,c = nor theσc·σ samet = 0.39cross-section MPa due to the instrumental (2) 2 limitations. As a consequence, conversion factors were used as indicated in [55] to com- pare2.3.6. the Results compressive Discussion strength values. Therefore, the compressive strength values were multipliedTo compare by two thefactors obtained k1 = 1/1.18 values = from0.85 and the compressivek2 = 1.02, the testfirst with to take those into of account the literature, the differentreported slenderness, in Table1, for and the the Young’s second modulus to take into it was account necessary the todifferent consider cross-section the value of di- the mensionmaximum of the peak specimens between 5with and the 30%, same through h/d [56]. linear At regression, the end, sincethe final the literaturecompressive data strengthwere calculated values, initially under suchequal conditions. to 2.58 MPa Most for S3_1–S3_6 of the studies are innow Table equal1 were to 2.23 carried MPa. out The on compressivecylindrical strength specimens results with of an S3_1–S3_6h/d ratio specimens equal to 2 andcomply a cross-section with NZS 4298 equal [12] to and 100 are mm, comparablewith the exception with the of studies [17,18], [19,20], who use since cubic the specimens. same methods Unfortunately, of soil correction in the present and study,me- chanicalh/d = 2compaction could not be are utilized, used. As nor a theresult, same the cross-section data of the case due tostudy the instrumentalcan be considered limita- consistent.tions. As a consequence, conversion factors were used as indicated in [55] to compare the compressiveIn the scientific strength literature, values. there Therefore, are several the compressive studies concerning strength the values Young’s were modulus multiplied (seeby Table two factors 1), that k 1show= 1/1.18 a great = 0.85 dispersion and k2 = for 1.02, both the for first the to takeun-stabilized into account andthe stabilized different rammedslenderness, earth within and the a secondrange of to 60–4150 take into MPa account [10,13–24]. the different This is cross-sectiondue to many dimensionfactors like of thethe type specimens of the soil, with the the workmanship same h/d [56 and]. At the the testing end, the procedures. final compressive In the present strength study, values, the ini- Young’stially equal modulus to 2.58 reaches MPa for values S3_1–S3_6 of around are now 70 MPa, equal as to in 2.23 [13–14,24]. MPa. The compressive strength resultsWith of reference S3_1–S3_6 to the specimens tensile strength, comply withthere NZS are few 4298 studies [12] and in the are literature comparable which with do the notstudies allow [a19 direct,20], since comparison. the same Furthermore, methods of soil in correctionsome studies, and as mechanical in [16], considering compaction the are relationshipused. As a that result, exists the databetween of the compressive case study can strength be considered and tensile consistent. strength in rammed earth constructions,In the scientific this literature, parameter there is assumed are several to be studies equal concerningto fct = 0.11 f thec. This Young’s occurs modulus for S3 (see Table1), that show a great dispersion for both for the un-stabilized and stabilized rammed earth within a range of 60–4150 MPa [10,13–24]. This is due to many factors like the type of the soil, the workmanship and the testing procedures. In the present study, the Young’s modulus reaches values of around 70 MPa, as in [13,14,24]. With reference to the tensile strength, there are few studies in the literature which do not allow a direct comparison. Furthermore, in some studies, as in [16], considering the relationship that exists between compressive strength and tensile strength in rammed Fibers 2021, 9, 30 13 of 20

earth constructions, this parameter is assumed to be equal to fct = 0.11 fc. This occurs for S3 Mix design (see Table6); in case this factor is higher, this means that the tensile strength of rammed earth with straw has interesting results.

Table 6. Average characteristic parameters of the analyzed S3 specimen.

Specimen Characteristic S3 Value Dimensions d = h = 150 mm Wet weight 6000 g Wet density 2.28 g/cm3 Maximum dry density 2.07 g/cm3 Compressive strength 2.58 MPa Young’s modulus (E) 71 MPa Tensile strength 0.24 MPa Internal friction angle (ϕ) 56◦ Cohesion (c) 0.39 MPa

The results achieved allow to consider S3 as the mix compliant with regulations and guidelines [12,40,45] for rammed earth constructions both from a quantitative and qualita- tive point of view. In fact, it is the suitable mix for erecting rammed earth constructions: it contains fine and coarse gravel for structural stability, sand for greater resistance to atmospheric agents, as well as silt and clay for greater cohesion [31].

2.4. Earth Mortar Characterization 2.4.1. Mix Design The experimental campaign involved the realization of four earth mortars to be used both as non-structural finishes (internal and external coatings) and as structural part connections and/or strengthening devices for rammed earth walls. In particular, it is used as compatible support for any seismic reinforcement with a fabric net, which is necessary to redistribute the stresses inside the bearing wall and increase its ductility. The investigated mixture does not consider any stabilizing, opting for clay as the only binding element. In this way, maximum compatibility with earth structures is obtained, guaranteeing strong adhesion, even in the case of subsequent installation. Before the use, the soil was sieved through a 2 mm sieve, according to the recommendations given by Röhlen and Volhard [47] for mortars in earth constructions. Each mixture was obtained by adding a different amount of river sand to S1 (Table7). Sand was added, necessary to avoid high shrinkage, and therefore to reduce the risk of cracking during drying; this also increases the workability of the material by acting on the water content.

Table 7. Compositions of mortars M1–M4.

Mix Design Earth Sand Clay S1 100% 0% 49% M1 30% 70% 21% M2 25% 75% 16% M3 20% 80% 12% M4 15% 85% 9%

2.4.2. Laboratory Tests: Realization of the Specimens For the realization of the four mixes, earth and dry sand (previously sieved [47]) were placed inside an industrial mixer and mixed for 30 s. Subsequently, water was added and kneading proceeded for 90 s more. Once the mixture was produced, three prismatic specimens were realized for each mixture, in a steel mold, with dimensions of (160 × 40 × 40) mm (Figure 11). Before the use, the molds were doused with oil, to facilitate sample removal. Once filled, the molds were placed on a mortar setter (60 consecutive Fibers 2021, 9, x FOR PEER REVIEW 14 of 20

Fibers 2021, 9, 30 14 of 20

mm (Figure 11). Before the use, the molds were doused with oil, to facilitate sample re- moval. Once filled, the molds were placed on a mortar setter (60 consecutive hits) to re- movehits) to any remove remaining any remaining cavities. cavities.After filling After all filling the gaps all the and gaps leveling and leveling the surface the surfaceby means by ofmeans a spatula, of a spatula, the molds the moldswere covered were covered with a with slide, a slide,and all and the all samples the samples were were stored stored in a climaticin a climatic chamber chamber for a for period a period of 28 of days, 28 days, at a constant at a constant temperature temperature and andhumidity humidity (at 22 (at ± 222 °C± and2 ◦C 50% and relative 50% relative humidity). humidity).

Figure 11. ManufacturingManufacturing of of earth earth mortar mortar specimens: specimens: (a) (equipment;a) equipment; (b) earth (b) earth mortar mortar mix design; mix design; and (c and) prismatic (c) prismatic speci- specimens.mens.

2.4.3. Laboratory Tests: Evaluation of Linear and Volumetric ShrinkageShrinkage Shrinkage is important as it can cause the earth mortar to crack. Shrinkage cracks due to drying trigger, in fact, could produce furtherfurther cracks; as a consequence, water can easily penetrate, negatively negatively affecting affecting the the mechanical mechanical characteristics. characteristics. Therefore, Therefore, low low shrinkage shrinkage is essentialis essential to tohave have a asafe safe and and compatible compatible mortar mortar layer. layer. The The New New Zealand standards [[12]12] state that the linear shrinkage of an un-stabilizedun-stabilized earth mortar should never exceed 3%, Gomes [30] [30] reduces this value to 2%. After 28 days of drying in a climatic chamber at a constant temperaturetemperature andand humidity, humidity, the the linear linear and and volumetric volumetric shrinkage shrinkage was was assessed assessed (see Table(see Table8). All 8). mortars All mortars have have verified verified the limit the linearlimit linear shrinkage shrinkage of 3%, of established 3%, established by [ 12 by], with[12], with the exception the exception of the of M1 the mortar. M1 mortar. Furthermore, Furthermore, M3 and M3 M4 and mortars M4 mortars satisfied satisfied the limit the shrinkage values of 2% as well. Therefore, M1 and M2 mortars were excluded, exceeding limit shrinkage values of 2% as well. Therefore, M1 and M2 mortars were excluded, ex- the limit of 2%. ceeding the limit of 2%. As expected, shrinkage was higher for mortars with higher clay content. The effect As expected, shrinkage was higher for mortars with higher clay content. The effect is is evident in both linear and volumetric shrinkage. In this case, the relationship seems to evident in both linear and volumetric shrinkage. In this case, the relationship seems to be be of a higher order than linear, increasing more with the clay content: a 4% variation in of a higher order than linear, increasing more with the clay content: a 4% variation in the the quantity of clay component determines approximately a doubling of the volumetric quantity of clay component determines approximately a doubling of the volumetric shrinkage value (Figure 12). shrinkage value (Figure 12). The test carried out allows to determine the dry bulk density of the mortars equal to The test carried out allows to determine the dry bulk density of the mortars equal to ρ = 1.75 ± 0.03 g/cm3. ρ = 1.75 ± 0.03 g/cm3. Table 8. Linear and volumetric shrinkage of mortars M1–M4. Table 8. Linear and volumetric shrinkage of mortars M1–M4. Shrinkage (mm) N◦ Specimens Shrinkage (mm) N° Specimens ∆L1 ∆L2 ∆H ∆V ΔL1 ΔL2 ΔH ΔV M1_1 3.6 1.5 1.6 7.7 M1_2M1_1 3.33.6 1.61.5 1.61.6 7.7 7.9 M1_3M1_2 3.13.3 1.31.6 1.51.6 7.9 7.3 Mean ValueM1_3 M1 3.33 (2.1%3.1) 1.47 (3.7%1.3) 1.57 (3.9%1.5 ) 7.637.3 Mean Value M1 3.33 (2.1%) 1.47 (3.7%) 1.57 (3.9%) 7.63 Standard Deviation M1 0.21 0.12 0.05 0.25 Standard Deviation M1 0.21 0.12 0.05 0.25 Fibers 2021, 9, 30 15 of 20

Fibers 2021, 9, x FOR PEER REVIEW 15 of 20 Table 8. Cont.

Shrinkage (mm) N◦ Specimens ∆L1 ∆L2 ∆H ∆V M2_1 3.8 1.0 1.2 6.0 M2_1 3.8 1.0 1.2 6.0 M2_2M2_2 0.90.9 1.11.1 1.21.2 3.2 3.2 M2_3M2_3 0.80.8 1.11.1 1.21.2 3.1 3.1 MeanMean Value Value M2 M2 1.83 (1.1%1.83) (1.1%1.07) (2.7%1.07) (2.7%1.2 ()3.0% 1.2) (3.0%) 4.1 4.1 StandardStandard Deviation Deviation M2 M2 1.39 1.39 0.05 0.05 0 0 1.34 1.34 M3_1M3_1 0.60.6 0.70.7 0.60.6 1.9 1.9 M3_2M3_2 0.70.7 0.80.8 0.70.7 2.2 2.2 M3_3M3_3 0.50.5 0.60.6 0.50.5 1.6 1.6 MeanMean Value Value M3 M3 0.60 (0.4%0.60) (0.4%0.70) (1.8%0.70) (1.8%0.60)( 1.5%0.60) (1.5%1.90) 1.90 StandardStandard Deviation Deviation M3 M3 0.08 0.08 0.08 0.08 0.08 0.08 0.24 0.24 M4_1M4_1 0.50.5 0.60.6 0.50.5 1.6 1.6 M4_2 0.6 0.5 0.5 1.6 M4_3M4_2 0.40.6 0.50.5 0.50.5 1.4 1.6 M4_3 0.4 0.5 0.5 1.4 Mean Value M4 0.50 (0.3%) 0.53 (1.3%) 0.5 (1.3%) 1.53 Mean Value M4 0.50 (0.3%) 0.53 (1.3%) 0.5 (1.3%) 1.53 Standard Deviation M4 0.08 0.05 0.00 0.09 Standard Deviation M4 0.08 0.05 0.00 0.09

8

6 Volume 4 Height

2 Width Shrinkage (%) Length 0 9 121621 Clay content (%)

FigureFigure 12. 12.Linear Linear and and volumetric volumetric average average shrinkage shrinkage for M1–M4 for M1–M4 mortars mortars vs. the vs. clay the content. clay content.

2.4.4.2.4.4. Laboratory Laboratory Tests: Tests: Three-Points Three-Points Bending Bending Test Test The bending strength was tested using a CONTROLS 65-L1200 automatic press com- The bending strength was tested using a CONTROLS 65-L1200 automatic press com- pliant with UNI EN 1015-11 [56]. The equipment is configured in a controlled load mode, withpliant a constant with UNI speed EN equal1015-11 to 0.01[56]. N/mm The equipment2s (minimum is configured value that can in bea controlled set) and permit load mode, 2 towith perform a constant the bending speed testequal on threeto 0.01 points, N/mm throughs (minimum the deflection value that of the can samples be set) up and to permit breakto perform in two stumpsthe bending (Figure test 13). on The three maximum points, load through was automatically the deflection recorded of the andsamples it is up to possiblebreak in to two determine stumps the (Figure bending 13). strength The maximumft in relation load to was its geometry: automatically recorded and it is possible to determine the bending strength ft in relation to its geometry: F ·l f = 1.5 ∗ t 1 (3) t 2 · l2·h =1.5∗ (3) ·ℎ where l1 is the distance between the supports, l2 and h are the width and height of the specimenwhere l1 andis theFt isdistance the maximum between force the applied. supports, The resultsl2 and areh are summarized the width in and Table height9. of the specimen and Ft is the maximum force applied. The results are summarized in Table 9. Fibers 2021, 9, 30Fibers 2021, 9, x FOR PEER REVIEW 16 of 20 16 of 20

Figure 13. Three-pointFigure 13. bendingThree-point test performed bending on test M1–M4 performed mort onar specimens M1–M4 mortar at the specimens “TecnoLab” at laboratory: the “TecnoLab” (a) pre-test; and (b) post-test.laboratory: (a) pre-test; and (b) post-test.

Table 9. Bending2.4.5. and Laboratory compressive Tests: strength Simple of mortars Compressive M1–M4. Test The portions resulting from the bending failure of the mortars were used for the sim- Compressive Strength f Bending Strength f N◦ Specimens c ct ple compression tests (Figure 14).(MPa) Each sample was split in two(MPa) parts, thus, a total of six compression tests were performed for each kind of mortar. M1_1The tests were carried 1.88 out according to 3.13 UNI EN 1015-11 [56] 0.91 at “TecnoLab” labora- M1_2 2.65 3.07 0.87 M1_3tory by means of the CONTROLS 2.13 65-L1200 1.97press; in this case, the 0.98 load is uniformly trans- mitted through a 40 × 40 mm plate to the specimen (Figure 14a). Therefore, the equipment Mean Value M1 2.47 0.92 considers only the contact square area between the plate and the specimen, equal to 1600 Standard Deviationmm2, to return M1 the value of the compressive 0.51 strength fc according 0.05 to the formula (4): M2_1 1.79 1.23 0.60 M2_2 1.89 1.25 = 0.69 (4) M2_3 1.70 1.90 0.61

Mean Valuewhere M2 A is the square area section 1.63 of the sample subject to compressive 0.63 load and Fc the Standard Deviationmaximum M2 force applied or compressive 0.28 strength. The results are 0.04reported in Table 9. M3_1 1.33 1.11 0.51 M3_2 1.80 1.50 0.55 M3_3 1.27 1.52 0.43 Mean Value M3 1.42 0.50 Standard Deviation M3 0.22 0.05 M4_1 0.92 1.13 0.47 M4_2 0.95 0.91 0.43 M4_3 0.71 1.16 0.35 Mean Value M4 0.96 0.42 Standard Deviation M4 0.15 0.05

2.4.5. Laboratory Tests: Simple Compressive Test The portions resulting from the bending failure of the mortars were used for the simple compression tests (Figure 14). Each sample was split in two parts, thus, a total of six compressionFigure tests 14. were Unconfined performed compression for each kindtest on of M1–M4 mortar. mortar specimens, at the “TecnoLab” labora- tory: (a) pre-test; and (b) post-test.

2.4.6. Results Discussion The mechanical characteristics of the mortar were evaluated from the three-point bending test and the compression test on prismatic samples. From Table 9, it can be noticed that the compressive and bending strength values increase as the amount of clay present in the mortar sample increases. In fact, the highest Fibers 2021, 9, x FOR PEER REVIEW 16 of 20

Figure 13. Three-point bending test performed on M1–M4 mortar specimens at the “TecnoLab” laboratory: (a) pre-test; and (b) post-test.

2.4.5. Laboratory Tests: Simple Compressive Test The portions resulting from the bending failure of the mortars were used for the sim- ple compression tests (Figure 14). Each sample was split in two parts, thus, a total of six compression tests were performed for each kind of mortar. The tests were carried out according to UNI EN 1015-11 [56] at “TecnoLab” labora- tory by means of the CONTROLS 65-L1200 press; in this case, the load is uniformly trans- mitted through a 40 × 40 mm plate to the specimen (Figure 14a). Therefore, the equipment considers only the contact square area between the plate and the specimen, equal to 1600 mm2, to return the value of the compressive strength fc according to the formula (4):

= (4) Fibers 2021, 9, 30 17 of 20 where A is the square area section of the sample subject to compressive load and Fc the maximum force applied or compressive strength. The results are reported in Table 9.

Figure 14. UnconfinedUnconfined compression testtest onon M1–M4M1–M4 mortarmortar specimens,specimens, at at the the “TecnoLab” “TecnoLab” laboratory: labora- tory:(a) pre-test; (a) pre-test; and ( band) post-test. (b) post-test.

2.4.6.The Results tests Discussion were carried out according to UNI EN 1015-11 [56] at “TecnoLab” laboratory by meansThe mechanical of the CONTROLS characteristics 65-L1200 of press; the mortar in this case,were theevaluated load is uniformlyfrom the transmittedthree-point bendingthrough test a 40 and× 40the mm compression plate to thetest specimenon prismatic (Figure samples. 14a). Therefore, the equipment considersFrom onlyTable the 9, it contact can be square noticed area that between the compressive the plate and and bending the specimen, strength equal values to 2 1600increase mm as, tothe return amount the of value clay ofpresent the compressive in the mortar strength samplefc accordingincreases. In to thefact, formula the highest (4): F f = c (4) c A

where A is the square area section of the sample subject to compressive load and Fc the maximum force applied or compressive strength. The results are reported in Table9.

2.4.6. Results Discussion The mechanical characteristics of the mortar were evaluated from the three-point bending test and the compression test on prismatic samples. From Table9, it can be noticed that the compressive and bending strength values increase as the amount of clay present in the mortar sample increases. In fact, the highest values are obtained for M1 mortar with a percentage of clay equal to 21% and on the contrary, for M4 mortar, characterized by a 9% clay component, the lowest values were ob- tained. The results are in line with those found in the literature in Table2; in particular, they confirm the Barroso [57] trend. Furthermore, the compressive strength is approximately 2.5 times greater than the bending strength, which is in line with expectations. After the test campaign on the mortars, the values were compared with the limit values imposed by the NZS 4298 standards [12] on the earth mortar, which establish for mortars a minimum compressive strength of 1.3 MPa and a minimum bending strength of 0.25 MPa. The shrinkage test was also performed and showed that the mortar shrinks more when there is more water in the mix design. For this reason, even if the maximum strength values were recorded for M1, M1 and M2 mortars were not considered, as they did not reach the theoretical limits established in the literature [30], while M3 mortar was considered the most compatible both for its excellent workability and low shrinkage.

3. Conclusions In the present study, the mechanical properties of raw earth added to straw fibers for the pisé construction technique, and of a compatible earth mortar, were tested and evaluated. The goal was to create a consistent mix design for the compacted earth technique in the construction field, starting from the definition of the base material through USCS Fibers 2021, 9, 30 18 of 20

classification. The compressive and tensile strength values were obtained and compared with other experimental studies. Scientific literature and mechanical characterization of the specific design mixes with the addition of a fiber percentage for the pisé technique is very scarce and even absent. The present study tends to fill this gap. The results confirm that: • Fibers would increase the ductility of earth elements and the consequent mechanical production and compaction methods for the specimens, as occurred for the pisé walls; • The compaction method is reliable as the compressive strength values are very similar to each other and therefore constant, settling around a value consistent for the raw earth material; • Due to the presence of straw in the earth mix, the tensile strength can be considered of interest for future studies, also taking into account the lack of data; • The compatible earth mortar results comply and are consistent with all the require- ments in terms of workability, linear shrinkage and mechanical characteristics. Further advancements in the field of rammed earth walls are under development: they will be the subject of future studies complementary to the results here presented.

Author Contributions: Methodology, formal analysis, investigation, M.T. and C.F.; data curation, M.F.S.; writing—original draft preparation, M.F.S., M.T. and C.F.; Conceptualization, writing—review and editing, supervision, D.F. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Data Availability Statement: The data presented in this study are available in the article. Acknowledgments: “Madeinterra” cooperative in San Vito dei Normanni (Brindisi), the “Geotechni- cal Laboratory” of the Polytechnic University of Bari, “TecnoLab” laboratory in Altamura (Bari) and “M. Salvati” testing and materials laboratory of the Polytechnic University of Bari are acknowledged for the support given to the present research. Conflicts of Interest: The authors declare no conflict of interest.

References 1. Galdieri, E. Le Meraviglie Dell’architettura in Terra Cruda; Editori Laterza: Bari, Italy, 1982. (In Italian) 2. Houben, H.; Guillaud, H. Traité de Construction en Terre; Parentheses: Marseille, , 2006. (In French) 3. Guadagni, A. Prontuario Dell’ingegnere, 3rd ed.; Hoepli Editore: Milano, Italy, 2010. (In Italian) 4. Anger, R.; Fontaine, L. B“atir en Terre. Du Grain de Sable à L’architecture; Belin: Paris, France, 2009. (In French) 5. Houben, H.; Guillaud, H. Earth Construction: A Comprehensive Guide, 2nd ed.; Practical Action Publishing: Rugby, UK, 2008. 6. Bertagnin, M. Il Pisé e La Regola; Alinea: Firenze, Italy, 1981. (In Italian) 7. Bui, Q.B.; Morel, J.C.; Hans, S.; Meunier, N. Compression behaviour of non-industrial materials in civil engineering by three scale experiments: The case of rammed earth. Mater. Struct. 2009, 42, 1101–1116. [CrossRef] 8. Plataforma Arquitectura. Available online: https://www.plataformaarquitectura.cl/cl/801069/25-manuales-de-autoconstruccion- diseno-y-arquitectura-participativa-en-mexico-parte-1?ad_medium=gallery (accessed on 5 February 2021). 9. Vargas-Neumann, J. Earthquake resistant rammed earth (tapial) buildings. In Proceedings of the 7th International Conference on the Study and Conservation of Earthen Architecture TERRA 93, Silves, Portugal, 24–29 October 1993; pp. 140–151. 10. Röhlen, U.; Ziegert, C. Lehmbau-Praxis: Planung und Ausfv¨hrung, 3rd ed.; Beuth Praxis (Beuth Verlag GmbH): Berlin, Germany, 2020. (In German) 11. New Zealand Standards: NZS 4297. In Engineering Design of Earth Buildings; Standards New Zealand; The trading arm of theStandards Council: Wellington, New Zealand, 1998. 12. New Zealand Standards: NZS 4298. In Materials and Workmanship for Earth Building; Standards New Zealand; The trading arm of theStandards Council: Wellington, New Zealand, 1998. 13. Jaquin, P.A.; Augarde, C.E.; Gerrard, C.M. Analysis of Historic Rammed Earth construction. In Proceedings of the 5th International Conference on Structural Analysis of Historical Constructions SAHC 2006, New Delhi, India, 6–8 November 2006; pp. 1091–1098. 14. Jaquin, P.A.; Augarde, C.E.; Gerrard, C.M. Historic rammed earth structures in Spain: Construction techniques and a preliminary classification. In Proceedings of the International Symposium on Earthen Structures, Bangalore, India, 22–24 August 2007; pp. 327–333. Fibers 2021, 9, 30 19 of 20

15. Dierks, K.; Ziegert, C. Kapelle der Versöhnung—Schlussbericht zur Fremdv¨berwachtung Forschungs– und Seminarbericht des Fachgebietes Tragwerkslehre und Baukonstruktion der TU Berlin. 2000. (In German) Available online: https://www.tek.tu- berlin.de/menue/forschung/lehmbau/ (accessed on 17 January 2021). 16. Bui, T.T.; Bui, Q.B.; Limam, A.; Maximilien, S. Failure of rammed earth walls: From observations to quantifications. Constr. Build. Mater. 2014, 51, 295–302. [CrossRef] 17. Hall, M.; Djerbib, Y. Rammed earth sample production: Context, recommendations and consistency. Constr. Build Mater. 2014, 18, 281–286. [CrossRef] 18. Lilley, D.M.; Robinson, J. Ultimate strength of rammed earth walls with openings. Proceedings of the Institution of Civil Engineers (ICE). Struct. Build. 1995, 110, 278–287. [CrossRef] 19. Maniatidis, V.; Walker, P.; Heath, A.; Hayward, S. Mechanical and thermal characteristics of rammed earth. In Proceedings of the International symposium on Earthen Structures, Bangalore, India, 22–24 August 2007; pp. 205–211. 20. Maniatidis, V.; Walker, P. Structural capacity of rammed earth in compression. J. Mater. Civil. Eng. 2008, 20, 230–238. [CrossRef] 21. Miccoli, L.; Mv¨ller, U. Characterization of earthen materials. A comparison between earth block masonry, rammed earth and cob. In Proceedings of the 8th International conference on Structural Analysis of Historical Constructions SAHC 2012, Wroclaw, Poland, 15–17 October 2012; pp. 841–849. 22. Miccoli, L.; Mv¨ller, U.; Fontana, P. Mechanical behaviour of earthen materials: A comparison between earth block masonry, rammed earth and cob. Constr. Build Mater. 2014, 61, 327–339. [CrossRef] 23. Romanazzi, A.; Oliveira, D.V.; Silva, R.A. Experimental Investigation on the Bond Behavior of a Compatible TRM-based solution for rammed earth heritage. Int. J. Archit. Herit. 2019, 19, 1042–1060. [CrossRef] 24. Yamín Lacouture, L.E.; Phillips Bernal, C.; Reyes Ortiz, J.C.; Ruiz Valencia, D. Estudios de vulnerabilidad sismica rehabilitacion y refuerzo de casas en adobe y tapia pisada. Apunt. Rev. Estud. Sobre Patrim. Cult. J. Cult. Herit. Stud. 2007, 20, 286–303. (In Spanish) 25. De Felice, G.; De Santis, S.; Garmendia, L.; Ghiassi, B.; Larrinaga, P.; Lourenço, P.B.; Oliveira, D.V.; Fabrizio Paolacci, F.; Papanicolaou, C.G. Mortar-based systems for externally bonded strengthening of masonry. Mater. Struct. 2014, 47, 2021–2037. [CrossRef] 26. Borri, A.; Corradi, M. FRP: Fiber Reinforced Polymers. I Materiali Innovativi Nell’edilizia e Nel Restauro; Recupero e Conservazione; De Lettera: Milano, Italy, 2001. (In Italian) 27. Torrealva, D. Static and dynamic testing for validating the polymer grid as external reinforcement in earthen buildings. In Proceedings of the 16th International and Block Masonry Conference, Padova, Italy, 26–30 June 2016. 28. Fagone, M.; Loccarini, F.; Ranocchiai, G. Strength evaluation of jute fabric for the reinforcement of rammed earth structures. Compos. Part B Eng. 2017, 133, 1–13. [CrossRef] 29. Figueiredo, A.; Varum, H.; Costa, A.; Silveira, D.; Oliveira, C. Seismic retrofitting solution of an adobe masonry wall. Mater. Struct. 2013, 46, 203–219. [CrossRef] 30. Gomes, M.I.; Gonçalves, T.D.; Faria, P. The compatibility of earth-based repair mortars with rammed earth substrates. In Proceedings of the 3rd Historic Mortars Conference HMC13, Glasgow, UK, 11–14 September 2013. 31. Van Gorp, M.; Verstrynge, E.; Oliveira, D.V.; Silva, R.A.M.; Romanazzi, A. Experimental Study of the In-Plane Behavior of Rammed Earth, Strengthened with TRM. 2019. Available online: http://hdl.handle.net/1822/69268 (accessed on 17 January 2021). 32. Silva, R.A.; Oliveira, D.V.; Schueremans, L.; Miranda, T.; Machado, J. Shear behaviour of rammed earth walls repaired by means of grouting. In Proceedings of the 9th International Masonry Conference 2014 9IMC, Guimarães, Portugal, 7–9 July 2014. 33. Domínguez Martínez, O. Preservation and Repair of Rammed Earth Constructions. Master’s Thesis, Universidade do Minho, Braga, Portugal, 14 July 2015. [CrossRef] 34. Atzeni, C.; Achenza, M.; Sanna, U.; Mocci, S. Il Manuale Tematico Della Terra Cruda; Libreria dei tipografi del genio civile: Roma, Italy, 2008. (In Italian) 35. ASTM D422: Standard Test Method for Particle-Size Analysis of Soils, American Society for Testing and Materials; Withdrawn 2016; (ASTM) International: West Conshohocken, PA, USA, 2007. 36. ASTM D2487: Standard Practice for Classification of Soils for Engineering Purposes (USCS: Unified Soil Classification System); American Society for Testing and Materials (ASTM) International: West Conshohocken, PA, USA, 2017. 37. Casagrande, A. The determination of pre-consolidation load and it’s practical significance. Proc. Int. Conf. Soil Mech. Found. Eng. Cambridge Mass. 1936, 3, 60. 38. ASTM D4318: Standard Test Methods for Liquid Limit, Plastic Limit, and Plasticity Index of Soils; American Society for Testing and Materials (ASTM) International: West Conshohocken, PA, USA, 2017. 39. Minke, G. Building with Earth: Design and of a ; Birkhäuser-De Gruyter: Basel, Germany, 2006. 40. Fuller, W.B.; Thomson, S.E. The Laws of Proportioning . Trans. Am. Soc. Civ. Eng. 1907, 59, 67–143. [CrossRef] 41. Viñuales, G.M.; Martins Neves, C.M.; Flores, M.O.; Ríos, L.S. Arquitecturas de Tierra en Iberoamérica, Habiterra; Buenos Aires Libros: Buenos Aires, Argentina, 2003. (In Spanish) 42. Harmonized standard for Rammed Earth Structures: Code of Practice (SADCSTAN TC 1/SC 5/CD SAZS 724); African Organisation for Standardisation: Nairobi, Kenya, 2014. 43. Doat, P.; Hays, A.; Houben, H.; Matuk, S.; Vitoux, F. Building with Earth (CRATerre, France); The Village Society: New Delhi, India, 1991. 44. Houben, H.; Guillaud, H. Earth Construction: A Comprehensive Guide Technology, 1st ed.; ITDG Publishing: London, UK, 2004. Fibers 2021, 9, 30 20 of 20

45. LNEC. O Uso da Terra Como Material de Construção; Laboratório Nacional de Engenharia Civil: Lisbon, Portugal, 1953. (In Portuguese) 46. Walker, P.; Keable, R.; Martin, J.; Maniatidis, V. Rammed Earth, Design and Construction Guidelines; BRE Press: Watford, UK, 2005. 47. Röhlen, U.; Volhard, F. Lehmbau Regeln: Begriffe, Baustoffe, Bauteile, 3rd ed.; Springer Vieweg Verlag: Berlin, Germany, 2009. (In German) 48. Gonzalo, B.R.; Pilar, B.B. Monografías de la Direccioón General para la Vivienda y Arquitectura. In Bases Para el Diseno y Construccion con Tapial; Centro de Publicaciones: Madrid, Spain, 1992. 49. Instituto de Ciencias de la Construcción Eduardo Torroja, Madrid, Spagna. Available online: http://www.ietcc.csic.es/ (accessed on 17 January 2021). 50. Walker, P. Standards Australia. In The Australian Earth Building Handbook; HB 195-2002; Standards Australia International: Sydney, Australia, 2002. 51. ASTM D698: Standard Test Methods for Laboratory Compaction Characteristics of Soil Using Standard Effort; American Society for Testing and Materials (ASTM) International: West Conshohocken, PA, USA, 2012. 52. Vargas, J.; Bariola, J.; Blondet, M. Seismic strength of adobe masonry. Mater. Struct. 1986, 19, 253–258. [CrossRef] 53. Scudo, G.; Narici, B.; Talamo, C. Costruire Con La Terra. Tecniche Costruttive, Campi di Utilizzo e Prestazioni; Editori Esselibri: Napoli, Italy, 2001. (In Italian) 54. UNI EN 12390-6:2010. In Testing Hardened Concrete-Part 6: Tensile Splitting Strength of Test Specimens; European Standard; BSI: London, UK, 2019. 55. UNI EN 1015-11:2019. In Methods of Test for Mortar for Masonry-Part 11: Determination of Flexural and Compressive Strength of Hardened Mortar; European Standard; BSI: London, UK, 2019. 56. UNI EN 12390-3:2019. In Testing Hardened Concrete-Part 3: Compressive Strength of Test Specimens; European Standard; BSI: London, UK, 2019. 57. Barroso, A. Reforço Sìsmico Inovador de Construção de Taipa. Msc Thesis, Universidade do Minho, Guimarães, Portugal, 2017.