The Effect of Soil Mineral Composition on the Compressive Strength Of

materials

Article The Eﬀect of Soil Mineral Composition on the Compressive Strength of Cement Stabilized Rammed Earth

Piotr Narloch 1,* , Piotr Woyciechowski 1 , Jakub Kotowski 2 , Ireneusz Gawriuczenkow 2 and Emilia Wójcik 2 1 Faculty of Civil Engineering, Warsaw University of Technology, Al. Armii Ludowej 16, 00-637 Warsaw, Poland; [email protected] 2 Faculty of Geology, University of Warsaw, Zwirki˙ i Wigury 93, 02-089 Warsaw, Poland; [email protected] (J.K.); [email protected] (I.G.); [email protected] (E.W.) * Correspondence: [email protected]; Tel.: +48-691-660-184

Received: 5 December 2019; Accepted: 6 January 2020; Published: 10 January 2020

Abstract: Cemented stabilized rammed earth (CSRE) is a building material used to build load bearing walls from locally available soil. The article analyzes the influence of soil mineral composition on CSRE compressive strength. Compression tests of CSRE samples of various mineral compositions, but the same particle size distribution, water content, and cement content were conducted. Based on the compression strength results and analyzed SEM images, it was observed that even small changes in the mineral composition significantly affected the CSRE compressive strength. From the comparison of CSRE compressive strength result sets, one can draw general qualitative conclusions that montmorillonite lowered the compressive strength the most; beidellite also lowered it, but to a lesser extent. Kaolinite lightly increased the compressive strength.

Keywords: rammed earth; cement stabilized rammed earth; SEM images; mineral composition; compressive strength; sustainable building material

1. Introduction

1.1. An Overview of Cement Stabilized Rammed Earth Cement stabilized rammed earth (CSRE) is a construction material used to build load bearing walls from locally available soil found under the layer of humus. The technology of erecting walls from CSRE involves dynamic ramming of a moist soil-cement mixture in the formwork. After proper compaction of the layer, the next layers are added until the planned height of the element is reached (Figure1). CSRE is a sustainable building material. As a result of the Building Research Establishment Environmental Assessment Method (BREEAM), external building partitions consisting of a CSRE bearing layer obtained the highest A+ rating [1,2]. Portland cement is a frequently used stabilizer added to mixes, which increases the mechanical strength and durability of the material. It is an ideal stabilizer for coarse soils that have less expansive clay minerals. Soils that do not contain clay fractions ease the mixing and cement stabilization processes. The increase of the compressive strength of CSRE from mixes based on these types of soil is more than proportional to the cement added, thanks to which it is often possible to reduce the amount of cement needed significantly to achieve a compressive strength at the intended level [3]. With soils containing clay fractions, lower CSRE compressive strengths are achieved [3]. This is due to the fact that these soils are much more chemically diverse, and the reactions that occur significantly affect the properties of the material [4]. Furthermore, the

Materials 2020, 13, 324; doi:10.3390/ma13020324 www.mdpi.com/journal/materials Materials 2020, 13, 324 2 of 21

Materials 2020, 13, x FOR PEER REVIEW 2 of 21 rateFurthermore, of the increase the inrate compressive of the increase strength in compressiv in soils containinge strength clayin soils fractions containing is slower clay comparedfractions is to mixesslower consisting compared of onlyto mixes sand consisting and gravel of fractions only sand [4 and]. The gravel pozzolanic fractions reaction [4]. The in pozzolanic soils containing reaction clay mineralsin soils stimulates containing aclay long minerals term increase stimulates in compressive a long term strengthincrease in [4 ].compressive The compressive strength strength [4]. The of rammedcompressive earth fromstrength mixtures of rammed containing earth from clays mixtures can even containing be reduced clays by can added even be cement, reduced in by particular added whencement, less thanin particular 5% is added when [ 4less]. than 5% is added [4].

FigureFigure 1. Right:1. Right: a cementa cement stabilized stabilized rammed rammed earthearth (CSRE)(CSRE) wall. Left: Left: CSRE CSRE wall wall erection erection scheme. scheme. A: A: thethe formwork formwork is is ﬁlled filled with with a a layer layer of of moist moist soil-cement soil-cement mixture.mixture. B: B: ramming ramming the the mixture. mixture. C: C: filling ﬁlling thethe formwork formwork with with another another layer layer of of the the mixture. mixture. D:D: successivesuccessive layers layers of of moist moist mixture mixture are are added added and and compacted.compacted. E: E: the the formwork formwork is is removed. removed.

1.2.1.2. Factors Factors A ffAffectingecting the the Compressive Compressive Strength Strength ofof CSRECSRE TheThe compressive compressive strength strength of CSRE of CSRE is influenced is influenc by aed variety by a of variety mixture of properties mixture andproperties technological and features,technological such as: features, such as: − - mineralmineral composition composition of of the the soil soil [5[5];]; − density of elements [3,6,7]; - density of elements [3,6,7]; − building element shape [5]; - − buildingporosity element of elements shape [7]; [5 ]; - − porosityparticle of size elements distribution [7]; of the soil used in the mixture [3,6,8]; - − particleage of sizethe building distribution element of the [7,9,10]; soil used in the mixture [3,6,8]; - − agemixture of the compaction building element method [7 ,[11];9,10 ]; - − mixturemoisture compaction content of methodelement [during11]; service [6,12,13]; − - moisturemoisture content content of of element the soil-cement during servicemixture [ 6during,12,13 ];construction [3,13–17]; − content and type of Portland cement [7,16,17]; - moisture content of the soil-cement mixture during construction [3,13–17]; − the exposure conditions of the building element [6,18]; - content and type of Portland cement [7,16,17]; - theAmong exposure the conditionstechnological of thefeatures building mentioned element ab [6,ove,18]; focus has been placed on the mineral composition of soil. This is a property of soil that is difficult to predict in a construction setting and is Amongoften overlooked the technological in scientific features papers. For mentioned this reason, above, the authors focus decided has been to placedverify the on significance the mineral compositionof the impact of soil.of the This soil’s is mineral a property composition of soil that on is the diffi compressivecult to predict strength in a construction of CSRE. setting and is often overlooked in scientific papers. For this reason, the authors decided to verify the significance of the1.3. impact Soil Mineral of the soil’sComposition mineral Effect composition on the Compressive on the compressive Strength of CSRE strength of CSRE.

1.3.1.3.1. Soil Introduction Mineral Composition Eﬀect on the Compressive Strength of CSRE Cohesive soils are composed of clay and non-clay minerals. Among clay minerals, illite is the 1.3.1. Introduction predominant mineral in the subsurface of Central and Northern European soils (the part that had beenCohesive glaciated) soils [19]. are Present composed as well of clayare smectite and non-clay and kaolinite minerals. [19]. Among There are clay no minerals, differences illite in isthe the predominantqualitative mineralcomposition in the of subsurface clay minerals of Central found in and the Northern variously European aged tills soils in Central (the part and that Northern had been glaciated)Europe [[19].19]. PresentAmong asnon-clay well are minerals, smectite andthe kaolinitemost commonly [19]. There found are mineral no diﬀerences is quartz, in the which qualitative can compositionconstitute up of to clay 90% minerals of cohesive found soils in [19]. the variouslyIron compounds aged tills such in as Central goethite, and siderite, Northern and carbonates Europe [19 ]. Among(among non-clay others, minerals,calcite) are the also most frequently commonly present found [19]. mineral is quartz, which can constitute up to

Materials 2020, 13, 324 3 of 21

90%Materials of cohesive 2020, 13, soilsx FOR [PEER19]. REVIEW Iron compounds such as goethite, siderite, and carbonates (among3 others,of 21 calcite) are also frequently present [19]. 1.3.2. Clay Minerals 1.3.2. Clay Minerals Clay minerals differ from other minerals in their specific structure. They are hydrated aluminosilicatesClay minerals differwith a from characteristic, other minerals layered in their crystall specificine structure. structure. One They of arethe hydrated layers contains aluminosilicates silicate withtetrahedrons a characteristic, whose layered centers crystalline are Si4+ and structure. are surrounded One of the by layers four oxygen contains atoms. silicate These tetrahedrons tetrahedrons whose 4+ centersconnect are with Si andone another are surrounded to form bythe four hexagonal oxygen arrangement atoms. These of tetrahedrons Si2O2-5. The second connect layer with consists one another of 3+ 3+ to formoctahedrons the hexagonal where Al arrangement is surrounded of Si by2O six2-5 .hydroxyl The second groups. layer The consists layers of combine octahedrons with one where other Al in is surroundeddistinctive by arrangements six hydroxyl called groups. pack Theets layers that determine combine with their one properti otheres. in distinctiveThe most important arrangements features called packetsof clay that minerals determine include their their properties. cation exchange The most capacity important and features their ability of clay to minerals absorb water include [17]. their About cation exchange15 minerals capacity are ordinarily and their classified ability to as absorb clay minera water [ls,17 and]. About these 15belong minerals to three are main ordinarily groups: classified kaolin, as clayillite, minerals, and smectite and these [20]. belong These togroups three of main minerals groups: are kaolin, the most illite, commonly and smectite occurring [20]. Theseclay minerals groups of mineralsin soils are [21–23]. the most commonly occurring clay minerals in soils [21–23]. Kaolinite is the most widespread mineral in the kaolin group. Its structure consists of one Kaolinite is the most widespread mineral in the kaolin group. Its structure consists of one tetrahedral layer and one octahedral layer. The hydrogen bond is strong, and crucially, the tetrahedral layer and one octahedral layer. The hydrogen bond is strong, and crucially, the intra-lamellar intra-lamellar spacing is very small; hence, it is extremely difficult to separate the layers, and as a spacing is very small; hence, it is extremely difficult to separate the layers, and as a result, kaolinite is result, kaolinite is relatively stable and water less able to penetrate between the layers; thus, it exhibits relatively stable and water less able to penetrate between the layers; thus, it exhibits relatively little relatively little swell on wetting [24,25]. swell onIllite wetting is the [most24,25 ].common mineral from the illite group. Its structure consists of one octahedral layerIllite and is two the mosttetrahedral common layers. mineral It is characterized from the illite by group.a rigid structure, Its structure which consists is due of to onethe presence octahedral layerof K and+ cations two tetrahedral in the spaces layers. between It is the characterized packets. These by acations rigid structure,connect negatively which is duecharged to the surfaces, presence + ofcausing K cations the inillites the not spaces to swell between as much the packets.as smectites These [26]. cations Illites are connect classified negatively as non-swelling charged or surfaces, low causingswelling the minerals illites not [26]. to swell as much as smectites [26]. Illites are classified as non-swelling or low swellingIn mineralsthe smectite [26 ].group, the most widespread minerals are beidellite and montmorillonite. These minerals,In the smectitelike illite, group, are made the of most two widespreadtetrahedral la mineralsyers and an are octahedral beidellite layer and montmorillonite.bound between them. These minerals,The thickness like illite, of this are packet made ofis twoabout tetrahedral 1 nm (10 Å). layers Montmorillonites and an octahedral are characterized layer bound by between their high them. Thewater thickness absorption, of this which packet results is about mainly 1 nm from (10 Å).their Montmorillonites structure and from are the characterized existence of by interlayer their high watercations. absorption, Smectites which differ results from mainly illites fromin that their they structure have a and labile from structure, the existence i.e., ofinterlayer interlayer spaces cations. Smectitesmay expand differ [27]. from illites in that they have a labile structure, i.e., interlayer spaces may expand [27]. TheThe swelling swelling potential potential (Figure (Figure2 )2) of of the the minerals minerals describeddescribed above indicates indicates that that soils soils containing containing eveneven small small amounts amounts of of smectites smectites will will swellswell intensively,intensively, while soils soils with with illite illite or or kaolinite kaolinite will will not. not. In In thethe opinion opinion of of the the authors, authors, it it is is therefore therefore notnot possiblepossible to treat treat the the clay clay faction faction in in soil soil as as always always having having a detrimental effect on the compressive strength of CSRE. This approach would indicate that soil a detrimental effect on the compressive strength of CSRE. This approach would indicate that soil without clay fractions (e.g., sand with cement) will not necessarily have a higher compressive without clay fractions (e.g., sand with cement) will not necessarily have a higher compressive strength strength compared to soil containing clay fractions. compared to soil containing clay fractions.

Figure 2. Relationship between swelling potential and clay fraction percentage (after Seed et al.) [28]. Figure 2. Relationship between swelling potential and clay fraction percentage (after Seed et al.) [28].

MaterialsMaterials 20202020,, 1313,, x 324 FOR PEER REVIEW 44 of of 21 21

The studies described in [4] showed that soil mixes containing a large quantity of kaolinite (>10%)The obtain studies higher described compressive in [4] showed strengths that compared soil mixes to containing sand and a cement large quantity mixes (Figure of kaolinite 3). In (> turn,10%) soilobtain mixes higher with compressive montmorillonite strengths obtain compared lower tocompressive sand and cementstrengths mixes compared (Figure 3to). Inmixes turn, with soil kaolinitemixes with [29]. montmorillonite Interestingly, obtainmixes lower with compressiveexpansive bentonite strengths comparedhave obtained to mixes similar with kaolinitecompressive [29]. strengthsInterestingly, to mixes mixes with with sand expansive alone. bentoniteTest results have also obtained indicate similar that granulation compressive has strengths a key impact to mixes on compressivewith sand alone. strength. Test resultsThe CSRE also soil indicate mix composed that granulation of 60% silty has aloam key impactand 40% on sand compressive obtained a strength. higher strengthThe CSRE than soil the mix rest. composed of 60% silty loam and 40% sand obtained a higher strength than the rest.

2 Compressive strength (MPa) Compressive strength 1

0 Sand 0-4 Bentonite:Sand = 1:9 Kaolinite:Sand = 1:9 Silty loam: Sand = 6:4

Figure 3. Compressive strengths of adobe bricks with the addition of 6% cement. Based on [4]. Figure 3. Compressive strengths of adobe bricks with the addition of 6% cement. Based on [4]. 1.3.3. Non-Clay Minerals 1.3.3. Non-Clay Minerals Quartz, SiO2, is among the most common minerals found in the Earth’s crust. It occurs in many diverseQuartz, varieties, SiO2, includingis among inthe the most form common of coated minerals grains found in sands, in the as well Earth’s as in crust. cohesive It occurs soils in where many it diverseforms aggregates varieties, including and microaggregates in the form with of coated clay minerals. grains in Itsands, is characterized as well as in by cohesive a high hardness soils where (seven it formson the aggregates Mohs Hardness and microaggregates Scale) and a density with of clay 2.65 minerals. g/cm3. These It is properties characterized increase by a the high compressive hardness 3 (sevenstrength on of the CSRE Mohs [5]. Hardness It is a mineral Scale) that and does a notdensity react of with 2.65 water. g/cm . These properties increase the compressiveAnother strength common of mineral CSRE [5]. is goethite, It is a mineral FeO(OH). that It does is a mineral not react found with in water. sedimentary rocks, usually occurringAnother in compact,common mineral granular, is and goethite, earthy FeO(OH). clusters. It It is a less mineral hard thanfound quartz in sedimentary (5.5 on the rocks, Mohs usuallyHardness occurring Scale), and in becausecompact, of granular, its high iron and content earthy (62.8%), clusters. it hasIt is a less notably hard high than density quartz (4–4.4 (5.5 on g/cm the3). MohsThis may Hardness influence Scale), the bulkand because density ofof aits CSRE high element. iron content An important (62.8%), it feature has a notably of goethite high is itsdensity color (4–4.4 g/cm3). This may influence the bulk density of a CSRE element. An important feature of (usually, various shades of brown), often having an effect on the color of CSRE. Siderite, FeCO3, occurs goethitemainly in is aits compact, color (usually, earthy, orvarious very fine shades crystalline of brown), cluster often form. having It has an a lowereffect hardnesson the color than of goethite CSRE. Siderite,(4–4.5 on FeCO the Mohs3, occurs Hardness mainly Scale), in a compact, and because earthy, it has or very a smaller fine crystalline iron content cluster (48.3%), form. it therefore It has a lower has a hardnesssmaller density than goethite (3.7–3.9 (4–4.5 g/cm3 on). Itthe usually Mohs hasHardness a dark-yellow, Scale), and orange because color. it Likehas a the smaller non-clay iron minerals content 3 (48.3%),mentioned it therefore above, it has does a notsmaller absorb density water. (3.7–3.9 Goethite g/cm and). siderite It usually in thehas presence a dark-yellow, of water orange are stable color. at Like the non-clay minerals mentioned above, it does not absorb water. Goethite and siderite in the temperatures from 0 to 100 ◦C. This means that they will not react with water [30]. presence of water are stable at temperatures from 0 to 100 °C. This means that they will not react Calcite, CaCO3, is an important rock forming mineral. It creates very diverse forms of clusters: withgranular, water compacted, [30] and fibrous. This mineral is the main component in sedimentary limestone rocks, bothCalcite, organic CaCO and chemical3, is an important in origin. rock In soils, forming calcite mineral. plays aIt critical creates role, very a ffdiverseecting forms a number of clusters: of their granular,properties, compacted, also acting and structurally. fibrous. ItThis has mineral a relatively is the low main hardness component (3.0 on in the sedimentary Mohs Hardness limestone Scale), rocks,similar both to clay organic minerals. and chemical The density in origin. of calcite In soils, varies calcite from plays 2.6 to a 2.8 critical g/cm role,3. It doesaffecting not absorba number water, of theirbut it properties, easily dissolves also acting in acid structurally. (e.g., carbonic It has acid). a relatively low hardness (3.0 on the Mohs Hardness Scale), similar to clay minerals. The density of calcite varies from 2.6 to 2.8 g/cm3. It does not absorb water,1.4. Cement but it Stabilizationeasily dissolves in acid (e.g., carbonic acid). As mentioned above, each mineral behaves differently in contact with water, which in turn has an effect on cement binding. Cement stabilization for CSRE is best suited for soils with less expansive

Materials 2020, 13, x FOR PEER REVIEW 5 of 21

1.4. Cement Stabilization As mentioned above, each mineral behaves differently in contact with water, which in turn has an effect on cement binding. Cement stabilization for CSRE is best suited for soils with less expansive minerals.Materials 2020 Unlike, 13, 324 concrete, where early stabilization is caused by cement, in CSRE, there is a 5long of 21 lasting curing of the material thanks to which the strength continues to grow past 28 days [4,31,32].

2.minerals. Materials Unlike and Methods concrete, where early stabilization is caused by cement, in CSRE, there is a long lasting curing of the material thanks to which the strength continues to grow past 28 days [4,31,32]. 2.1. Materials 2. Materials and Methods The study examined the relationship between the mineral composition of soil-cement mixtures and2.1. Materialsthe compressive strength of the CSRE. Soil mixtures with the same particle size curves were prepared. Accordingly, soil mixtures consisting of sieved sand and gravel fractions, as well as seven The study examined the relationship between the mineral composition of soil-cement mixtures different loams were used. In terms of mineral composition, sand was composed of pure quartz. and the compressive strength of the CSRE. Soil mixtures with the same particle size curves were Gravel fractions were composed of quartz (75% by mass) and carbonate crumbs (25% by mass). The prepared. Accordingly, soil mixtures consisting of sieved sand and gravel fractions, as well as seven clays used in the study differed in both mineral composition (Table 1), as well as granulation diﬀerent loams were used. In terms of mineral composition, sand was composed of pure quartz. (Figure 4). Gravel fractions were composed of quartz (75% by mass) and carbonate crumbs (25% by mass). The clays used in the study diﬀeredTable in both 1. Mineral mineral composition composition of each (Table loam1), (%). as well as granulation (Figure4).

Loam Table 1. Mineral composition of each loam (%). Organic Quartz and Montmorillonite Beidellite Kaolinite Illite Goethite Siderite Calcite Symbol Substance Others Loam Organic Quartz and II Montmorillonite- Beidellite8.6 Kaolinite1.3 Illite2.8 Goethite1.0 Siderite- 22.7 Calcite 0.6 63.0 Symbol Substance Others VII - 8.6 4.7 - 3.0 2.8 - 1.3 79.6 II - 8.6 1.3 2.8 1.0 - 22.7 0.6 63.0 VIIXI -- 8.68.9 4.73.1 29.8 - - 3.0 4.1 2.8 - -- 1.3 79.654.1 IIIXI -- 8.921.7 3.16.1 29.8- 3.0 - - 4.1 43 -0.3 - 54.125.9 IVIII -- 21.7 - 6.172.7 -- 1.0 3.0 - -- 430.4 0.3 25.9 IVV -- - - 72.779.8 -- - 1.0 - -- -0.5 0.4 25.919.7 V - - 79.8 - - - - 0.5 19.7 X 10 10 18.1 18.1 27.227.2 11.3 - -4.3 4.3 - - 1.4 1.4 27.6

Figure 4. Granulation curves of selected clays.

By adding an appropriate amount of dry ingredients (clays, as well as sand and gravel fractions; Table 2), mixtures were obtained with granulation curves as shown in Figure 5. These mixtures simulated two potential soils that could be used as the main component of CSRE. These mixtures differed first and foremost in the proportion of clay fractions. The MC mixture consisted of a 16% Materialsclay fraction,2020, 13 15%, 324 silt fraction, 39% sand fraction, and 30% gravel fraction. Meanwhile, the LC mixture6 of 21 consisted of a 4% clay fraction, 18% silt fraction, 48% sand fraction, and 30% gravel fraction. The LC mixtures were designed with clays II, VII, and XI, while the MC mixtures with clays III, IV, V, and X. consisted of a 4% clay fraction, 18% silt fraction, 48% sand fraction, and 30% gravel fraction. The LC mixtures were designedTable with 2. clays Percentage II, VII, mass and XI,compositions while the of MC each mixtures soil mixture. with clays III, IV, V, and X.

(%) Mass Table 2. Percentage(%) mass Mass compositions of Silt and ofSand each Fractions soil mixture. of Gravel (%) Mass (%) Mass FSi MSi(%) Mass of SiltCSi and Sand FractionsFSa MSa CSa FGr (%)of Mass Loam of Gravel (0.002; 0063) (0.0063; 0.02) (0.02; 0.063) (0.063; 0.2) (0.2; 0.63) (0.63; 2.0) (2.0; 4.0) of Loam FSi MSi CSi FSa MSa CSa FGr Particle Size Distribution

Loam Symbol (mm) (mm) (mm) (mm) (mm) (mm) (mm) (0.002; 0063) (0.0063; 0.02) (0.02; 0.063) (0.063; 0.2) (0.2; 0.63) (0.63; 2.0) (2.0; 4.0) Particle Size Distribution II 30.1 0.0 1.9 1.0 0.0 27.7 9.2 30.0 Loam Symbol (mm) (mm) (mm) (mm) (mm) (mm) (mm) LC VIIII 30.125.8 0.0 3.0 1.9 1.5 1.0 0.3 0.0 3.2 27.728.4 9.27.7 30.030.0 LC XIVII 25.811.3 3.0 3.0 1.5 5.3 0.3 5.3 3.2 5.8 28.429.4 7.79.9 30.030.0 XI 11.3 3.0 5.3 5.3 5.8 29.4 9.9 30.0 III 30.9 3.8 1.6 0.2 1.8 24.6 7.1 30.0 IVIII 30.929.8 3.8 0.0 1.6 0.2 0.2 2.3 1.8 4.0 24.625.7 7.17.9 30.030.0 IV 29.8 0.0 0.2 2.3 4.0 25.7 7.9 30.0 MCMC VV 26.326.3 0.1 0.1 2.7 2.7 2.9 2.9 4.1 4.1 25.825.8 8.18.1 30.030.0 X 25.225.2 2.9 2.9 1.8 1.8 1.8 1.8 4.3 4.3 25.925.9 8.18.1 30.030.0

Figure 5. Particle size distribution of soil mixtures used in the CSRE compressive strengthstrength tests.tests.

Portland cement is a frequently usedused stabilizerstabilizer forfor rammedrammed earthearth mixtures.mixtures. Due to its hydration withwith water,water, itit signiﬁcantlysignificantly increasesincreases thethe compressivecompressive strength of CSRE.CSRE. In [[3],3], itit waswas shownshown thatthat for rammedrammed earth samples (from soil mixtures containingcontaining 30% clay and silt, 40% sand, and 30% gravel, withwith optimumoptimum humidity, humidity, seasoned seasoned for 28for days), 28 days), by adding by adding 9% cement, 9% cement, compressive compressive strength increasedstrength fromincreased about from 2.68 about MPa to 2.68 about MPa 9.30 to MPa.about Cement 9.30 MPa. CEM Cement I 42.5 RCEM (Table I 42.53) was R (Table added 3) to was these added mixtures, to these in themixtures, amount in of the 6% amount and 9% of by 6% mass and of 9% the by remaining mass of the dry remaining ingredients. dry These ingredients. values wereThese chosen values based were onchosen the authors’ based on own the authors’ previous own study, previous which foundstudy, thatwhich 6% found cement that ensured 6% cement an adequate ensured compressivean adequate strengthcompressive for the strength CSRE [3 ,5for,30 ],the while CSRE 9% cement [3,5,30] was, while needed 9% in regardscement to durabilitywas needed [33 –37in]. regards to durability [33–37].

Materials 2020, 13, 324 7 of 21

Table 3. Chemical composition of Portland Cement CEM I 42.5 R.

Constituent % by Weight

SiO2 20.6 Al2O3 6.2 Fe2O3 2.9 CaO 64 MgO 1.4 Na2O 0.7 SO3 2.5 N2O 0.05 Cl- 0.04 Loss of ignition 2.5

To the 14 soil-cement mixtures (Table4) that were obtained in this way, water was added, ensuring that the optimum moisture content (OMC) of the mixture was achieved, i.e., the moisture at which compaction by ramming achieved the maximum value of dry density (the method of determining the OMC is given in Section 2.2.2). The OMC is the water content at which a soil can be compacted to the maximum dry density by a given compactive eﬀort [38]. Therefore, OMC is a direct measure of CSRE workability.

Table 4. List of mixtures from which CSRE samples were prepared.

Particle Size Distribution Loam Used in the Cement Addition Mixture Symbol Water Content (%) Curve Mixture (%) LC II 6% 6 II LC II 9% 9 LC VII 6% 7 6 LC VII LC VII 9% 9 LC XI 6% 6 XI LC XI 9% 9 MC III 6% 6 III MC III 9% 9 MC IV 6% 6 IV MC IV 9% MC 8 9 MC V 6% 6 V MC V 9% 9 MC X 6% 6 X MC X 9% 9

The workability of the CSRE mixture is the ability to accurately fill molds while maintaining its uniformity and tightness and the use of a specific compaction method. This is an important feature of the mixture, affecting the mechanical strength and durability of the material, as well as the work needed to erect vertical partitions [4]. Similar to the concrete mixture, the workability of the CSRE mixture will be influenced, among others, by the amount of water added to the mixture, the grain size of the soil used for the mixture, the shape and texture of the aggregate grains, the content of cement, or other stabilizers, admixtures, and additives used [39]. Proper workability of the CSRE mixture is a condition for its effective compacting. In CSRE technology, compaction occurs by ramming a moist soil mixture in the formwork. The way the mixture is compacted plays an important role in shaping the appropriate compressive strength of the material [4]. For rammed earth, no consistency test methods shall be used [4,14,16]. In this case, the measure of workability is the OMC [40]. For this reason, the assessment of whether the mixture has adequate workability will be affected by both the ramming energy and the thickness of the compacted material layers. Materials 2020, 13, x FOR PEER REVIEW 8 of 21 mixes (LC) was 7% and for higher clay content mixes (MC) was 8%, regardless of the mineral composition and cement addition in the mixtures.

Preparation of Samples Materials 2020, 13, 324 8 of 21 Clay, sieved sand fractions, and gravel were mixed together in a dry state. Then, cement was added and everything mixed again. Finally, water was poured. After reaching a uniform consistency, 100 ×The 100 OMC × 100 for mm a speciﬁccubic samples mixture were design formed. is based The on theformation maximum of the dry samples density atwas which carried maximum out by compressiveramming the strengthmixture in is three achieved equal [3 layers]. Determining using a 6.5 the kg OMCrammer. of theThe soil-cement samples were mixture formed guaranties by freely thelowering proper the workability. rammer from It isa height one of of the 30 key cm issuesto the insurface CSRE of technology. the moist mixture The OMC filled for in lower the form. clay contentSamples mixes were (LC)preliminarily was 7% and formed for higher at a height clay content of 105 mixes mm. (MC)The day was before 8%, regardless planned of compressive the mineral compositionstrength testing, and they cement were addition cut to a in height the mixtures. of 100 mm (Figure 6). The removed 5 mm slabs slowly dried to an air-dry state (20 °C and 50% RH). In this state, Preparation of Samples polished thin sections for SEM analyzes were made. Samples were not artificially dried in dryers at any preparatoryClay, sieved stage. sand fractions,Therefore, and the gravelprocess were of the mixed preparation together of in samples a dry state. for SEM Then, testing cement did wasnot addedsignificantly and everything affect the mixed observed again. cracks, Finally, which, water according was poured. to the After authors, reaching were a uniform largely consistency, caused by 100 100 100 mm cubic samples were formed. The formation of the samples was carried out by natural,× slow× drying, such as occurs in construction conditions. rammingSince the the mixture rammed in earth three equalkeeps layersthe layered using astructur 6.5 kg rammer.e, the samples The samples in the werecompressive formed bystrength freely loweringtest were loaded the rammer in the fromdirection a height of ramming. of 30 cm For to each the surface series, 10 of samples the moist were mixture prepared. ﬁlled The in the samples form. Sampleswere demolded were preliminarily after 24 h. Then, formed they were at a heightcured for of 10528 days mm. in Thea condition day before of a high planned relative compressive humidity strengthof 95% (±2%) testing, and they a temperature were cut to of a height20 °C (±1 of 100°C). mm (Figure6).

Figure 6. Part of the CSRE samples on which the compressive strength tests were made.

2.2. MethodsThe removed 5 mm slabs slowly dried to an air-dry state (20 ◦C and 50% RH). In this state, polished thin sections for SEM analyzes were made. Samples were not artificially dried in dryers at2.2.1. any The preparatory Method for stage. Determining Therefore, the the OMC process of Soil-Cement of the preparation Mixtures of samples for SEM testing did not significantly affect the observed cracks, which, according to the authors, were largely caused by The OMC of a soil-cement mixture is dependent on the particle size distribution of the soil, natural, slow drying, such as occurs in construction conditions. cement content in the mixture, and mineral composition of the soil. With small amounts of clay Since the rammed earth keeps the layered structure, the samples in the compressive strength test fractions, the mineral compositions of the mixtures would not change much. Soil particle size and were loaded in the direction of ramming. For each series, 10 samples were prepared. The samples cement addition have more impact [32]. The method and energy of ramming would also have an were demolded after 24 h. Then, they were cured for 28 days in a condition of a high relative humidity effect on this value. For this reason, a method to determine the optimum moisture content of soil of 95% ( 2%) and a temperature of 20 C( 1 C). mixtures± used for preparing rammed ◦earth± samples◦ for compressive strength test proposed in [39] 2.2.was Methodsused. This method is based on rammed earth soil compaction method given by NZS 4298: 1998 [40]. According to it, the moisture of a soil mixture for which the maximum dry 2.2.1.density The is Methodobtained for should Determining be determined the OMC using of Soil-Cementthe ramming Mixtures method such as the samples that were compacted. This in turn should reflect the nature and force of the dynamic compaction used for The OMC of a soil-cement mixture is dependent on the particle size distribution of the soil, cement building rammed earth walls. The paper adopts the method of ramming samples described in content in the mixture, and mineral composition of the soil. With small amounts of clay fractions, the mineral compositions of the mixtures would not change much. Soil particle size and cement addition have more impact [32]. The method and energy of ramming would also have an effect on this value. For this reason, a method to determine the optimum moisture content of soil mixtures used for preparing rammed earth samples for compressive strength test proposed in [39] was used. This Materials 2020, 13, 324 9 of 21 method is based on rammed earth soil compaction method given by NZS 4298: 1998 [40]. According to it, the moisture of a soil mixture for which the maximum dry density is obtained should be determined using the ramming method such as the samples that were compacted. This in turn should reflect the nature and force of the dynamic compaction used for building rammed earth walls. The paper adopts the method of ramming samples described in NZS 4298: 1998 [40]. The maximum dry density of samples from the mixtures was determined for this compaction method. The components of the soil mixture were dried to a constant mass and then mixed with each other. Then, water was added to obtain the OMC. The mold was weighed before measuring the density of the mixture. Next, ramming was carried out as specified in p. 2.1.1. After completion, the excess soil mixture above the mold surface was removed and then weighed. Then, the mass of the sample was calculated. Knowing the volume of the mold, the bulk density of the soil mixture ρ was calculated for a given moisture: m ρ = m (1) V where: mm: sample soil mass in (kg) V: volume of the sample in (m3) Knowing the bulk density of the soil mixture, ρ, the dry density, ρd, was calculated:

100ρ ρ = (2) d 100 + w

where: kg : bulk density of soil mixture ρ m3 W: moisture of sample mixture (%) For every soil mix of a given moisture content, measurements of dry density were carried out on three samples; the arithmetic mean of the measurements was taken as the value of the dry density.

2.2.2. Methods for Determining the Granulation of Loams In order to determine the grain size distribution of each loam, an aerometric analysis was carried out for each of them in accordance with standard PN-B-04481:1988 [41]. In this analysis, the relationship between the falling speed of particles in water and their diameter was used to determine the granulometric composition. Aerometric analysis was used in determining the content of soil particles with equivalent diameters less than 0.06 mm in cohesive soils [20].

2.2.3. SEM Methodology The SEM-EDS analysis was carried out in order to identify the mineral composition precisely and to present the texture of the samples in the micro-area. Electron microscopy combined with X-ray microanalysis gave the possibility of precise identification of the chemical composition and thus also the mineralogical of the samples tested. The SEM method allowed the observation of the surface of the sample at high magnifications, up to 100,000 . High magnifications allowed observing clusters and × aggregates of clay minerals and the clay-silt matrix between the mineral framework. Observations and analyses of aggregate minerals were made by scanning electron microscopy (SEM) at the National Multidisciplinary Laboratory of Functional Nanomaterials at the Faculty of Geology, University of Warsaw. Thin carbon sputtered (20 nm) sections were examined using the ZEISS Sigma VP FE-SEM (scanning electron microscope with field emission, Carl Zeiss AG, Oberkochen, Germany) equipped with a dispersive X-ray spectrometer (EDS). To record the spectra, the 20 kV electron beam’s acceleration voltage was used. To obtain the maximum current, required for high contrast imaging and element mapping, a 120 µm aperture was chosen, as well as an optional, high-current Zeiss mode. Mappings (spatial arrangement) of selected chemical elements were carried out at the above mentioned conditions and a time of counting of 10 min/frame. Materials 2020, 13, x FOR PEER REVIEW 10 of 21 Materials 2020, 13, 324 10 of 21 2.2.4. CSRE Compressive Strength Test 2.2.4.Compressive CSRE Compressive strength Strength tests on Test CSRE samples were carried out in a testing machine with two test modules:Compressive with 0–3000 strength kN and tests 0–250 on CSRE kN samplesranges of were measurement, carried out inwith a testing a measurement machine with error two of testless modules:than 1%. The with tests 0–3000 were kN carried and out 0–250 in a kN similar ranges fash ofion measurement, to the test method with afor measurement the compressive error strength of less thanof concrete 1%. The samples tests were given carried in standard out in aEN similar 12390-3 fashion [42]. toThe the only test difference method for was the the compressive method of strength placing ofsamples concrete in samplesthe strength given testing in standard machine. EN Rammed 12390-3 [42 earth,]. The laid only and diff compactederence was in the layers, method retained of placing this sampleslayered instructure the strength in both testing a machine.monolithic Rammed wall and earth, in laidmolded and compacted samples. inDue layers, to the retained material this layeredcharacteristics, structure as in well both as a the monolithic forming wallmethod, and init was molded considered samples. representati Due to theve material of rammed characteristics, earth that asthe well method as the of forming testing method, the compressive it was considered strength representative was by loading of rammed the sample earth in that the the direction method ofof testingits formation. the compressive strength was by loading the sample in the direction of its formation.

2.2.5.2.2.5. Moisture Content of CSRE Samples afterafter thethe CuringCuring PeriodPeriod AfterAfter testingtesting the the compressive compressive strength, strength, the crushedthe crushed fragments fragments of one of CSRE one sample CSRE weresample collected were fromcollected each from of the each 14 tested of the series.14 tested The series. fragments The fragments were weighed were andweighed then driedand then to a dried dry mass to a atdry 105 mass◦C. Massat 105 measurements °C. Mass measurements were done everywere done 24 h. every When 24 two h. consecutiveWhen two consecutive measurements measurements diﬀered by less differed than 1by g, less the than elements 1 g, the were elements considered were toconsidered have reached to have their reached dry mass. their Knowing dry mass. the Knowing dry mass the of dry crushed mass CSREof crushed sample, CSRE the sample, moisture the content moisture of the content CSRE of (w) the was CSRE calculated. (w) was calculated.

𝑚 −𝑚 𝑤 =m humid mdry (3) w = 𝑚− (3) mdry where: where:𝑚: mass of the humid crushed CSRE sample (kg); m𝑚humid: mass: mass of of the the dry humid crushed crushed CSRE CSRE sample sample (kg). (kg); mdry: mass of the dry crushed CSRE sample (kg). 3. Results 3. Results 3.1. CSRE Compressive Strength and Moisture Content 3.1. CSRE Compressive Strength and Moisture Content The compressive strength results after the curing period for all tested CSRE series are shown in The compressive strength results after the curing period for all tested CSRE series are shown in Figure 7, Figure 8, and Table 5. The average moisture content for all series is shown in Figure 9 and Figure7, Figure8, and Table5. The average moisture content for all series is shown in Figure9 and Table 5. By comparing the results from the LC and MC series, it can be seen that granulation had a Table5. By comparing the results from the LC and MC series, it can be seen that granulation had a key influence on compressive strength. The average moisture contents of the MC sample series (with key influence on compressive strength. The average moisture contents of the MC sample series (with more clay) were higher than those of the LC sample series (with less clay). As expected, significant more clay) were higher than those of the LC sample series (with less clay). As expected, significant differences were observed between the moisture content of samples with the same granulation, but differences were observed between the moisture content of samples with the same granulation, but different mineral composition. different mineral composition.

16.0 14.0 12.0 10.0 8.0 [MPa] 6.0 4.0 2.0 Average compressive strength Average compressive 0.0 LC VII LC XI LC II MC III MC IV MC V MC X

Figure 7. Compressive strength of the CSRE sample series with 9% cement content. Horizontal Figure 7. Compressive strength of the CSRE sample series with 9% cement content. Horizontal lines lines mark the minimum and maximum value in the sample series, while the round points mean the mark the minimum and maximum value in the sample series, while the round points mean the average value. average value.

MaterialsMaterials 20202020, ,1313, ,x 324FOR PEER REVIEW 1111 of of 21 21

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0 Average compressive strength [MPa] strength Average compressive 0.0 LC VII LC XI LC II MC III MC IV MC V MC X

FigureFigure 8. 8. CompressiveCompressive strength strength of the of theCSRE CSRE sample sample series series with with 6% cement 6% cement content. content. Horizontal Horizontal lines marklines the mark minimum the minimum and maximum and maximum value value in the in thesample sample series, series, while while the the round round points points mean mean the the averageaverage value. value. Table 5. Mineral compositions of soil mixtures given in percentages. The table also gives the percentage of5.0% water and CEM I 42.5 R cement additions and the average obtained compressive strength of CSRE samples in each series.

4.0% (%) (%) (%) 3.0% (%) Illite (%) Calcite (%) Siderite (%) Goethite (%) Compressive Kaolinite (%) Beidellite (%) Strength (MPa) Mixture Symbol Montmorillonite Cement Addition Water Content (%) Quartz and Others 2.0% Organic Substance LC II 6% 6 6.43 0.0 0.0 7 LC II 9% 2.6 0.4 0.8 0.3 6.8 0.2 88.9 9 14.09 LC[%] content Moisture VII 6% 6 11.07 1.0% 0.0 0.0 0.0 7 LC VII 9% 2.3 1.2 0.8 0.7 0.3 94.6 9 13.54 LC XI 6% 6 8.01 0.0 0.0 0.0 0.0 7 LC XI 9% 1.8 0.4 2.7 0.5 94.6 9 13.26 0.0% MC III 6% 6 6.96 LC0.0 VII LC XI LC0.0 II0.0 MC III MC IV MC V MC8 X MC III 9% 6.6 1.9 0.9 13.1 0.1 77.3 9 8.76 MC IV 6% 9%CEM 6%CEM 6 4.51 0.0 0.0 0.0 0.0 0.0 8 MC IV 9% 21.8 0.3 0.1 77.8 9 5.63 FigureMC V 6%9. Average moisture content of each CSRE sample series during the compressive strength6 tests.4.30 0.0 0.0 0.0 0.0 0.0 0.0 8 MC V 9% 21.1 0.1 78.7 9 5.45 MC X 6% 6 3.60 The MC III mixture contained the least amount0.0 of clay minerals.0.0 Besides beidellite and8 kaolinite, MC X 9% 3.0 4.1 6.9 2.9 1.1 0.4 81.7 9 4.27 the most likely mineral to comprise the clay fraction was calcite. For this reason, CSRE made from Legend (%) 0.1–1 1–5 5–10 10–25 >25 this mixture had the highest compressive strength of those in the CSRE group containing a high clay content. Mixtures MC IV and MC V were characterized by their almost identical mineral composition; The MC III mixture contained the least amount of clay minerals. Besides beidellite and kaolinite, hence, they obtained similar compressive strength results. These almost identical results confirmed the most likely mineral to comprise the clay fraction was calcite. For this reason, CSRE made from that the mineral composition affected the strength of the samples. The lack of swelling minerals this mixture had the highest compressive strength of those in the CSRE group containing a high clay (smectites, i.e., beidellite and montmorillonite) in the MC IV and MC V mixes resulted in a noticeably content. Mixtures MC IV and MC V were characterized by their almost identical mineral composition; higher compressive strength compared to the MC X mix. The MC X mixture was the only mix hence, they obtained similar compressive strength results. These almost identical results conﬁrmed that containing montmorillonite, a highly swelling mineral with high hydrophilicity. For this reason, the mineral composition aﬀected the strength of the samples. The lack of swelling minerals (smectites, CSRE samples from this mixture had the highest moisture content at the time of testing and i.e., beidellite and montmorillonite) in the MC IV and MC V mixes resulted in a noticeably higher consequently the lowest compressive strength. The MC III mix contained 6.6% beidellite, which was compressive strength compared to the MC X mix. The MC X mixture was the only mix containing also a swelling mineral. For this reason, this mixture was characterized by a higher moisture content, montmorillonite, a highly swelling mineral with high hydrophilicity. For this reason, CSRE samples compared to mixes MC IV and MC V. from this mixture had the highest moisture content at the time of testing and consequently the lowest

Materials 2020, 13, x FOR PEER REVIEW 11 of 21

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0 Average compressive strength [MPa] strength Average compressive 0.0 Materials 2020 13 , , 324LC VII LC XI LC II MC III MC IV MC V MC X12 of 21

Figure 8. Compressive strength of the CSRE sample series with 6% cement content. Horizontal lines compressive strength. The MC III mix contained 6.6% beidellite, which was also a swelling mineral. mark the minimum and maximum value in the sample series, while the round points mean the For this reason, this mixture was characterized by a higher moisture content, compared to mixes MC average value. IV and MC V.

5.0%

4.0%

3.0%

2.0% Moisture content [%] content Moisture 1.0%

0.0% LC VII LC XI LC II MC III MC IV MC V MC X

9%CEM 6%CEM FigureFigure 9. 9. AverageAverage moisture moisture content content of of each each CSRE CSRE sample sample series series during during the the compressive compressive strength strength tests. tests.

Figure 10 presents the compressive strength results of all tested CSRE series, depending on The MC III mixture contained the least amount of clay minerals. Besides beidellite and kaolinite, the clay fraction content and the content of individual clay minerals in the soil mixture. From the the most likely mineral to comprise the clay fraction was calcite. For this reason, CSRE made from graph showing the relationship between the CSRE compressive strength and kaolinite content, one this mixture had the highest compressive strength of those in the CSRE group containing a high clay could conclude that as the kaolinite content increased, the compressive strength decreased. However, content. Mixtures MC IV and MC V were characterized by their almost identical mineral composition; comparing the results of mixtures MC IV, MV V, and MC X showed the opposite result, that with hence, they obtained similar compressive strength results. These almost identical results confirmed more kaolinite, there as a slightly higher compressive strength. The high compressive strength of the that the mineral composition affected the strength of the samples. The lack of swelling minerals CSRE from the MC III mix (low kaolinite mix) was due to the high calcite content. The impact of (smectites, i.e., beidellite and montmorillonite) in the MC IV and MC V mixes resulted in a noticeably illite is difficult to determine; the graph revealed that the other minerals affected the result. The CSRE higher compressive strength compared to the MC X mix. The MC X mixture was the only mix series containing montmorillonite obtained the lowest compressive strength. The higher compressive containing montmorillonite, a highly swelling mineral with high hydrophilicity. For this reason, strength of the samples from the MC III mixture compared to samples from mixtures containing a CSRE samples from this mixture had the highest moisture content at the time of testing and higher (16%) clay fraction content (MC IV, MC V, and MC X) did not result from the beneficial effect of consequently the lowest compressive strength. The MC III mix contained 6.6% beidellite, which was calcite, but from a lower content of clay minerals in the clay fraction in the MC III mixture. also a swelling mineral. For this reason, this mixture was characterized by a higher moisture content, Calcite is not a clay mineral and does not swell at all. Even kaolinite swells more than calcite. compared to mixes MC IV and MC V. Hence, the highest compressive strength of the samples from the MC III mixtures resulted. Calcite,

however, undergoes weathering processes, which also has a destructive eﬀect on the CSRE compressive strength. This also explains why the LC II mixture obtained the worst compressive strength in the

group of mixtures with a low clay fraction content (4%, LC). Figure 11 shows the compressive strength results of all tested CSRE series, depending on the clay fraction content and the content of certain non-clay minerals in the soil mixture. For CSRE with a higher clay content (16%), a significant increase in compressive strength was observed along with an increase in the calcite content. It is worth mentioning in the group with the lower clay fraction content that the opposite was true. CSRE samples made from high clay content mixtures, which had a high percentage of calcite and were stabilized with 6% cement, produced comparable compressive strength results to samples from mixtures with low clay content that did not contain calcite. The results for series with mixtures containing a 9% addition of cement differed significantly. Samples from these two series obtained the largest discrepancies in the results. Materials 2020, 13, 324 13 of 21 Materials 2020, 13, x FOR PEER REVIEW 13 of 21

FigureFigure 10. 10.Compressive Compressive strength strength in in comparison comparison to theto the clay clay fraction fraction content content and and the amountthe amount of each of clayeach mineral:clay mineral: kaolinite kaolinite (top left), (top illite left), (top right),illite (top smectites right), (bottom smectites left), (bottom and montmorillonite left), and montmorillonite (bottom right). (bottom right). Goethite’s low influence was noticeable in mixtures with high clay content and was also beneficial. The effGoethite’sect of siderite low on influence compressive was strength noticeable was in unfavorable mixtures andwith visible high withclay highcontent clay contents.and was Thealso influencebeneficial. of The quartz effect was of favorable, siderite on which compressive resulted fromstrength the factwas thatunfavorable the clay content and visible was lower,with high and clay the compositecontents. The became influence similar of to quartz cement was concrete. favorable, The ewhichffect of resulted calcite was from debatable the fact and that depended the clay content on the formwas inlower, which and it occurred.the composite became similar to cement concrete. The effect of calcite was debatable and depended on the form in which it occurred.

Materials 2020, 13, 324 14 of 21 Materials 2020, 13, x FOR PEER REVIEW 14 of 21

FigureFigure 11. 11.CSRE CSRE compression compression strength strength in in comparison comparison to to the the clay clay fraction fraction content content and and on on the the amount amount ofof certain certain non-silicate non-silicate minerals: minerals: goethitegoethite (top(top left),left), sideritesiderite (top(top right),right), quartzquartz (bottom(bottom left), left), calcite calcite (bottom(bottom right). right).

3.2. SEM Image Analysis 3.2. SEM Image Analysis Comparing SEM images of CSRE samples from mixtures MC V 9% CEM (Figure 12A) and MC X Comparing SEM images of CSRE samples from mixtures MC V 9% CEM (Figure 12A) and 9% CEM (Figure 12B) showed that the MC X 9% CEM sample had more cracks. This may be due to the MC X 9% CEM (Figure 12B) showed that the MC X 9% CEM sample had more cracks. This may be presencedue to the of presence montmorillonite, of montmorillonite, which when which absorbing when water, absorbing caused water, swelling. caused swelling. ClustersClusters of of unreacted unreacted cement cement stood stood out out as as light light “grains” “grains” against against other other mineral mineral grains grains or or mixture mixture ofof clay clay minerals minerals constituting constituting the the dark dark grey grey background background of of the the sample. sample. Unreacted Unreacted cement, cement, in in addition addition toto the the characteristic characteristic brightness brightness threshold, threshold, also also had had a a relatively relatively characteristic characteristic texture, texture, it it occurred occurred as as a a ﬁllingfilling in in the the background background of of the the sample, sample, bonding bonding clusters clusters of of clay clay minerals. minerals. InIn contrast,contrast, ironiron oxides,oxides, barite,barite, or or other other rock rock forming forming minerals minerals occurred occurred as as single single grains grains embedded embedded in in clay clay (the (the details details of of SEM SEM imageimage analysis analysis basedbased onon the the brightness brightness threshold threshold supported supported byby pointpoint EDSEDS analyzes analyzes ofof the the speciﬁc specific grainsgrains with with distinctive distinctive shades shades of of gray gray are are included included in in Appendix AppendixA Aof of the the article). article). In both the MC III 9% CEM (Figure 12C) and MC V 9% CEM (Figure 12A) samples, unreacted cement (white areas) could be seen, but more so in the MC V 9% CEM sample. Higher amounts of unreacted cement reduced the CSRE’s compressive strength. Figures 13 and 14 show BSE images of samples with a high clay content (MC) mixture with the addition of 9% cement. The yellow indicates

Materials 2020, 13, x FOR PEER REVIEW 15 of 21

Materials 2020, 13, 324 15 of 21 visible calcium compounds in the form of limestone grains in MC III (Figure 13) and calcium compound rich clay minerals in MC X (Figure 14). A high occurrence of calcium compounds in clay minerals (including montmorillonite) was correlated with the occurrence of a large number of microcracks in the CSRE structure (Figure 14). This can be treated as a reason for the low compressive strength results obtained by the MC X series. Materials 2020, 13, x FOR PEER REVIEW 15 of 21

Figure 12. SEM images of samples MC V 9% CEM (A), MC X 9% CEM (B), and MC III 9% CEM (C). Symbols in the figures: WD—working distance; EHT—accelerating voltage; AsB—backscattered electrons (BSE) detector.

In both the MC III 9% CEM (Figure 12C) and MC V 9% CEM (Figure 12A) samples, unreacted cement (white areas) could be seen, but more so in the MC V 9% CEM sample. Higher amounts of unreacted cement reduced the CSRE’s compressive strength. Figures 13 and 14 show BSE images of samples with a high clay content (MC) mixture with the addition of 9% cement. The yellow indicates visible calcium compounds in the form of limestone grains in MC III (Figure 13) and calcium compound rich clay minerals in MC X (Figure 14). A high occurrence of calcium compounds in clay mineralsFigureFigure (including 12.12. SEMSEM images imagesmontmorillonite) of of samples samples MC MC was V V9% 9%correla CEM CEM (tedA (),A MC),with MC X 9% Xthe 9% CEM occurrence CEM (B), ( Band), and MCof MC aIII large 9% III 9%CEM number CEM (C). of microcracks(SymbolsC). Symbols in in the the in CSRE the figures: ﬁgures: structure WD—working WD—working (Figure 14).distance; distance; This canEHT—accelerating EHT—accelerating be treated as a reasonvoltage; for AsB—backscatteredAsB—backscattered the low compressive strengthelectronselectrons results (BSE)(BSE) obtained detector.detector. by the MC X series.

Figure 13.13. BSEBSE image image of of a sample a sample of MC of IIIMC 9% III CEM, 9% showingCEM, showing the occurrence the occurrence of calcium. of Thecalcium. numerous The carbonatenumerous grainscarbonate are markedgrains are in yellow.marked in yellow.

Figure 13. BSE image of a sample of MC III 9% CEM, showing the occurrence of calcium. The numerous carbonate grains are marked in yellow.

Materials 2020, 13, 324x FOR PEER REVIEW 16 of 21 Materials 2020, 13, x FOR PEER REVIEW 16 of 21

Figure 14. BSE image of a sample of MC X 9% CEM. Visible yellow calcium patches are not carbonates, Figurebut clay 14. minerals BSEBSE image image rich of ofin a aCa. sample sample of of MC MC X X 9% 9% CEM. CEM. Visi Visibleble yellow yellow calcium calcium patches patches are are not not carbonates, carbonates, but clay minerals rich in Ca. When analyzing the SEM images of samples from a low clay fraction mixture (Figure 15), numerousWhen cracks, analyzinganalyzing gaps, the the and SEM SEM pores images images we ofre samples clearlyof samples visible from from a lowas in claya thelowfraction image clay offraction mixture the LC (Figuremixture XI 6% 15CEM (Figure), numerous sample. 15), numerousThecracks, image gaps, cracks,of and the pores samplegaps, were and from pores clearly the we LC visiblere XI clearly 9% as inCEM visible the imagemixture as in of the theshowed image LC XI that 6%of the CEMthe LC granular sample. XI 6% CEMskeleton The imagesample. was of Thecompact,the sampleimage with of from the a significantly thesample LC XI from 9% reduced CEMthe LC mixture number XI 9% showed CEMof ga psmixture thatcompared the showed granular to the that sample skeleton the granular with was 6% compact, skeletoncement. withThiswas compact,resulteda significantly within a a reducedsignificant significantly number difference reduced of gaps numberin compared the obof tainedga tops thecompared compressive sample to with the strength 6%sample cement. with(over This 6% 5 cement. resultedMPa). ThisThe in a resultedcompressivesignificant in di affstrength erencesignificant in of the the difference obtained MC XI series compressive in the with ob atained strength6% cement compressive (over addition, 5 MPa). strength amounting The compressive (over to approximately5 MPa). strength The of compressive8the MPa, MC allowed XI series strength the with material aof 6% the cement toMC be XI used addition, series as a with stru amounting cturala 6% cementcomponent. to approximately addition, On the amounting 8other MPa, hand, allowed to theapproximately theSEM material image 8analysisto MPa, be used allowed suggested as a the structural materialthat samples component. to be usedwith as On6% a stru thecementctural other may component. hand, have the insufficient SEM On imagethe other durability, analysis hand, the suggested especially SEM image that in analysisclimatessamples withwheresuggested 6% CSRE cement that partitions maysamples have may with insu be ffi 6%exposedcient cement durability, to cyclicalmay especiallyhave freezing insufficient in and climates thawing, durability, where as CSREwell especially as partitions cyclical in climateswettingmay be exposed andwhere drying. CSRE to cyclical Evenpartitions freezingat the may optimum and be exposed thawing, moisture to as cyclical well content as freezing cyclical of the wettingand LC thawing, 6% and CEM drying. as (7%) well Evenmixture, as cyclical at the it wettingremainedoptimum and moisturepoorly drying. workable content Even at ofand, the as LCoptimum a 6%result CEM ofmoisture (7%)its compaction, mixture, content it ofproduced remained the LC poorly6%a highly CEM workable porous(7%) mixture, structure and, as it a remainedwithresult a oflow its poorly compressive compaction, workable strength. produced and, as a a highly result porous of its compaction, structure with produced a low compressive a highly porous strength. structure with a low compressive strength.

Figure 15. SEMSEM images of samples LC XI 6% CEM (left) and LC XI 9% CEM (right). Figure 15. SEM images of samples LC XI 6% CEM (left) and LC XI 9% CEM (right). 4. Conclusions Conclusions 4. Conclusions The research presented in this article showed that mineral composition has a signiﬁcantsignificant impact on the The compressive research presented strength in of thisCSRE. article For this showed reason reason, that, the mineral authors composition suggested hasthat a all significant research articles impact on the rammed compressive earth should strength include of CSRE. mine mineral Forral thiscomposition, reason, the along authors with suggested othe otherr key that soil all properties research such articles as ongranulation, rammed earthmoisture should content, include and mine stabilizers.ral composition, The authors’ along intention with othe was r keynot soilto determine properties the such effect eﬀect as granulation,of individual moisture minerals content, on the compressive and stabilizers. strength The authors’ of of CSRE intention samples, samples, was due due not to to tothe the determine fact fact that that in inthe nature, effect ofsuch individual monomineralmonomineral minerals soils soils on are arethe very compressivevery rare. rare. Another Another strength problem problem of CSRE is the interactionissamples, the interaction due between to the between individualfact that individual in minerals. nature, suchminerals.From monomineral the comparisonFrom the comparisonsoils of the are CSRE very of compressivethe rare. CSRE Another compre strength problemssive result strength is sets,the result oneinteraction can sets, draw one between general can draw qualitativeindividual general minerals.qualitativeconclusions: From conclusions: the comparison of the CSRE compressive strength result sets, one can draw general qualitative conclusions:

Materials 2020, 13, 324 17 of 21

Montmorillonite, as a hydrophilic mineral, set the moisture balance between the sample and the • environment at a higher level than for samples not containing this mineral. The result was a higher moisture content at the time of testing of CSRE samples containing montmorillonite, and consequently a lower strength. Beidellite slightly reduced the compressive strength. • Illite did not explicitly affect the compressive strength. • Kaolinite, as a non-swelling mineral, slightly increased the compressive strength. • Due to the low percentage of non-clay minerals with the exception of quartz, it was difficult to • clearly assess their impact on compressive strength. The higher compressive strength of the samples from the MC III mixture compared to samples • from mixtures containing a higher (16%) clay fraction content (MC IV, MC V, and MC X) did not result from the beneficial effect of calcite, but from a lower content of clay minerals in the clay fraction in the MC III mixture. Calcite is not a clay mineral and does not swell at all. Even kaolinite swells more than calcite. Hence, the highest compressive strength of the samples was from the MC III mixtures. Calcite, however, undergoes weathering processes, which also has a destructive effect on the CSRE compressive strength. This also explained why the LC II mixture obtained the worst compressive strength in the group of mixtures with a low clay fraction content (4%, LC).

The study also conﬁrmed that soil granulation has a primary eﬀect on CSRE compressive strength. For mixtures stabilized with cement, it is recommended to use soils with a low clay fraction.

Author Contributions: Conceptualization P.N.; methodology P.N.; formal analysis, P.N., J.K., I.G., and P.W.; investigation P.N., J.K., I.G., and E.W.; resources, P.N.; writing, original draft preparation P.N. and I.G.; writing, review and editing, P.N., I.G., P.W., and J.K.; visualization, P.N. and J.K.; supervision, P.N.; project administration, P.N. All authors have read and agreed to the published version of the manuscript. Funding: This research received no external funding. Acknowledgments: This research was supported by the Cryo-SEM laboratory, National Multidisciplinary Laboratory of Functional Nanomaterials NanoFun Project No. POIG.02.02.00-00-025/09 at the Faculty of Geology, University of Warsaw. This research received financial support from the Faculty of Civil Engineering, Warsaw University of Technology. Conflicts of Interest: The authors declare no conflict of interest.

Appendix A Figure A1 is a BSE photo of the sample along with energy-dispersive X-ray spectroscopy (EDS) analysis locations (an1–an6). Figure A2 shows full EDS spectra. Measurements were done with 20 kV voltage and point analysis with 30 s counting. When analyzing a sample in SEM using a BSE (backscattered electrons) detector, the image in various shades of gray, with so called, material contrast was received. This means that a material made of a substance with a higher electron density (in most cases, with a higher atomic number) will have a lighter shade than a material made of a substance with lighter elements. It is often the case that mineral grains with diﬀerent chemical compositions have a very similar shade of gray (Figure A2, an3—dolomite, an4—calcite, an6—quartz). In that case reliable method of identifying the given grains in SEM is EDS analysis. However, some grains may have such a characteristic shade of gray that they can be easily distinguished from other. Clusters of unreacted cement (Figure A2, an1) stand out as light “grains” against other mineral grains (Figure A2, an3, an4, an6) or mixture of clay minerals (Figure A2, an5) constituting the dark grey background of the sample. The most vivid white grains present in the sample belong to naturally occurring grains of iron oxides and/or barite (Figure A2, an2). Unreacted cement, in addition to the characteristic brightness threshold, also has a relatively characteristic texture, it occurs as a ﬁlling the background of the sample, bonding clusters of clay minerals. In contrast, iron oxides, barite or other rock-forming minerals occur as single grains embedded in clay. Materials 2020, 13, 324 18 of 21 Materials 2020, 13, x FOR PEER REVIEW 18 of 21 Materials 2020, 13, x FOR PEER REVIEW 18 of 21

FigureFigureFigure A1.A1. A1. BSE BSE photo photo photo of ofof the thethe sample sample along along with with with EDS EDS EDS analysis analysis analysis locations. locations. locations.

an1—cementan1—cement

an2—baritean2—barite

an3—dolomite an3—dolomite

Figure A2. Cont.

Materials 2020, 13, 324 19 of 21 Materials 2020, 13, x FOR PEER REVIEW 19 of 21

an4—calcite

an5—mix of clay minerals

an6—quartz

FigureFigure A2. A2.EDS EDS spectraspectra of grains grains with with different diﬀerent shades shades of ofgrey grey indicate indicate different diﬀerent chemical chemical composition. composition.

ReferencesReferences

1. 1. Hall,Hall, M.R.; M.R.; Swaney, Swaney, W. EuropeanW. European modern modern earth earth construction. construction. In Modern In Modern Earth Buildings.Earth Buildings. Materials, Materials, Engineering, ConstructionEngineering, and Construction Applications and; Applications Woodhead; PublishingWoodhead Limited:Publishing Cambridge, Limited: Cambridge, UK, 2012; UK, pp. 2012; 650–687. pp. 650– 2. Piasecki,687. M.; Kozicki, M.; Firl ˛ag,S.; Goljan, A.; Kostyrko, K. The Approach of Including TVOCs Concentration

2. inPiasecki, the Indoor M.; EnvironmentalKozicki, M.; Firl Qualityąg, S.; ModelGoljan, (IEQ)-CaseA.; Kostyrko, Studies K. The of BREEAMApproach Certifiedof Including Offi ceTVOCs Buildings. Concentration in the Indoor Environmental Quality Model (IEQ)-Case Studies of BREEAM Certified Office Sustainability 2018, 10, 3902. [CrossRef] Buildings. Sustainability 2018, 10, 3902, doi:10.3390/su10113902. 3. Anysz, H.; Narloch, P. Designing the composition of cement stabilized rammed earth using artificial neural 3. Anysz, H.; Narloch, P. Designing the composition of cement stabilized rammed earth using artificial neural networks. Materials 2019, 12, 1396. [CrossRef][PubMed] networks. Materials 2019, 12, 1396, doi:10.3390/ma12091396; DOI:10.3390/ma12091396. 4. Minke, G. Building with Earth: Design and Technology of a Sustainable Architecture; Birkhäuser–Publishers for 4. Minke, G. Building with Earth: Design and Technology of a Sustainable Architecture; Birkhäuser–Publishers for Architecture:Architecture: Kassel, Kassel, Germany,Germany, 2006. 5. 5. Narloch,Narloch, P.L.; P.L.; Woyciechowski, Woyciechowski, P.; P.; J˛eda,P. Jęda, The P. influenceThe influence of loam of loam type andtype cement and cement content content on the compressiveon the strengthcompressive of rammed strength earth. of rammedArch. Civ.earth. Eng. Arch.2015 Civ., 61Eng., 73–88. 2015, 61 [CrossRef, 73–88, doi:10.1515/ace-2015-0005.] 6. 6. Ciancio,Ciancio, D.; D.; Jaquin, Jaquin, P.; Walker, P. Advances on on the the assessment assessment of ofsoil soil suitability suitability for rammed for rammed earth. earth. Constr.Constr. Build.Build. Mater. Mater.2013 2013,, 4242,, 40–47.40–47, doi:10.1016/j.conbuildmat.2012.12.049. [CrossRef]

Materials 2020, 13, 324 20 of 21

7. Consoli, N.C.; Festugato, L.; Gravina da Rocha, C.; Caberlon Cruz, R. Key parameters for strength control of rammed sand–cement mixtures: Influence of types of portland cement. Constr. Build. Mater. 2013, 49, 591–597. [CrossRef] 8. Lina, H.; Zhenga, S.; Lourençoa, S.; Jaquin, P. Characterization of coarse soils derived from igneous rocks for rammed earth. Eng. Geol. 2017, 228, 137–145. [CrossRef] 9. Bui, Q.-B.; Morel, J.-C. First Exploratory Study on the Ageing of Rammed Earth Material. Materials 2015, 8, 1–15. [CrossRef] 10. Bui, Q.B.; Morel, J.C.; Venkatarama Reddy, B.V.; Ghayad, W. Durability of rammed earth walls exposed for 20 years to natural weathering. Build. Environ. 2009, 44, 912–919. [CrossRef] 11. Venkatarama Reddy, B.V.; PKumar, P. Cement stabilised rammed earth. Part A: Compaction characteristics and physical properties of compacted cement stabilised soils. Mater. Struct. 2011, 44, 681–693. [CrossRef] 12. Bui, Q.-B. Assessing the Rebound Hammer Test for Rammed Earth Material. Sustainability 2017, 9, 1904. [CrossRef] 13. Bui, Q.-B.; Morel, J.-C.; Hans, S.; Walker, P. Effect of moisture content on the mechanical characteristics of rammed earth. Constr. Build. Mater. 2014, 54, 163–169. [CrossRef] 14. Christopher, B.; Ciancio, D. Effect of compaction water content on the strength of cement-stabilized rammed earth materials. Can. Geotech. J. 2014, 51, 583–590. [CrossRef] 15. Hall, M.; Allinson, D. Influence of cementitious binder content on moisture transport in stabilised earth materials analysed using 1-dimensional sharp wet front theory. Build. Environ. 2009, 44, 688–693. [CrossRef] 16. Hall, M.; Allinson, D. Assessing the moisture-content-dependent parameters of stabilised earth materials using the cyclic-response admittance method. Energy Build. 2008, 40, 2044–2051. [CrossRef] 17. Pagacz, J.; Pielichowski, K. Preparation and characterization of PVC/montmorillonite nanocomposites–A review. J. Vinyl Addit. Technol. 2009, 15, 61–76. [CrossRef] 18. Arrigoni, A.; Grillet, A.-C.; Pelosato, R.; Dotelli, G.; Beckett, C.; Woloszyn, M.; Ciancio, D. Reduction of rammed earth’s hygroscopic performance under stabilisation: An experimental investigation. Build. Environ. 2017, 115, 358–367. [CrossRef] 19. Grabowska-Olszewska, B. Geologia Stosowana; Wła´sciwo´sciGruntów Nienasycoych; Wydawnictwo Naukowe PWN: Warszawa, Poland, 1998. 20. Punmia, B.C.; Jain, A.K.; Jain, A.K. Soil Mechanics and Foundations; Laxmi Publications: New Delhi, India, 2005. 21. Reddy, B.V. Stabilised soil blocks for structural masonry in earth construction. In Modern Earth Buildings; Materials, Engineering, Construction and Applications; Woodhead Publishing Limited: Cambridge, UK, 2012; pp. 324–363. 22. Reddi, L.N.; Jain, A.K.; Yun, H.-B. Soil materials for earth construction properties, classification and suitability testing. In Modern Earth Buildings; Materials, Engineering, Construction and Applications; Woodhead Publishing Limited: Cambridge, UK, 2012; pp. 155–171. 23. Houben, H.; Guillaud, H. Earth Construction: A Comprehensive Guide; Intermediate Technology Publications: London, UK, 1994. 24. Yanguatin, H.; JTobón, J.; JRamírez, J. Pozzolanic reactivity of kaolin clays, a review. Revista Ingenieria de Construccion 2017, 32, 13–24. [CrossRef] 25. Grim, R.E. Clay Mineralogy, International Series in the Earth and Planetary Sciences; Press, F., Ed.; McGraw-Hill Book Company: New York, NY, USA, 1968. 26. Karpinski, B.; Szkodo, M. Clay minerals–mineralogy and phenomenon of clay swelling in oil & gas industry. Adv. Mater. Sci. Eng. 2015, 15, 37–55. [CrossRef] 27. Kaczyński,R.; Grabowska-Olszewska, B. Soil mechanics of the potentially expansive clays in Poland. Appl. Clay Sci. 1997, 11, 337–355. [CrossRef] 28. Seed, H.B.; Woodward, R.J.; Lundgren, R. Prediction of Swelling Potential for Compacted Clays. Trans. Am. Soc. Civ. Eng. 1962, 88, 53–88. 29. Oliver, M.; Mesbah, A. The earth as a material. In Proceedings of the International Symposium on Modern Earth Construction, Peking, China, 1–4 November 1985. 30. Borkowska, M.; Smulikowski, K. Mineraly Skałotworcze; Wydawnictwa Geologiczne: Warszawa, Poland, 1973; p. 165. Materials 2020, 13, 324 21 of 21

31. Narloch, P.; Hassanat, A.; Tarawneh, A.S.; Anysz, H.; Kotowski, J.; Almohammadi, K. Predicting Compressive Strength of Cement-Stabilized Rammed Earth Based on SEM Images Using Computer Vision and Deep Learning. Appl. Sci. 2019, 9, 5131. [CrossRef] 32. Narloch, P.L. Ziemia Ubijana Stabilizowana Cementem Jako Konstrukcyjny Materiał Budowlany w Klimacie Umiarkowanym. Ph.D. Thesis, Warsaw University of Technology, Warsaw, Poland, 2017. 33. Narloch, P.L.; Lidner, M.; Kunicka, M.; Bielecki, M. Flexural tensile strength of construction elements made out of cement stabilized rammed earth. Procedia Eng. 2015, 111, 589–595. [CrossRef] 34. Narloch, P.L.; Woyciechowski, P.P.; Dmowska, E.; Halemba, K. Durability assessment of monolithic rammed earth walls. Arch. Civ. Eng. 2015, 61, 73–88. [CrossRef] 35. Kariyawasam, K.K.G.K.D.; Jayasinghe, C. Cement stabilized rammed earth as a sustainable construction material. Constr. Build. Mater. 2016, 105, 519–527. [CrossRef] 36. Rempel, A.W.; Rempel, A.R. Frost Resilience of Stabilized Earth Building Materials. Geosciences 2019, 9, 328. [CrossRef] 37. Firl ˛ag,S.; Piasecki, M. NZEB Renovation Deﬁnition in a Heating Dominated Climate: Case Study of Poland. Appl. Sci. 2018, 8, 1605. [CrossRef] 38. Rahmat, M.R.; Ismail, N. Eﬀect of optimum compaction moisture content formulations on the strength and durability of sustainable stabilised materials. Appl. Clay Sci. 2018, 157, 257–266. [CrossRef] 39. Hall, M.; Djerbib, Y. Rammed earth sample production: Context, recommendations and consistency. Constr. Build. Mater. 2004, 18, 281–286. [CrossRef] 40. NZS 4298. Materials and Workmanship for Earth Buildings; Standards New Zealand: Wellington, New Zealand, 1998. 41. PN-B-04481:1988. Grunty Budowlane—Badania Próbek Gruntu; Polski Komitet Normalizacyjny: Warszawa, Poland, 1988. 42. BS EN 12390-3. Testing Hardened Concrete; Compressive Strength of Test Specimens; British Standards Institution: London, UK, 2019.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).