Hydraulic lime is manufactured by using a low-grade limestone having high SiO₂ and Al₂O₃. Since properties of hydraulic lime can be expressed according to contents of mineral phase which is produced during calcination, the mixing process of other materials for improving physical properties is not included in the hydraulic lime manufacturing process.¹⁾ Therefore, manufacturing process of hydraulic lime includes somewhat simple processes like calcination, hydration, and fine grinding. Also, if physical properties are needed to be improved, an adequate admixture is mixed in the finished sample. Fig. 1 shows manufacturing process of D-NHL using a low-grade dolomitic limestone. D-NHL was manufactured by the same process as general hydraulic lime.

Low-grade dolomitic limestone was collected from the mines of H company located at Danyang, Chungcheongbukdo, Korea. The collected limestone was crushed to a size (10 ~ 20) mm suiting to calcination and hydration process. Calcination was executed at 1,250°C for two hours, hydraulic lime was quickly cooled, and dry type hydration was continued at water ratio 0.4% (by weight) for 24 h. The samples, after completing hydration, was dried at (105 ± 5)°C for 24 h, and fine grinding was executed using an air jet mill to prepare D-NHL with particle size of hydraulic lime less than 0.09 mm, complying BS EN 459-1: 2010. Paste and mortar were prepared to evaluate the physical properties of prepared D-NHL itself according to curing, and changes in the physical properties were reviewed by mixing blast furnace slag (hereinafter referred to as BFS) to further raise applicability in the field.

Paste of D-NHL single component and another paste made of ternary system D-NHL + BFS + gypsum were prepared to study hydration properties of D-NHL. In case of general hydraulic lime, hydration and carbonation are complicatedly occurred, thereby shows a long-term curing mechanism wherein physical properties are continuously improved even until one year. D-NHL prepared in this study also, though it was expected that complicate reaction of hydration and carbonation would occur as a lime binder to which hydraulic properties is added, since D-NHL has somewhat less contents of hydraulic mineral phase, it was intended to supplement physical properties in an experimental way by mixing mineral additives BFS and gypsum at a specific ratio.

Table 1 and Fig. 2 show chemical analysis results and mineral phase analysis results for BFS and gypsum. Chemical analysis of BSF revealed that it was composed of CaO at 42.44%, SiO₂ at 27.40%, and Al₂O₃ at 14.79%, indicating that SiO₂ and Al₂O₃ were contained in large quantities. Mineral phase analysis showed that BFS contained as amorphous mineral, quartz (SiO₂) and dihydrate (CaSO₄·2H₂O). Chemical analysis of gypsum showed that CaO at 42.20% and SO₃ at 51.00% were main components, and minute quantities of SiO₂ and Al₂O₃ were also contained, while mineral, phase analysis showed that gypsum was presented as anhydrous form. Gypsum has been used as an alkali stimulant which promotes potential hydraulic properties of BSF, and if ion elution of BSF is smooth, it would affect greatly on the overall types of hydrates and production rate. Particle size analysis as can be seen in Fig. 3 showed that overall particle size of BFS and gypsum was more than 10 μm, which is similar as large size particles normally presented in the hydraulic lime. Prepared paste was divided in the petri-dish as thin layer, hydration was stopped with acetone according to curing, and then dried before sampling, ready for analysis with instrument. Reaction properties of each sample which received pre-treatment according to each curing was investigated with X-ray diffraction analysis (XRD: D/max 2500V/P, Rigaku Co. Ltd. Japan), differential scanning calorimetry, thermogravimetric analyzer (TG/DSC: STA 449C Jupiter, NETZSCH Co, Ltd. Germany), scanning electron microscope (SEM: S-4300, HITACHI Co. Ltd. Japan), and porosity measurement (Auto Pore IV 9520, Micromeritics Co. Ltd. USA).

Generally, raw materials which are utilized to manufacture hydraulic lime are low-grade limestones having high SiO₂ and Al₂O₃ contents. During calcination process, CaO is produced by thermal degradation of CaCO₃, and CaO reacted with SiO₂ or Al₂O₃ through solid state reaction to produce hydraulic minerals such as C₂S, C₃S, and C₃A. According to contents of mineral phase thus produced, hydration properties of hydraulic binder differ. Those researchers, who proposed that dolomitic lime can show hydraulic properties, forecasted that semi hydraulic lime would match hydraulic lime having somewhat lower hydraulic lime.⁶⁾ Table 3 shows chemical analysis result of low-grade dolomitic limestone used as a raw material for D-NHL production, showing SiO₂ at 5.18% and MgO at 14.44%, which indicates that the dolomitic limestone is very low-grade as compared with general industrial purpose limestone and contains large amount of SiO₂ and MgO. Fig. 4 shows XRD analysis results of raw materials used for manufacture of D-NHL in this study. Fig. 4(a) shows low-grade dolomitic limestone, (b) shows sample after calcination at an optimum calcination temperature in which solid phase reaction can take place, and (c) shows analysis results of D-NHL ground to comply EU standard after dry type hydration of calcined sample. Major mineral phases of the low-grade dolomitic limestone in Fig. 4(a) are CaCO₃, CaMg(CO₃)₂, SiO₂, and muscovite, having higher SiO₂ contents than general high grade limestone. Besides, they contain high amount of CaCO₃ and CaMg(CO₃)₂, thus expected to produce hydraulic mineral phase smoothly by solid phase reaction after thermal decomposition. After calcination, major mineral phase of low-grade dolomitic lime (Fig. 4(b)) were CaO, MgO, C₂S, and C₃S, revealing that MgO was produced by thermal degradation of CaMg(CO₃)₂, while C₂S and C₃S were produced by solid phase reaction. In case of finally produced D-NHL (Fig. 4(c)), Ca(OH)₂ and Mg(OH)₂, which attribute carbonation and hydrate production, were presented in a large quantity. Hydraulic mineral phases such as C₂S and C₃S, were also remained, indicating that complicated curing would be proceeded. Particle analysis of D-NHL in Fig. 5 shows that average particle size is lower than 90 μm, complying EU standard. These compositional mineral phases of D-NHL are same as the manufacturing properties of general hydraulic lime, which implies that D-NHL manufacture is possible by utilizing domestically available raw limestone.

Curing mechanism as a hydraulic lime was checked by investigating hydration properties of D-NHL and changes in mineral phase by hydration and carbonation were investigated. D-NHL paste was prepared and changes in mineral phase according to curing was reviewed. XRD analysis revealed that major mineral phases of D-NHL were Ca(OH)₂, CaCO₃, Mg(OH)₂, MgO, and C₂S as shown in Fig. 6. Part of Ca(OH)₂ was converted to CaCO₃ by carbonation, and CaCO₃ peak was increased as curing was increased. It indicates that carbonation was continuously progressed, and remaining Ca(OH)₂ also would attribute carbonation when it enters into a long curing after 28 days. However, in case of Mg(OH)₂, there was no changes in the peak according to curing. Further, carbonates and hydrates including MgCO₃ which is a carbonate of Mg(OH)₂ were not presented. According to preceding literatures, carbonation of Mg(OH)₂ is difficult to be proceeded under atmosphere, and physical properties should be observed for at least some years to check carbonates.^7,8) Mg(OH)₂ included in D-NHL, prepared in this study, also might exhibited the same trend in the preceding studies. It is judged that additional research has to be performed, which can prove physical properties change and soundness of Mg base hydrate in terms of a long-term reactivity, in order to propose a potential for the production of lime binder as a high value added product of low-grade dolomitic limestone. For more accurate approach about mineral phases that could be confirmed by the XRD analysis, DSC analysis was carried out (Fig. 7).

DSC analysis results showed that endothermic peaks by thermal decomposition of Mg(OH)₂, Ca(OH)₂, and CaCO₃ were clearly formed at near 400°C, 500°C and 800°C. Further, some of the endothermic peaks at near 100°C, which was corresponding to thermal decomposition temperature of C-S-H phase produced by hydration of C₂S, were appeared. However, peaks by hydrate and carbonate of Mg(OH)₂ could not be found. This result shows a similar trend with the XRD analysis result.

For the physical properties of hydraulic lime complying EU standard, mandatory measurement items and properties values corresponding to each item are specified. Compressive strength is one of the basic properties for evaluations that can forecast practical implementation of hydraulic lime. Fig. 8 shows measurement results of compressive strength for the D-NHL mortar prepared as per EU standard, which shows increases in the compressive strength as curing was increased. This enhancement of strength might be because inner compactness was raised by the reaction product which was produced by carbonation and hydration. While, compressive strength of hydraulic lime as per EU standard is specified as at least more than 2 MPa on the 28 days. However, in case of D-NHL prepared in this study, it showed a low compressive strength as 1 MPa on 28 days, which does not comply EU standard. Same as the result from XRD analysis, content of Ca(OH)₂ which could not be converted to CaCO₃ by carbonation was high, and durability enhancement of hardened body might be very slowly appeared due to C₂S and Mg(OH)₂, which have a low reactivity. Such expression of strength is due to mineral phase of D-NHL itself, thus research about how mineral phase of D-NHL behaviors in long-term physical properties might be required.

Table 4 shows measurement results of physical properties for D-NHL performed as per physical properties measurement items of hydraulic lime that is specified in BS EN 459-1: 2010. The specification of physical properties of hydraulic lime according to EU standard is air contents within 5%, soundness within ± 2 mm, and in case of setting, initial setting for more than one hour, and final setting time within 40 h. Physical properties values of D-NHL as in Table 4 showed air contents 1.10%, soundness −1.00 mm, and properties values were in conformity with EU standard. However, the initial setting time 72 h for D-NHL in this study was not in conformity with physical properties of hydraulic lime specified in EU standard, while final setting time was more than 168 h, which could not be confirmed as an accurate properties’ value. This might be because non-availasility of C-S-H, and ettringite, or Al series compound such as C₄AH₁₃ that attribute initial setting.^9,10) Expression of overall physical properties of D-NHL showed the similar trend with hydraulic lime, and physical properties also were in conformity with EU standard except compressive strength and setting.

Hydraulic lime according to EU standard is categorized into NHL, HL, and FL, and its criteria is as per the types and characteristics of additives. In case of D-NHL, it has been confirmed that supplement of properties such as compressive strength and setting is required through physical properties study. Mineral admixtures BFS and gypsum were added to supplement physical properties of hydraulic lime and the results were observed, considered practical application and physical properties changes. Used mineral admixtures were those which are widely used to enhance physical properties of cement and lime binder, and these can be distinguished as HL (NHL + inorganic additives) among hydraulic limes as per EU standard. Fig. 9 shows XRD analysis results according to curing of paste wherein BFS and gypsum were mixed into D-NHL. Physical properties change was investigated by adding BFS at 10%, 20%, and 30%, respectively. Major mineral phases of the paste were Ca(OH)₂, CaCO₃, Mg(OH)₂, MgO, C₂S, ettringite, gehlenite, C₄AH₁₃, and non-reacted gypsum. Overall mineral phases were similar regardless of admixture contents. As curing became extended, CaCO₃ contents were increased, which indicated that carbonation was continuously progressed. Further, with only XRD analysis, production of calcium base hydrates by C₂S could not be confirmed. Besides, carbonates by Mg(OH)₂ were not produced, which shows a similar trend with characteristics of mineral phase of D-NHL as shown in Fig. 6. Still, according to addition of BFS and gypsum, ettringite at its initial setting time (as on 3 days) was produced.

DSC analysis results (Fig. 10) showed that endothermic peak by Al compound such as C₄AH₁₃ and etteringite appeared at near 120°C, peaks at near 400°C by Mg(OH)₂, at near 500°C by Ca(OH)₂, and at near 800°C by CaCO₃, showing the similar trends as XRD analysis results. The largest difference between DSC thermal analysis result of D-NHL in this study as in Fig. 7 and D-NHL with added BFS and gypsum thermal analysis results in Fig. 10 was that the peak was formed at near 120°C by Al base hydrates, which was increased as curing become extended, thus confirming hydrates production by adding mineral admixture.

Such Al base hydrates directly affect initial strength and setting of hardened body, thus supplementation of physical characteristics on the D-NHL by adding BFS and gypsum would be possible. However, the production of hydrates by increasing BFS contents was not large enough, which might be due to limited SO₃²⁻ ion contents supplied by gypsum as against BFS contents.¹¹⁾

Figure 11 shows compressive strength measurement results for D-NHL mortar with added BFS and gypsum, indicating increased in the compressive strength as BFS content was increased and curing was extended. Physical properties more than HL 2 could be appeared with compressive strength 3.08 MPa on 28 days even by adding 10% of BFS, if analyzed with added BFS contents without considering gypsum contents which were at a constant level. When BFS was added at 20%, physical properties more than equivalent of HL 5, which exhibited the best physical properties among hydraulic lime, could be obtained with more than 2 MPa on 7 days and more than 5 MPa on 28 days, respectively. Increases in the compressive strength by adding BFS and gypsum, as shown in the XRD analysis results in Fig. 9 and DSC analysis as in Fig. 10, might be due to Al base compound which was produced from the initial stage of curing. While, expression of far excellent physical properties could be expected in the long-term when carbonation of Ca(OH)₂ and hydration of hydraulic mineral phase was progressed after 28 days.

Table 5 shows physical properties evaluation results for ternary system D-NHL wherein BFS and gypsum were mixed. Air contents measurement results showed that as BFS contents were increased, air contents were also increased. Besides, higher values in overall than physical properties of D-NHL itself could be observed. Air contents were decreased as powder became increased. This might be due to increases in the air contents as admixture contents were increased since particle sizes of BFS and gypsum were larger than D-NHL particles (Fig. 3 and Fig. 5). As far as soundness is concerned, changes in the length was within −1.00 mm when additives were added, which was in conformity with EU standard. Meanwhile, initial setting time within (13 ~ 18) h and final setting time within 26 h were matching with the soundness levels HL 3.5 ~ HL 5. Difference in the setting time before and after adding admixture revealed that physical properties were improved by longer than 54 h for initial setting, while physical properties were improved by longer than 142 h for final setting time, which might be due to enhancement of hydrate production as mineral additives were added. Fig. 12 shows SEM analysis result for ternary system D-NHL paste wherein D-NHL paste, BFS, and gypsum were mixed. In the D-NHL itself, mineral phase by hydrates could not be found, but in the D-NHL into which BFS and gypsum were added, minerals of ettringite and C₄AH₁₃ could be observed. Further, difference in the inner mineral phase by mineral additives could be confirmed same as in the XRD and DSC analysis results.