Resistance to Moisture-Induced Damage of Half-Warm-Mix Asphalt Concrete with Foamed Bitumen

Hot-mix asphalt (HMA) remains the predominant material for pavement surfacing. Mixing is performed at about 180 °C, depending on the bitumen used. Environmental concerns in terms of emissions and energy demand are fostering new sustainable technologies in road construction. Warm-mix asphalt (WMA) and half-warm-mix asphalt (HWMA) mixtures meet current expectations in that they are produced at lower temperatures, 100–130 °C, ensured by foaming the bitumen with water. The extent of temperature reduction requires that the mixture has adequate moisture and frost resistance, which is particularly important in countries that have a low-temperature climate. Asphalt concrete AC 8 S with 50/70-grade foamed bitumen modified with 0.6 wt.% surface-active agent (SAA) was used in the tests. To provide the AC mixture with the required resistance to climatic factors (water, temperature below 0), hydrated lime was added at 0, 15, 30, and 45 wt.% as limestone filler replacement. The influence of the hydrated lime addition on the air void content and resistance to moisture and frost damage was investigated according to the WT-2 2014 methodology based on EN 12697-12: 2008 and to the modified AASHTO T283 method. The optimum content of hydrated lime for filler replacement was determined through statistical analysis of the test results. With the optimum hydrated lime replacement of 30%, the required level of moisture and frost resistance of HWMA concrete with foamed bitumen is achieved. The results of this study confirmed the suitability of HWMA concrete with foamed bitumen for application in road construction practice.


Introduction
The sustainability of road construction processes is understood as minimizing negative environmental impacts through reducing greenhouse emissions and energy use of asphalt mixing plants. Conventional hot-mix asphalt (HMA) mixtures are produced at high temperatures reaching 180 • C, depending on the bitumen type used. During production, a large amount of CO 2 and other harmful gases are emitted, and a large amount of energy is needed to dry the aggregate and bring the bitumen from solid to liquid with a viscosity of about 0.2 Pa·s. The need to find more eco-friendly technologies has been reflected in the number of research and development activities leading to the implementation of low-temperature processes. Warm-mix asphalt (WMA) is produced and placed at lower temperatures compared with HMA [1,2]. This temperature reduction results from adding a variety of chemical agents [3][4][5][6][7], such as low-viscosity additives that decrease bitumen viscosity [8][9][10][11][12] and foamed bitumen based on natural [13][14][15] or synthetic zeolite [16,17].
Parameters V a , ITSR and RW WM were determined by compacting the specimens using a Marshall hammer with the number of blows depending on the procedure used. All characteristics were determined on a total of nine specimens [59] that met the required physical and geometrical criteria [58].

Air Void Content (V a )
The air void content is the volume of air voids in the bituminous specimen, expressed as a percent of the total volume of that specimen. This parameter was determined as per EN 12697-8:2005 using the formula below where: Va>V a -volume of air voids in the bituminous specimen (vol %), ρ m -density of the bituminous mixture (kg/m 3 ), ρ b -bulk density of the bituminous mixture (kg/m 3 ).

Water Sensitivity (ITSR) in Accordance with WT-2 [58]
The test was performed on Marshall specimens 63.5 mm ± 2.5 mm high and 101 mm in diameter. The compaction procedure consisted of applying 35 blows to each face of the specimen in the Marshall apparatus.
The prepared specimens were divided into two equal sets before conditioning. One set was stored at room temperature (20 ± 5 • C) without additional conditioning (so-called "dry set"). The other set (the so-called "wet set") was conditioned according to the WT-2 2014 procedure.
Conditioning of the specimens from the "wet set" began with placing them on a perforated shelf in a vacuum tank (chamber) filled with distilled water at a temperature of (20 ± 5 • C). After immersion, their upper surfaces were at least 20 mm below the water level. Then the vacuum apparatus was activated and absolute pressure (6.7 ± 0.3) kPa was obtained within 10 ± 1 min. To avoid damage to the specimens, the pressure was lowered slowly and evenly, and then maintained for a period of 30 ± 5 min. After that, the pressure was slowly increased up to the level of atmospheric pressure and then the specimens were left in water for another 30 ± 5 min. After removing from water, the dimensions of the specimens were measured in accordance with EN 12697- 29: 2006 and their volumes were calculated. Specimens that increased their volume by more than 2% were discarded and the degree of water saturation was determined according to the formula: where: N w -degree of water saturation of the specimen [%], A-air-dry specimen mass before vacuum saturation [g], B-specimen mass in air after vacuum saturation, [g] after surface drying, C-specimen mass in water after vacuum saturation [g], V-air void content in the sample, expressed as a decimal number.
The specimens with a saturation level above 80% were discarded. If the saturation level was below 55%, the saturation procedure was repeated. Specimens from the "wet set" were placed in a 40 ± 1 • C water bath for a period of 68 h. After removal from the water bath, the dripping wet specimens were tightly wrapped in a stretch film and placed in a plastic bag containing (10 ± 1) mL of water, then sealed and frozen at −18 ± 3 • C for a minimum of 16 h, counting from the time the freezer reached the required temperature. From the freezer the specimens were moved to a 25 ± 2 • C water bath. Shortly after submerging the specimens in water, they were removed from the plastic bags, unwrapped and quickly placed back in the water for 24 ± 1 h from the time of first insertion into the water after storage in the freezer. After conditioning, the indirect tensile strength of all the specimens was determined according to EN 12697-23.
The moisture and frost resistance indicator, ITSR, was calculated using the formula: where: ITS d -average indirect tensile strength of air-conditioned specimens, ITS w -average indirect tensile strength of water-conditioned specimens as per WT-2 2014.

Resistance to Moisture-Induced Damage in Accordance with Modified AASHTO T283 Method
The tests were conducted in accordance with the modified AASHTO T283 method [47,48,50,51] on Marshall-compacted specimens 63.5 mm ± 2.5 mm high and 101 mm in diameter. The compaction procedure was modified to obtain 6% to 8% voids. The prepared two sets of specimens differed by conditioning method. The dry set was air-conditioned at room temperature until testing.
The wet set specimens were placed in a vacuum chamber filled with distilled water. The upper surface of the specimens was covered by at least 20 mm of water. The pressure in the vacuum chamber was reduced to reach 6.7 kPa for 30 min. Then the specimens were put to a 40 • C ± 1 • C water bath for 68 h. After that, the specimens were wrapped in a plastic foil and frozen at −18 • C for at least 4 h, then thawed for 4 h at 20 • C. The number of freeze-thaw cycles was 18 as recommended by the adopted procedure [50,51]. After removal from the freezer, the specimens were submerged in water at 60 • C ± 1 • C for 24 h.
Directly before conditioning, all the specimens were stored in a 25 • C water bath for 3 h. The indirect tensile strength test at 25 • C was run in the universal testing machine at a rate of 50 mm/min. The load was applied to failure by caps 12 mm in width and with 50.5 mm curvature radius. The loading scheme was identical to that used during the indirect tensile strength test. The ratio for moisture-induced damage RW WM according to the modified AASHTO T283 procedure was calculated from where: ITS d A -average indirect tensile strength of air-conditioned specimens, ITS w A -average indirect tensile strength of water-conditioned specimens undergoing a freeze-thaw cycle as per AASHTO T283,

Materials and Mix Design Procedure
In the laboratory tests, 5.6%, 5.9%, 6.2% and 6.5% foamed bitumen was used in the HWMA concrete mixture, as required for the wearing course. [58] The bitumen was a 50/70 paving-grade bitumen, commonly used in the countries of central and eastern Europe in bituminous mixes for road pavements under traffic characterized by 2.5 × 10 6 < ESAL 100 kN < 7.3 × 10 6 (ESAL-equivalent single axle load) [58].
For the use in HWMA, the 50/70 bitumen was modified with fatty acid amide-based SAA in the amount of 0.6 wt.% (in relation to the mass) of the binder. The SAA properties are summarized in Table 1. The bitumen containing 0.6% SAA was subjected to water-based foaming. The foam characteristics, expansion ratio ER [19,29] and half-life HL [19,29], were determined with 9 replicates in the second stage of the study [59].
The test results of unmodified and 0.6% SAA modified bitumen are compiled in Table 2.  Figure 1 shows the characteristics of 50/70-grade foamed bitumen after 0.6% SAA modification.  Figure 1 shows the characteristics of 50/70-grade foamed bitumen after 0.6% SAA modification. The 50/70-grade bitumen containing 0.6% SAA has very high foaming parameters. Hence, its use in the mineral asphalt concrete mixture should ensure that all aggregate particles are well coated. Detailed results of the study of SAA modified bitumen 50/70 were described in [32].
The basic frame-compositions of the mineral mixture and bituminous mixture are summarized in Table 3 and the particle size design of AC 8 S is plotted in Figure 2. The research used a variable bitumen contents to determine the effect of the bitumen amount on asphalt concrete properties in terms of hydrated lime dosing. While increasing the amount of foamed bitumen from 5.9% to 6.5% according to the experimental design, the grading of the mineral mixture was adjusted accordingly. The 50/70-grade bitumen containing 0.6% SAA has very high foaming parameters. Hence, its use in the mineral asphalt concrete mixture should ensure that all aggregate particles are well coated. Detailed results of the study of SAA modified bitumen 50/70 were described in [32].
The basic frame-compositions of the mineral mixture and bituminous mixture are summarized in Table 3 and the particle size design of AC 8 S is plotted in Figure 2.  The 50/70-grade bitumen containing 0.6% SAA has very high foaming parameters. Hence, its use in the mineral asphalt concrete mixture should ensure that all aggregate particles are well coated. Detailed results of the study of SAA modified bitumen 50/70 were described in [32].
The basic frame-compositions of the mineral mixture and bituminous mixture are summarized in Table 3 and the particle size design of AC 8 S is plotted in Figure 2. The research used a variable bitumen contents to determine the effect of the bitumen amount on asphalt concrete properties in terms of hydrated lime dosing. While increasing the amount of foamed bitumen from 5.9% to 6.5% according to the experimental design, the grading of the mineral mixture was adjusted accordingly.   The research used a variable bitumen contents to determine the effect of the bitumen amount on asphalt concrete properties in terms of hydrated lime dosing. While increasing the amount of foamed bitumen from 5.9% to 6.5% according to the experimental design, the grading of the mineral mixture was adjusted accordingly.
To ensure moisture resistance of the asphalt concrete, hydrated lime was added at 15%, 30% and 45% by weight as a lime filler replacement. The AC 8 S mixture was prepared in a heated 60 l mechanical mixer to which foamed bitumen produced in the WLB-10S plant was added. The composite was mechanically mixed with a stirrer for a maximum of 5 min. The production temperature of AC 8 S with additives did not exceed 100 • C.
The mixture prepared in this way was compacted using the Marshall impact compactor. The number of impacts was dependent on the test type. After making, the asphalt concrete specimens were allowed to cool at room temperature for 48 h.
The algorithm for the design of factorial experiments was used [60]. The physical and mechanical parameters of AC 8 S were determined based on the adopted 4 × 4 factorial design, in accordance with the testing program. The developed plan of the experiment is shown in Figure 3. To ensure moisture resistance of the asphalt concrete, hydrated lime was added at 15%, 30% and 45% by weight as a lime filler replacement.
The AC 8 S mixture was prepared in a heated 60 l mechanical mixer to which foamed bitumen produced in the WLB-10S plant was added. The composite was mechanically mixed with a stirrer for a maximum of 5 min. The production temperature of AC 8 S with additives did not exceed 100 °C.
The mixture prepared in this way was compacted using the Marshall impact compactor. The number of impacts was dependent on the test type. After making, the asphalt concrete specimens were allowed to cool at room temperature for 48 h.
The algorithm for the design of factorial experiments was used [60]. The physical and mechanical parameters of AC 8 S were determined based on the adopted 4 × 4 factorial design, in accordance with the testing program. The developed plan of the experiment is shown in Figure 3.

The Effects of the Foamed Bitumen and Hydrated Lime on Air Void Content in Asphalt Concrete
Air voids play an important role in shaping the structure of the bituminous mixture. The amount of air voids has a large impact on the material properties of asphalt concrete. Too high an air void content negatively affects moisture resistance and performance under high loads. On the other hand, the insufficient air void content decreases the resistance of the mixture to permanent deformations despite being beneficial in terms of moisture resistance. Therefore, when examining new bituminous materials or their modifications, particular attention should be paid to this parameter. HWMA mixtures should have an air void content comparable to that obtained by traditional hot-mix asphalts if proper operation in asphalt pavement and durability are to be ensured.
The air void content of the mixture was studied in accordance with the presented methodology as per EN 12697-6: 2008. Asphalt concrete AC 8 S should have 2.0% to 4.0% air voids [58]. Test results of the AC 8 HWMA mixture with 0.6% SAA modified foamed bitumen and hydrated lime are summarized in Table 4 together with the basic statistical parameters.

The Effects of the Foamed Bitumen and Hydrated Lime on Air Void Content in Asphalt Concrete
Air voids play an important role in shaping the structure of the bituminous mixture. The amount of air voids has a large impact on the material properties of asphalt concrete. Too high an air void content negatively affects moisture resistance and performance under high loads. On the other hand, the insufficient air void content decreases the resistance of the mixture to permanent deformations despite being beneficial in terms of moisture resistance. Therefore, when examining new bituminous materials or their modifications, particular attention should be paid to this parameter. HWMA mixtures should have an air void content comparable to that obtained by traditional hot-mix asphalts if proper operation in asphalt pavement and durability are to be ensured.
The air void content of the mixture was studied in accordance with the presented methodology as per EN 12697-6: 2008. Asphalt concrete AC 8 S should have 2.0% to 4.0% air voids [58]. Test results of the AC 8 HWMA mixture with 0.6% SAA modified foamed bitumen and hydrated lime are summarized in Table 4 together with the basic statistical parameters.
Graphic representation of the quantity of hydrated lime and foamed bitumen on the quantity of air voids V a in AC 8 S is shown in Figure 4. Graphic representation of the quantity of hydrated lime and foamed bitumen on the quantity of air voids Va in AC 8 S is shown in Figure 4.  As the amount of foamed bitumen in AC 8 S mixtures increases to 6.2%, the air void content in asphalt concrete decreases, which is consistent with the general trend in this respect. However, with the binder amount of 6.5%, the air void content increases, which results from excessive binder content in the mix and its segregation. The use of 15%, 30% and 45% hydrated lime to replace mineral lime filler significantly influences variations in the value of this parameter. The addition of 15% m/m hydrated lime reduces the content of air voids, which is certainly the effect of improved adhesion of the mineral mixture-bitumen interface. Increasing hydrated lime content to 45% m/m leads to an increase in the air void content. This can be attributed to insufficient amount of binder due to the use of hydrated lime, which has a larger specific surface area than the lime filler. As a result, compaction As the amount of foamed bitumen in AC 8 S mixtures increases to 6.2%, the air void content in asphalt concrete decreases, which is consistent with the general trend in this respect. However, with the binder amount of 6.5%, the air void content increases, which results from excessive binder content in the mix and its segregation. The use of 15%, 30% and 45% hydrated lime to replace mineral lime filler significantly influences variations in the value of this parameter. The addition of 15% m/m hydrated lime reduces the content of air voids, which is certainly the effect of improved adhesion of the mineral mixture-bitumen interface. Increasing hydrated lime content to 45% m/m leads to an increase in the air void content. This can be attributed to insufficient amount of binder due to the use of hydrated lime, which has a larger specific surface area than the lime filler. As a result, compaction of the mixture becomes difficult and the content of air voids increases [61]. Regardless of the hydrated lime content, AC 8 S containing 6.2% foamed bitumen has the smallest amount of air voids. This binder content may impede maintaining adequate resistance to permanent deformations [24,25].
To comprehensively describe the variations in the AC 8 S air void content due to the changed content of the 50/70 foamed bitumen with 0.6% SAA and hydrated lime content, a statistical model using a second-degree polynomial [62] was adopted: where: In the first step, a significance test was used with analysis of variance (ANOVA) [63] (Table 5). Analysis of the parameters listed in Table 5 indicates clearly that the content of foamed bitumen and hydrated lime is an important factor that has an effect on the air void content in the AC 8 S mixture, as demonstrated by the p-value being lower than the pre-set significance level α = 0.05 (values in red). An interaction effect is present between the content of foamed bitumen and hydrated lime, which affect the air void content in the mixture (p-value less than α = 0.05).
The values describing the parameters of the regression model for the relationship between the air void content in terms of the amount of foamed bitumen and hydrated lime are summarized in Table 6. Analysis of the procedure used shows that the value of the adjusted coefficient of determination R 2 is 88% which confirms the adequacy of the model. The amounts of foamed bitumen and hydrated lime as well as the interaction of these factors have a significant impact on the air void content in the mixture.
Graphic representation of the variation in air void content in AC 8 S as a function of the amount of foamed bitumen and hydrated lime is shown in Figure 5. Graphic representation of the variation in air void content in AC 8 S as a function of the amount of foamed bitumen and hydrated lime is shown in Figure 5. Analysis of the results presented in Figure 5 confirms that as the amount of foamed bitumen and hydrated lime increases, the air void content in the asphalt concrete decreases throughout the experiment. At the same time, hydrated lime has a significant impact on the assessed parameter in the range from 5.6% to 5.9% of 50/70 foamed bitumen as it contributes to an increase in air voids; its recommended content (max. 4.0%) is even exceeded. The increase in foamed bitumen content in the range from 5.9% to 6.2% has a positive effect on this parameter and enables it to obtain the recommended values [58]. A further increase in foamed bitumen content to 6.5% results in a significant air voids reduction to the level below the required level. [58].

The Effects of the Foamed Bitumen and Hydrated Lime on Moisture Resistance of Asphalt Concrete
In accordance with the adopted research plan, the assessment of moisture and frost resistance of the AC 8 S mixture with foamed bitumen and hydrated lime was based on the WT-2 2014 procedure [58] and on the modified AASHTO T283 procedure [50,51]. The test methodology according to the modified AASHTO T283 method is based on a fairly aggressive asphalt concrete conditioning scheme, which is to simulate very variable and stringent climatic conditions.

Water Sensitivity According to WT-2 2014
The procedure for assessing water sensitivity of an asphalt mix according to the requirements of WT-2 2014 [58] has only one freeze-thaw cycle on water-saturated specimens. The basis for the assessment is the ITSR, which should be greater than 90% for the bituminous mixture designed for the wearing course to be resistant to moisture and frost. When the mixture is characterized by high fineness and rather high binder content, this ratio reaches a value greater than 100% [58,64].
In the first stage of AC 8 moisture and frost resistance assessment, first the indirect tensile strength ITSd for unconditioned specimens and then ITSw for moisture-conditioned specimens were determined to WT-2 2014 [58]. The test results are summarized in Table 7 and graphically represented in Figure 6. Analysis of the results presented in Figure 5 confirms that as the amount of foamed bitumen and hydrated lime increases, the air void content in the asphalt concrete decreases throughout the experiment. At the same time, hydrated lime has a significant impact on the assessed parameter in the range from 5.6% to 5.9% of 50/70 foamed bitumen as it contributes to an increase in air voids; its recommended content (max. 4.0%) is even exceeded. The increase in foamed bitumen content in the range from 5.9% to 6.2% has a positive effect on this parameter and enables it to obtain the recommended values [58]. A further increase in foamed bitumen content to 6.5% results in a significant air voids reduction to the level below the required level. [58].

The Effects of the Foamed Bitumen and Hydrated Lime on Moisture Resistance of Asphalt Concrete
In accordance with the adopted research plan, the assessment of moisture and frost resistance of the AC 8 S mixture with foamed bitumen and hydrated lime was based on the WT-2 2014 procedure [58] and on the modified AASHTO T283 procedure [50,51]. The test methodology according to the modified AASHTO T283 method is based on a fairly aggressive asphalt concrete conditioning scheme, which is to simulate very variable and stringent climatic conditions.

Water Sensitivity According to WT-2 2014
The procedure for assessing water sensitivity of an asphalt mix according to the requirements of WT-2 2014 [58] has only one freeze-thaw cycle on water-saturated specimens. The basis for the assessment is the ITSR, which should be greater than 90% for the bituminous mixture designed for the wearing course to be resistant to moisture and frost. When the mixture is characterized by high fineness and rather high binder content, this ratio reaches a value greater than 100% [58,64].
In the first stage of AC 8 moisture and frost resistance assessment, first the indirect tensile strength ITS d for unconditioned specimens and then ITS w for moisture-conditioned specimens were determined to WT-2 2014 [58]. The test results are summarized in Table 7 and graphically represented in Figure 6.  An increase in the 50/70-grade foamed bitumen content in the AC 8 S mixture lowers ITSd and ITSw values, which is in line with the general trend in this respect. The use of hydrated lime increases ITSd and ITSw. With 5.6% binder in the mixture, this relationship only occurs with the addition of 15% hydrated lime. When its content increases, the value of the analyzed parameter decreases. This may be due to the high concentration of hydrated lime and a small amount of binder, which stiffens the An increase in the 50/70-grade foamed bitumen content in the AC 8 S mixture lowers ITS d and ITS w values, which is in line with the general trend in this respect. The use of hydrated lime increases ITS d and ITS w . With 5.6% binder in the mixture, this relationship only occurs with the addition of 15% hydrated lime. When its content increases, the value of the analyzed parameter decreases. This may be due to the high concentration of hydrated lime and a small amount of binder, which stiffens the AC 8 S mixture. Higher hydrated lime content tends to increase ITS d and ITS w . This trend increases with the amount of hydrated lime increased to 30%. It is most likely the effect of a favorable improvement of the binder adhesion to the aggregate and its interaction as a mineral lime filler with very small particle size [61]. Increasing the hydrated lime content to 45% has a negative effect as a result of introducing too much fine-grained material and insufficient amount of binder to coat it, leading to excessive stiffening of the mixture. Attention should be paid to adsorption at the interface between the strongly basic mineral material (hydrated lime) and the binder, which contributes to a decrease in the content of "free binder" in the structure of the mixture [61]. This further increases its rigidity and lowers the values of ITS d and ITS w [40].
Overall, with the binder content of 5.9% and hydrated lime content of 30% in the AC 8 S HWMA mixture, the indirect tensile strength ITS d and ITS w met the normative requirements of WT-2 2014 [58] most favorably.
To comprehensively describe the relationship of indirect tensile strength ITS d and ITS w of the AC 8 HWMA mixture with foamed bitumen 50/70, 0.6% SAA and hydrated lime, the second-degree polynomial model was adopted. In the first stage of model assessment, the significance test was performed using ANOVA [63] (Table 8). Analysis of the parameters listed in Table 8 indicates clearly that foamed bitumen and hydrated lime constitute a factor that has a significant effect on the indirect tensile strength, ITS d and ITS w , of the AC 8 S, as demonstrated by p-values being less than the pre-defined significance level α = 0.05. Statistical significance was not observed in the assessment of linear effects of hydrated lime on ITS d and foamed bitumen on ITS w . Interactions found between the foamed bitumen and hydrated lime contents also affect the value of this parameter (p-value less than α = 0.05).
The parameters describing the regression model for ITS d and ITS w of AC 8 S as a function of the amounts of foamed bitumen and hydrated lime are given in Table 9. The adjusted coefficient of determination R 2 of the model for ITS d and ITS w is nearly 53%, which confirms the adequacy of the model. Both the amount of foamed bitumen 50/70 alone and the amount of hydrated lime alone have a significant impact on the flexural tensile strength, ITS d and ITS w of the AC 8 HWMA mixture. An interaction of these factors is observed (p-value is less than the pre-defined significance level α = 0.05).
Graphic representation of the variation in the ITS d and ITS w values of the AC 8 S mixture in terms of foamed bitumen and hydrated lime is shown in Figure 7. The adjusted coefficient of determination R 2 of the model for ITSd and ITSw is nearly 53%, which confirms the adequacy of the model. Both the amount of foamed bitumen 50/70 alone and the amount of hydrated lime alone have a significant impact on the flexural tensile strength, ITSd and ITSw of the AC 8 HWMA mixture. An interaction of these factors is observed (p-value is less than the pre-defined significance level α = 0.05).
Graphic representation of the variation in the ITSd and ITSw values of the AC 8 S mixture in terms of foamed bitumen and hydrated lime is shown in Figure 7. Analysis of the results presented in Figure 6 confirms that with an increase in the amount of foamed bitumen and hydrated lime, the value of dry tensile strength ITSd as per WT-2 2014 [58] reaches a maximum at a 6.2% bitumen content in the mixture. At the same time, hydrated lime has a significant impact on the ITSd in virtually the entire dosing range.
A comprehensive analysis of the results based on the assessment of the relationship presented in Figure 7 shows that as the amounts of foamed bitumen increase and at hydrated lime increase from 15% to 30%, the wet tensile strength ITSw of the specimens conditioned as per WT-2 2014 increases significantly [58]. Please note the synergy of hydrated lime and foamed bitumen that significantly affect the ITSw of the mixture.
Assessment of the resistance of the AC 8 S mixture to climatic factors according to WT-2 2014 is based on the ITSR expressed as the ratio of the tensile strength of moisture-conditioned specimen to the tensile strength of air-conditioned specimen (ITSw/ITSd).
The methodology developed as per WT-2 2014 assumes that the AC 8 S bituminous mixture is resistant to climatic factors (water and frost) when the ITSR reaches at least 90% [58].
Graphic representation of the ITSR as a function of technology used, the amount of binder, and Analysis of the results presented in Figure 6 confirms that with an increase in the amount of foamed bitumen and hydrated lime, the value of dry tensile strength ITS d as per WT-2 2014 [58] reaches a maximum at a 6.2% bitumen content in the mixture. At the same time, hydrated lime has a significant impact on the ITS d in virtually the entire dosing range.
A comprehensive analysis of the results based on the assessment of the relationship presented in Figure 7 shows that as the amounts of foamed bitumen increase and at hydrated lime increase from 15% to 30%, the wet tensile strength ITS w of the specimens conditioned as per WT-2 2014 increases significantly [58]. Please note the synergy of hydrated lime and foamed bitumen that significantly affect the ITS w of the mixture.
Assessment of the resistance of the AC 8 S mixture to climatic factors according to WT-2 2014 is based on the ITSR expressed as the ratio of the tensile strength of moisture-conditioned specimen to the tensile strength of air-conditioned specimen (ITS w /ITS d ).
The methodology developed as per WT-2 2014 assumes that the AC 8 S bituminous mixture is resistant to climatic factors (water and frost) when the ITSR reaches at least 90% [58].
Graphic representation of the ITSR as a function of technology used, the amount of binder, and the hydrated lime content is shown in Figure 8. Analysis of the test results indicates that the AC 8 S mixture is resistant to moisture and frost in the entire range of use of 50/70-grade foamed bitumen with 0.6% SAA and hydrated lime-the ITSR always reaches values higher than 90%. The variation in its value depending on the constituents used (foamed bitumen, hydrated lime) is analogous to that of ITSd and ITSw.
Overall, it was observed that with 5.9% binder and 30% hydrated lime in the filler, the AC 8 S asphalt concrete has obtained the greatest increase in the water sensitivity described by the ITSR. The mixtures with higher binder contents have yielded slightly higher ITSR results; however, the high bitumen content could be a concern as for the resistance to permanent deformation performance and cost of such mixtures.
The relationship between the moisture resistance of the AC 8 HWMA mixture with 0.6% SAA modified 50/70-grade foamed bitumen and hydrated lime was described using the second-degree polynomial model. In the first stage in the model assessment, the significance test was performed using ANOVA variance analysis [63] (Table 10).  Analysis of the test results indicates that the AC 8 S mixture is resistant to moisture and frost in the entire range of use of 50/70-grade foamed bitumen with 0.6% SAA and hydrated lime-the ITSR always reaches values higher than 90%. The variation in its value depending on the constituents used (foamed bitumen, hydrated lime) is analogous to that of ITS d and ITS w .
Overall, it was observed that with 5.9% binder and 30% hydrated lime in the filler, the AC 8 S asphalt concrete has obtained the greatest increase in the water sensitivity described by the ITSR. The mixtures with higher binder contents have yielded slightly higher ITSR results; however, the high bitumen content could be a concern as for the resistance to permanent deformation performance and cost of such mixtures.
The relationship between the moisture resistance of the AC 8 HWMA mixture with 0.6% SAA modified 50/70-grade foamed bitumen and hydrated lime was described using the second-degree polynomial model. In the first stage in the model assessment, the significance test was performed using ANOVA variance analysis [63] (Table 10).
Analysis of the parameters listed in Table 10 indicates clearly that foamed bitumen and hydrated lime contents constitute a factor that has a significant effect on the indirect tensile strength ratio ITSR in the AC 8 S, as demonstrated by the p-values being less than the pre-defined significance level α = 0.05. No statistical significance was observed only in the case of the quadratic component of the model, related to the bitumen amount, and of the interaction term. The parameters describing the regression model of the ITSR of the AC 8 S as a function of foamed bitumen and hydrated lime contents are given in Table 11. Table 11. Parameters of the model of the relationship between ITSR and the amounts of foamed bitumen with 0.6% SAA and hydrated lime. The adjusted coefficient of determination R 2 is nearly 58%, which indicates the adequacy of the model. Hydrated lime has a significant effect on the value of ITSR for the AC 8 S HWMA mixture (p-value is less than the pre-defined significance level α = 0.05). A deviation from the presented trend exists for the linear relationship between the lime and the linear and squared terms related to the foamed bitumen content in the mixture.

Response
Graphic representation of the variation in the ITSR value of the AC 8 S mixture in terms of the contents of foamed bitumen-grade 50/70 and hydrated lime is shown in Figure 9. Analysis of the parameters listed in Table 10 indicates clearly that foamed bitumen and hydrated lime contents constitute a factor that has a significant effect on the indirect tensile strength ratio ITSR in the AC 8 S, as demonstrated by the p-values being less than the pre-defined significance level α = 0.05. No statistical significance was observed only in the case of the quadratic component of the model, related to the bitumen amount, and of the interaction term.
The parameters describing the regression model of the ITSR of the AC 8 S as a function of foamed bitumen and hydrated lime contents are given in Table 11. The adjusted coefficient of determination R 2 is nearly 58%, which indicates the adequacy of the model. Hydrated lime has a significant effect on the value of ITSR for the AC 8 S HWMA mixture (pvalue is less than the pre-defined significance level α = 0.05). A deviation from the presented trend exists for the linear relationship between the lime and the linear and squared terms related to the foamed bitumen content in the mixture.
Graphic representation of the variation in the ITSR value of the AC 8 S mixture in terms of the contents of foamed bitumen-grade 50/70 and hydrated lime is shown in Figure 9. Comprehensive analysis of the results based on the relationships shown in Figure 9 indicates clearly that the contents of foamed bitumen increasing from 5.9% to 6.5% and hydrated lime increasing from 0% do 30% have a significant effect of increased ITSR, which corresponds to the resistance of the AC 8 S mixture to water and frost as per WT-2 2014 [58]. An interaction effect between the hydrated lime and foamed bitumen provides the mixture with water and frost resistance. Comprehensive analysis of the results based on the relationships shown in Figure 9 indicates clearly that the contents of foamed bitumen increasing from 5.9% to 6.5% and hydrated lime increasing from 0% do 30% have a significant effect of increased ITSR, which corresponds to the resistance of the AC 8 S mixture to water and frost as per WT-2 2014 [58]. An interaction effect between the hydrated lime and foamed bitumen provides the mixture with water and frost resistance.
At the same time, it is important to determine the correlation between the air void content V a and ITSR ( Figure 10). The correlation presented is linear and statistically significant, as the p-value is less than the predefined significance level α = 0.05.

Resistance to Water Damage According to the Modified AASHTO T283 Method
The procedure of assessing moisture and frost resistance according to the modified AASHTO T283 method consists of subjecting specimens of the bituminous mixture to 18 freeze-thaw cycles. In comparison with WT-2 (one freezing cycle), this procedure is more severe in terms of exposure of asphalt concrete to climatic factors. The simulation of the in-service conditions is closer to natural conditions. Thus, the results of moisture and frost resistance tests according to the modified AASHTO T283 method represent the behavior of the mixture in real conditions better than those obtained from the WT-2 2014 procedure.
The moisture-induced damage resistance (AASHTO T283) is reflected in the value of RWwm, which is defined as the indirect tensile strength of moisture-conditioned specimens divided by the indirect tensile strength of dry specimens.
In the first step of the assessment of AC 8 S moisture e resistance as per AASHTO T283, indirect tensile strength tests for dry (ITSd A ) and moisture-conditioned specimens (ITSw A ) were performed, as shown in Table 12. Graphic representation of the results is in Figure 11. The correlation presented is linear and statistically significant, as the p-value is less than the pre-defined significance level α = 0.05.

Resistance to Water Damage According to the Modified AASHTO T283 Method
The procedure of assessing moisture and frost resistance according to the modified AASHTO T283 method consists of subjecting specimens of the bituminous mixture to 18 freeze-thaw cycles. In comparison with WT-2 (one freezing cycle), this procedure is more severe in terms of exposure of asphalt concrete to climatic factors. The simulation of the in-service conditions is closer to natural conditions. Thus, the results of moisture and frost resistance tests according to the modified AASHTO T283 method represent the behavior of the mixture in real conditions better than those obtained from the WT-2 2014 procedure.
The moisture-induced damage resistance (AASHTO T283) is reflected in the value of RW wm , which is defined as the indirect tensile strength of moisture-conditioned specimens divided by the indirect tensile strength of dry specimens.
In the first step of the assessment of AC 8 S moisture e resistance as per AASHTO T283, indirect tensile strength tests for dry (ITS d A ) and moisture-conditioned specimens (ITS w A ) were performed, as shown in Table 12. Graphic representation of the results is in Figure 11.  Figure 11. Relationship between the indirect tensile strength of dry specimens ITSd A and moistureconditioned specimens ITSw A (modified AASHTO T283) of the AC 8 S mixture versus the technology used, amount of hydrated lime and foamed bitumen: (a) 5.6%, (b) 5.9%, (c) 6.2%, (d) 6.5%.
Analysis of the results of flexural tensile strength ITSd A for the dry-conditioned AC 8 S mixture indicates that hydrated lime up to 30% increases the analyzed parameter for the HWMA mixture. Increasing its content to 45% leads to a decrease in ITSd A . With increasing concentration of the binder in the mixture, the value of the analyzed parameter decreases, which is consistent with the observations made so far [33].
It should also be noted that the presented trend of dry indirect tensile strength (modified AASHTO T283) ITSd A for the AC 8 S in terms of the applied technology, amount of binder, and hydrated lime content has the same character as that obtained from the WT-2 2014 procedure [58].
It can thus be concluded that a favorable flexural tensile strength was obtained with the 5.9% binder content (bitumen 50/70 + 0.6% SAA by weight) and 30% concentration of hydrated lime in the filler added to the AC 8 S HWMA mixture.
Analysis of the results of the effect of 50/70-grade bitumen content with 0.6% SAA and hydrated lime on the flexural indirect tensile strength ITSw A of the moisture-conditioned mixture indicates that they exhibit the same trend as those from "dry" conditioned specimens.
Overall, the binder content of 5.9% and hydrated lime content of 30% in the AC 8 S HWMA mixture resulted in a favorable flexural tensile strength.
Comprehensive description of the relationships of ITSw A of the AC 8 S HWMA mixture with foamed 50/70 bitumen, 0.6% SAA addition and hydrated lime was prepared using the second-degree polynomial model. The first stage of the model evaluation involved performing a significance test using the ANOVA [63] (Table 13). Analysis of the results of flexural tensile strength ITS d A for the dry-conditioned AC 8 S mixture indicates that hydrated lime up to 30% increases the analyzed parameter for the HWMA mixture. Increasing its content to 45% leads to a decrease in ITS d A . With increasing concentration of the binder in the mixture, the value of the analyzed parameter decreases, which is consistent with the observations made so far [33]. It should also be noted that the presented trend of dry indirect tensile strength (modified AASHTO T283) ITS d A for the AC 8 S in terms of the applied technology, amount of binder, and hydrated lime content has the same character as that obtained from the WT-2 2014 procedure [58]. It can thus be concluded that a favorable flexural tensile strength was obtained with the 5.9% binder content (bitumen 50/70 + 0.6% SAA by weight) and 30% concentration of hydrated lime in the filler added to the AC 8 S HWMA mixture.
Analysis of the results of the effect of 50/70-grade bitumen content with 0.6% SAA and hydrated lime on the flexural indirect tensile strength ITS w A of the moisture-conditioned mixture indicates that they exhibit the same trend as those from "dry" conditioned specimens. Overall, the binder content of 5.9% and hydrated lime content of 30% in the AC 8 S HWMA mixture resulted in a favorable flexural tensile strength.
Comprehensive description of the relationships of ITS w A of the AC 8 S HWMA mixture with foamed 50/70 bitumen, 0.6% SAA addition and hydrated lime was prepared using the second-degree polynomial model. The first stage of the model evaluation involved performing a significance test using the ANOVA [63] (Table 13).  Table 13 shows that the contents of foamed bitumen and hydrated lime constitute a significant factor affecting the indirect tensile strength ITS d A and ITS w A of the AC 8 S, as the p-values are less than the pre-defined significance level α = 0.05. No statistical significance was observed for the quadratic terms related to the binder effect.

Analysis of the parameters compiled in
No interaction effects were found between the foamed bitumen content and hydrated lime content. This can be attributed to the stringent conditioning of the specimens as a result of which this factor did not show a significant effect on the parameter. Consequently, statistically significant synergy of the hydrated lime and foamed bitumen did not occur.
The values describing the parameters of the regression model of the response surface for the indirect tensile strength ITS d A and ITS w A of the AC 8 S mixture in terms of the foamed bitumen and hydrated lime contents are summarized in Table 14. Hydrated lime (p-value is less than the pre-defined significance level α = 0.05) has a statistically significant effect on the indirect tensile strength ITS w A of the AC 8 S HWMA mixture. However, no effect of the foamed bitumen content and no interaction between the contents of bitumen and hydrated lime were observed, which seems to be a consequence of rigorous conditions of specimen conditioning in the freeze-thaw process. Graphic representation of the variation in the flexural indirect tensile strength ITS d A and ITS w A of the AC 8 S mixture versus the contents of foamed 50/70-grade bitumen and hydrated lime is shown in Figure 12. The adjusted coefficient of determination R 2 for ITSd A is nearly 62% and 60% for ITSw A , which indicates the adequacy of the models. Hydrated lime has a significant influence on the dry indirect tensile strength ITSd A of the AC 8 S HWMA mixture (p-value is less than the pre-defined significance level α = 0.05).
Hydrated lime (p-value is less than the pre-defined significance level α = 0.05) has a statistically significant effect on the indirect tensile strength ITSw A of the AC 8 S HWMA mixture. However, no effect of the foamed bitumen content and no interaction between the contents of bitumen and hydrated lime were observed, which seems to be a consequence of rigorous conditions of specimen conditioning in the freeze-thaw process.
Graphic representation of the variation in the flexural indirect tensile strength ITSd A and ITSw A of the AC 8 S mixture versus the contents of foamed 50/70-grade bitumen and hydrated lime is shown in Figure 12. Comprehensive analysis of the results based on the response surface shown in Figure 12 indicates that increasing the content of foamed bitumen and the content of hydrated lime from 15% to 30% has a significant effect on the increase in wet indirect tensile strength ITSw A of the AC 8 S mixture (the modified AASHTO T283 procedure).
The resistance of the AC 8 S mixture to climatic factors as per AASHTO T283 is assessed based Comprehensive analysis of the results based on the response surface shown in Figure 12 indicates that increasing the content of foamed bitumen and the content of hydrated lime from 15% to 30% has a significant effect on the increase in wet indirect tensile strength ITS w A of the AC 8 S mixture (the modified AASHTO T283 procedure). The resistance of the AC 8 S mixture to climatic factors as per AASHTO T283 is assessed based on the RW WM , which is an ITS w A /ITS d A ratio. Calculation results are shown graphically in Figure 13. Comprehensive analysis of the results based on the response surface shown in Figure 12 indicates that increasing the content of foamed bitumen and the content of hydrated lime from 15% to 30% has a significant effect on the increase in wet indirect tensile strength ITSw A of the AC 8 S mixture (the modified AASHTO T283 procedure).
The resistance of the AC 8 S mixture to climatic factors as per AASHTO T283 is assessed based on the RWWM, which is an ITSw A /ITSd A ratio. Calculation results are shown graphically in Figure 13. According to the methodology developed based on the modified AASHTO T283 procedure, a bituminous mixture is resistant to climatic factors-moisture and frost-when the RW WM is at least 80%.
Analysis of the results shows that increasing content of the binder in the AC 8 S mixture increases the value of RW WM , hence, the resistance to moisture damage. This response was expected and compliant with the general principles of asphaltic materials technology. The use of hydrated lime increases the value of this parameter and at 30% lime in the limestone dust, the resistance to moisture, and frost satisfies the requirements. A further increase in the lime content decreases the value of this parameter which can be attributed to a significant increase in the specific area of the filler (mineral dust + hydrated lime) in relation to the binder content.
A comprehensive description of moisture and frost resistance of the AC 8 S HWMA mixture with foamed bitumen, 0.6% SAA and hydrated lime was prepared using the second-degree polynomial model. The first stage of the model evaluation included a significance test using ANOVA variance analysis (Table 15). Analysis of the parameters presented in Table 15 shows that the contents of foamed bitumen and hydrated lime constituted a significant factor that influenced the moisture and frost resistance (RW WM ) of the AC 8 S mixture, as demonstrated by the p-values being less than the pre-defined significance level α = 0.05. This relationship was not observed only for the square term describing the influence of bitumen content. The existence of interaction between the content of foamed bitumen and hydrated lime, which influence the analyzed parameter (p-value is less than α = 0.05), is important, as it supports the use of foamed bitumen and hydrated lime in the mixture.
The value of the adjusted R 2 is almost 76%, which indicates the adequacy of the adopted model. The parameters of the developed regression model of the relationship between the RW WM of the AC 8 S mixture and the foamed bitumen and hydrated lime contents are presented in Table 16. Analysis of the parameters listed in Table 16 shows that hydrated lime is a significant factor influencing the resistance of the AC 8 S mixture to moisture, characterized by the RW WM , as the p-value is less than the pre-defined significance level α = 0. An interaction exists of the influence of foamed asphalt and hydrated lime with respect to the RW WM (p-value less then α = 0.05) as per the modified AASHTO T283 method.
Graphic representation of moisture and frost resistance relationship model-RW WM (AASHTO T283) of the AC 8 S HWMA mixture in terms of foamed asphalt and hydrated lime contents is presented in Figure 14 together with the developed model. value is less than the pre-defined significance level α = 0. An interaction exists of the influence of foamed asphalt and hydrated lime with respect to the RWWM (p-value less then α = 0.05) as per the modified AASHTO T283 method.
Graphic representation of moisture and frost resistance relationship model-RWWM (AASHTO T283) of the AC 8 S HWMA mixture in terms of foamed asphalt and hydrated lime contents is presented in Figure 14 together with the developed model. Comprehensive analysis of the results based on the relationship shown in Figure 12 indicates that the contents of foamed bitumen increasing from 5.9% to 6.2% and hydrated lime from 15% to 30% have a significant effect on the RWWM increase, which corresponds to the resistance to moisture and frost (modified AASHTO T283). Hydrated lime and foamed bitumen interact to provide resistance to moisture-induced damage.
To assess the effect of hydrated lime and 50/70-grade foamed bitumen with an addition of 0.6% SAA on the properties of the AC 8 S bituminous mixture, the correlation between the air void content Va and resistance to water and frost RWWM ( Figure 15) was analyzed. Comprehensive analysis of the results based on the relationship shown in Figure 12 indicates that the contents of foamed bitumen increasing from 5.9% to 6.2% and hydrated lime from 15% to 30% have a significant effect on the RW WM increase, which corresponds to the resistance to moisture and frost (modified AASHTO T283). Hydrated lime and foamed bitumen interact to provide resistance to moisture-induced damage.
To assess the effect of hydrated lime and 50/70-grade foamed bitumen with an addition of 0.6% SAA on the properties of the AC 8 S bituminous mixture, the correlation between the air void content V a and resistance to water and frost RW WM (Figure 15) was analyzed. The statistically significant correlation between the air void content Va of AC 8 S and its resistance to water and frost (RWWM) has a linear character (p-value less than α = 0.05).

Optimization of the Foamed Bitumen and Hydrated Lime Contents in the AC 8 S Mixture in Terms of its Resistance to Moisture and Frost
To determine the recommended amount of foamed asphalt and hydrated lime to obtain the most favorable parameters characterizing the resistance of AC 8 S to moisture and frost, the following The statistically significant correlation between the air void content V a of AC 8 S and its resistance to water and frost (RW WM ) has a linear character (p-value less than α = 0.05).

Optimization of the Foamed Bitumen and Hydrated Lime Contents in the AC 8 S Mixture in Terms of its Resistance to Moisture and Frost
To determine the recommended amount of foamed asphalt and hydrated lime to obtain the most favorable parameters characterizing the resistance of AC 8 S to moisture and frost, the following parameters were analyzed: • air void content (V a ), • water sensitivity to WT-2 (ITSR), • moisture-induced damage resistance to modified AASHTO T283 method (RW WM ).
Analysis of the relationships between the evaluated properties of asphalt concrete AC 8 S is an important element in the assessment of the effects of hydrated lime and 50/70-grade foamed bitumen with the addition of 0.6% SAA. Correlations between these properties are summarized in Table 17. The main correlation parameter was the air void content V a , which has a significant effect on other properties of AC 8 S. The correlation values represent most of the results obtained in this experiment. From the results it follows that they are mostly non-linear or that statistically significant interactions exist.
The characteristics of the models describing the analyzed relationships of the AC 8 S parameters in terms of the contents of 50/70-grade foamed bitumen and hydrated lime are presented in Table 18. To assess the performance of AC 8 S, criteria were adopted, according to which the most desirable values of the parameters were assigned the performance indicator equal to 1 and the values least desired: 0. Intermediate values obtained indicators from the range 0 to 1 in the linear relationship. The used optimization procedure has been described in detail in [65]. The following criteria were applied for individual parameters of asphalt concrete: Air void content V a (max: 0, min: 1), ITSR according to WT-2 (max: 1, min: 0), RW WM according to the modified AASHTO T283 method (max: 1, min: 0).
Then the values of the utility function of the asphalt concrete were calculated. The approximated results are plotted in Figure 16.
relationship. The used optimization procedure has been described in detail in [65]. The following criteria were applied for individual parameters of asphalt concrete: Air void content Va (max: 0, min: 1), ITSR according to WT-2 (max: 1, min: 0), RWWM according to the modified AASHTO T283 method (max: 1, min: 0). Then the values of the utility function of the asphalt concrete were calculated. The approximated results are plotted in Figure 16.  On the basis of the utility analysis of the approximated values it was found that in order to ensure the moisture and frost resistance of asphalt concrete AC 8 S, the recommended content of 50/70 foamed bitumen with 0.6 % SAA should be 6.05 % and that of hydrated lime in the mineral filler 22.5 %.

Conclusion
The following conclusions were drawn based on the moisture resistance tests of AC 8 S HWMA mixture:

•
The use of 50/70-grade foamed bitumen with an addition of 0.6% SAA and hydrated lime in the AC 8 S mixture has a significant effect on its properties. The intensity of the effects of these constituents varies by parameter.

•
Hydrated lime used as a replacement for limestone dust is important for air void content. With up to 30%, it has a beneficial effect on the air void content regardless of the foamed bitumen content in the asphalt concrete. When it is used at more than 30%, the trend changes, which can be attributed to the fact that at higher concentrations hydrated lime impedes compaction due to higher demand for bitumen as a result of a larger specific area compared to that of the mineral dust. • Similar relationship is observed in the case of the resistance to moisture and frost, ITSR, determined to WT-2 2014. The interaction between the contents of hydrated lime and foamed bitumen plays a role here as they influence this parameter of the mixture.

•
Moisture-induced damage test to the modified AASHTO T283 method indicates that 15% and more hydrated lime used in the mixture composition provides optimum values of this characteristic. There is also an interaction between the contents of hydrated lime and binder, with a significant effect on the resistance of the mixture to water and frost (the modified AASHTO T283 method).

•
The authors found the synergy of the hydrated lime and foamed bitumen, depending on their contents in the asphalt concrete, for ensuring the resistance of the AC 8 S HWMA mixture to moisture and frost.

•
Regardless of the research methods used, resistance to moisture and frost can be ensured by optimizing the contents of foamed bitumen and hydrated lime in the HWMA mixture.

•
Optimization of the AC 8 S HWMA mixture in terms of the evaluated parameters determined the contents of hydrated lime and foamed bitumen at, considering the dosing tolerance, 30% and 5.9. The in-service resistance of the pavement wearing course to moisture and frost will be ensured.
The conducted research has proven that correct mixture design permits superior moisture and frost resistance of the tested AC 8 S HWMA with foamed bitumen. It was shown that incorporation of hydrated lime as a significant part of the filler fraction (30%) strongly contributed to the performance of the HWMA mixture when optimum amount of foamed bitumen was used. Despite the high surface area of the hydrated lime requiring increased amounts of bitumen, the hydrated lime bearing HWMA mixtures exhibited adequate volumetric performance and increased resistance to frost and moisture damage tested by the means of different methods. The incorporation of hydrated lime has proven to be a feasible method for ensuring the adequate performance and durability of the AC 8 S HWMA mixture with foamed bitumen.