From Waste to Roads: Improving Pavement Performance and Achieving Sustainability with Recycled Steel Slag and Low-Density Polyethylene

Abstract

The use of waste, recycled, and modified materials is increasingly popular in roadway construction for sustainability and pavement longevity. This research examines the combination of steel slag (SS) and low-density polyethylene (LDPE), commonly used in plastic bags and steel manufacturing by-products, to mitigate environmental pollution. LDPE was tested as a binder modifier in two bitumen grades, 60–70 and 80–100, at concentrations of 3%, 5%, and 7% by weight. SS was used as a replacement for coarse aggregate. The physical properties of both modified and unmodified bitumen grades and SS were analyzed before creating and testing hot-mix asphalt (HMA) samples. The dynamic modulus of these samples was measured at temperatures of 4.4 °C, 21.1 °C, 37.8 °C, and 54.4 °C with frequencies of 0.1 Hz, 0.5 Hz, 1 Hz, 5 Hz, 10 Hz, and 25 Hz. Master curves were developed, and the dynamic modulus data underwent design of experiment (DOE) and computational intelligence (CI) analyses. Using KENPAVE, a mechanistic–empirical tool, the analysis assessed the design life and enhancements in damage ratio for each modifier and grade. The results showed that adding LDPE increases the softening point and penetration grade but decreases ductility due to increased bitumen stiffness, leading to premature fatigue failure at higher LDPE levels. Both 3% LDPE and 3% SS-modified LDPE improved Marshall Stability and dynamic modulus across all temperature and frequency ranges. Specifically, 3% LDPE enhanced stability by 13–16% and 3% SS-LDPE by 30–32%. The KENPAVE results for 3% LDPE showed a design life improvement of 19–25% and a damage ratio reduction of 15–18%. In comparison, 3% SS-LDPE demonstrated a design life improvement of 50–60% and a damage ratio reduction of 25–35%. Overall, this study concludes that 3% LDPE- and 3% SS-LDPE-modified HMA in both bitumen grades 60–70 and 80–100 provide optimal results for improving pavement performance.

1. Introduction

A country’s transportation infrastructure is crucial to its growth and a mark of an industrialized country and developed nation. Most roads are typically made of hot-mix asphalt (HMA), which is a combination of bitumen and aggregates [1,2]. Many authors have researched the stability of bitumen and the replacement of aggregates to improve the bonding between them, either to provide better stability or to improve ride quality [3,4,5]. The need for aggregate replacements increases the utilization of the waste produced during different construction and demolition projects, which makes up for the shortcoming of aggregates [6]. Similarly, the need for the stability of bitumen increases to maintain a pavement’s load-carrying capability, prevent moisture damage, and improve the bond between aggregates and bitumen [7,8].

Shortage of materials is a rising issue in construction work; thus, using sustainable materials is now a developing area of research [9]. The current situation requires an economic and attractive road system with low economic costs. This method has become important for the conservation and management of our resources. Pavement repair has come a long way in the last 25 years. It is a professional and environmentally friendly process to modify and improve existing asphalt [10]. Many researchers have used different industrial wastes such as recycled construction waste, wastes from the steel industry in the form of slag, over-burnt brick waste, reclaimed asphalt waste, treated soil, etc., which are recommend materials for achieving sustainability [11,12,13,14,15,16,17]. With the steel industry’s aggressive growth, the use of steel slag has expanded [18]. Asphalt concrete can be successfully mixed with steel slag in place of aggregate and can improve its stability and thermal efficiency [19]. The use of steel slag is restricted to substitution either as coarse aggregates or as a fine aggregate only, because the entire replacement increases angularity, which can cause an increase in the void ratio and bulking [20]. Flexible pavement suffers from premature failure due to friction and fatigue cracks, which causes the road materials to fail before they reach the end of their intended lifespan. This is mainly due to the dramatic increase in road traffic that has occurred in the last twenty years. Shear flow, a form of rutting, is a typical failure of road materials that is mostly caused by vehicle stress, high temperatures, and heavy ESALs [21]. Three main causes of road material disintegration have been identified, namely temperature and surrounding conditions, heavy vehicle use, and negligence or quality compromise in construction.

The temperature of bitumen greatly influences the hardness and stiffness properties of bituminous mix, which is solely due to the viscoelastic nature of bitumen, which changes with the thermal gradient [22]. Lower winter temperatures stiffen the bitumen, which causes fatigue cracking, while higher summer temperatures increase the viscosity of bitumen, which causes rutting. As a result, higher temperatures, as seen in Pakistan, significantly degrade the strength of the asphalt material. In addition, if a sustained load is maintained on the asphalt for a longer period, damage is greatly increased, and shear deformation and plastic flow occur due to excessive shear and overloaded truck stresses [23].

Asphalt modification and recycling meet all the social goals of providing efficient and safe roads while significantly reducing energy (oil) consumption and environmental impact compared to road improvements [24]. Thus, it has a lasting impact and allows supplying agencies to increase their available funds. However, they must be used properly, as not all highways are ideal candidates for asphalt recycling [25]. This will provide a reliable and safe driving surface for travelers.

Bitumen is a viscoelastic material that is obtained from crude oil and is mainly composed of hydrocarbons; very few materials can replace it. One of the problems of bitumen is its temperature susceptibility, specifically its propensity to become brittle at lower temperatures and softened at higher temperatures, which may be modified by some additives and polymers [26]. Additives such as polyethylene, styrene, crumb rubber, ultra-high-molecular-weight polyethylene, and many others have been mixed in asphalt concrete to improve certain engineering characteristics of asphaltic mix [27,28,29,30]. In comparison to conventional asphaltic pavements, the deterioration of pavement materials can be reduced by partially adding waste plastic to bitumen [31].

The physical behavior and mechanical characteristics of flexible pavements after partially replacing waste plastic with low- and high-density polyethylene reveal that the waste both reduces the density and increases the stability [32]. Zhang et al. [33] studied the influence of polyethylene wax and selected three types of additives with different molecular weights. The results indicated that all types of additives improved the asphalt mixture’s ability to resist moisture and maintain stability at high temperatures. Recycling waste materials in HMA is the best choice for constructing pavements while keeping other necessary volumetric properties since several motorway restoration projects have been planned with HIR and CIR technologies (hot-in-place and cold-in-place recycling, respectively), which have already been deployed and tested in a few locations [34].

Incorporating bitumen additives into waste recycled coarse aggregates (steel slag) can reduce the overall project cost while also increasing the dynamic stability of the pavements. The steel manufacturing industry produces steel slag, a waste product with high mechanical properties and excellent surface morphology. In addition to its high mechanical properties, steel slag requires proper treatment due to its high-water absorption and volume expansion. Researchers have applied and compared various treatments, including polyvinyl alcohol (PVA), silane coupling agent (SCA), cement paste (CP), epoxy acrylic organic-silicone resin (EAOR), and low-density polyethylene (LDPE), to the surface of steel slag [35,36]. The treatment results showed that treating steel slag made the material much more stable, improved Marshall and flow values, and made it more resistant to freeze–thaw. It also made volume expansion problems a lot less severe [37].

The increasing demand for sustainable construction practices has prompted the investigation of innovative solutions for addressing waste management issues. Steel slag, a by-product of steel production, is plentiful but frequently underused, often ending up in landfills. Similarly, LDPE is commonly utilized in packaging and is considerably contributing to plastic pollution owing to its low recycling rate. The integration of these materials in road construction provides a dual advantage: minimizing waste accumulation and optimizing pavement design life.

To improve pavement performance and durability, many researchers have explored recycled waste materials as an alternative to conventional coarse aggregates, such as rubber from tire waste [38], reclaimed asphalt pavement (RAP) [39], recycled crushed concrete (RCC) [40], and ceramic tile waste (CTW) from construction and demolition operations [41]. Despite the economic and environmental benefits of using these recycled wastes in bitumen pavement, they are also subject to limitations, for example, the use of RCA [42], while the use of rubber waste reduces the mechanical properties due to its elastic nature and thus reduces the stiffness of bitumen, leading to increased deformation under heavy loads [43]. Additionally, despite the widespread application of RAP and its use as a highly expanded aggregate in various paving applications as an alternative to aggregates, it is also subject to many challenges, including processing requirements, significant variation in material quality depending on the source, age, and condition of the reclaimed materials, mix design, workability, compaction, and separation risks. Also, a high RAP content may weaken resistance to rutting under heavy loads if not designed properly [44,45]. Similarly, the use of CTW produced from residential building waste as an alternative to aggregate also has problems, including the tendency of its particles to crack early due to its brittle nature and high porosity, which makes the asphalt mixture susceptible to damage resulting from moisture sensitivity. Its excessive use also weakens its workability during mixing and compaction due to its angular and coarse texture, and it also requires more energy for the crushing and treatment process [46,47]. Compared to these recycled materials, steel slag shows high durability compared to natural aggregate due to its high strength and density, better resistance to slip due to its rough texture, and high stability under harsh weather conditions due to its strong and stable composition, which allows it to be chemically stable, being less susceptible to alkali and carbonate reactions. These advantages make steel slag a promising alternative to natural aggregate in the production of HMA [48,49,50]. This study addresses several critical gaps in the field of asphalt pavement research and introduces a novel, effective, and sustainable approach for transport infrastructure design by combining LDPE as a binder modifier and steel slag as a coarse aggregate replacement. Firstly, while LDPE and steel slag have been studied individually for their use in asphalt mixtures, little attention has been paid to their combined effects on pavement performance. Secondly, prior research lacks comprehensive evaluations of the dynamic modulus over a wide range of temperatures and frequencies, which is critical for understanding material behavior under diverse traffic and climate conditions. Thirdly, this work contributes to sustainability by utilizing industrial waste materials and offering practical solutions for reducing environmental impacts. Lastly, the integration of computational intelligence and advanced modeling techniques, such as multi-expression programming, fills a methodological gap in optimizing material design and performance predictions.

2. Scope and Objectives of the Study

This study seeks to address environmental sustainability by utilizing industrial waste while improving the mechanical and thermal properties of asphalt pavements to mitigate premature failures. The primary aim of this study is to evaluate the combined effect of low-density polyethylene (LDPE) with varying percentages (3%, 5%, and 7%) as a binder modifier in two different grades of bitumen (PEN 60/70 and 80/100) and steel slag (SS) as a coarse aggregate replacement in enhancing the performance of hot-mix asphalt (HMA) to provide practical solutions for durable and sustainable road construction. The LDPE-modified binders were evaluated for their physical properties to determine their performance characteristics. These binders were then used to prepare Marshall samples, both with virgin aggregates and with SS as a coarse aggregate replacement. To assess the effectiveness of these modifications, the dynamic modulus was measured at temperatures of 4.4 °C, 21.1 °C, 37.8 °C, and 54.4 °C and frequencies of 0.1 Hz, 0.5 Hz, 1 Hz, 5 Hz, 10 Hz, and 25 Hz. The dynamic modulus results were analyzed using the NCHRP09-29_Mastersolver2-2 Excel spreadsheet. Furthermore, prediction equations for the modified materials were derived from a Python 3.12-based dataset using multi-expression programming (MEPX), an artificial intelligence (AI) technique. Statistical analyses were conducted using a two-level factorial design in Minitab to investigate the influence of LDPE, SS, and their interactions on asphalt performance.

3. Materials and Methods

3.1. Material Selection

The materials used in this study include fine aggregate, coarse aggregate, and bitumen with two different grades (grades 60–70 and 80–100). Steel slag was obtained from a local factory (see Figure 1) as the by-product produced from impurities during steel making in the blast furnace. Steel slag is taken in its raw form, then crushed into coarse aggregates using a crushing machine, and cleaned with a brush to remove dust and suspended metals, similar to the process applied to coarse aggregate. LDPE was used as a modifier and was obtained from local industry in the form of pellets, as shown in Figure 2. The physicochemical properties of steel slag and LDPE are displayed in

Table 1. Physicochemical properties of steel slag (SS).

Chemical Properties
Property	Value
SiO2 (%)	15.04
Al2O3 (%)	4.12
Fe2O3 (%)	22.55
CaO (%)	41.53
MgO (%)	6.17
K2O (%)	0.05
Na2O (%)	0.14
SO3 (%)	0.08
LOI (%)	0.25
Physical properties
Color	Light to dark brown
Shape and Surface texture	Highly angular and rough
Specific gravity (gm/cm3)	3.25
Bulk density (Kg/m3)	1911.11
Moisture content (%)	0.54

and

Table 2. Physicochemical properties of LDPE.

Property	Value
Color	white
Specific gravity (gm/cm3)	0.921
Tensile strength (MN/m2)	10.11
Flexural modulus (GN/m2)	0.203
Melting point (°C)	113.20
Thermal degradation temperature (°C)	406
Chemical form	(−CH2 − CH2−)n

, respectively.

The physical properties of aggregates/steel slag and bitumen grades 60–70 and 80–100 are given in

Table 3. Test results performed on natural aggregates and steel slag.

Description of Test	Test Standard		Result		Limits
			Natural Aggregates	Steel Slag
Abrasion Value	ASTM C-131 [51]		26	25	45 (max)
Impact Value	BS 812-112 [52]		15	18	30 (max)
Crushing Strength	BS 812-110 [53]		27	26	30 (max)
Specific Gravity	Coarse	ASTM C-127 [54]	2.73	3.25	-
	Fine	ASTM C-128 [55]	2.66	-	-
	Filler	AASHTO T-133 [56]	2.63	-	-
Water Absorbance	Coarse	ASTM C-127 [54]	0.53	0.83	3% (max)
	Fine	ASTM C-128 [55]	0.78	-	3% (max)

and

Table 4. Test results of bitumen physical properties (grade 60–70 and 80–100).

Test Description	Test Standard	Bitumen Grade 60–70		Bitumen Grade 80–100
		Result	Specification	Result	Specification
Penetration, 25 °C, 0.1 mm	ASTM D-5 [57]	65	60/70	96	80/100
Softening Point, °C	ASTM D-36 [58]	47	(46–57) °C	45	(43–54) °C
Specific Gravity, 25 °C, cm	ASTM D-70 [59]	0.99	0.99 (min)	1.01	0.99 (min)
Ductility, g/cm3	ASTM D-113 [60]	>100	100 cm (min)	>100	100 cm (min)

, respectively. All the volumetric properties of control asphalt mix samples are given in

Table 5. Marshall results at OBC for HMA (grade 60–70 and 80–100).

Descriptions	Measured Value (60–70)	Measured Value (80–100)	Standards
Optimum Asphalt Contents (%)	4.0	4.0	4% voids—air
Unit Weight (g/cm3)	2.290	2.275	n/a
Voids in Mineral Aggregate (%)	13.25	13.75	13 (Minimum)
Voids filled with Asphalt (%)	70.5	72	65 to 75
Stability (kN)	11.38	11.07	8.0 (Minimum)
Flow(mm)	3.50	3.00	2.0 to 3.5

. All the results of the tests performed were within the limits of specifications. The aggregate gradation used in preparing HMA samples is shown in

Table 6. NHA specifications for gyratory sample preparations.

Sieve Size	NHA-B Specification Range (% Passing)	Selection (% Passing)
(mm)
19	100	100
12.5	75–90	82.5
9.5	60–80	70
4.75	40–60	50
2.38	20–40	30
1.18	5–15	10
0.075	3–8	5.5

. In this study, bitumen was modified with 3, 5, and 7% LDPE by weight of bitumen. Moreover, the coarse aggregate in HMA was also replaced by 3% steel slag (SS).

3.2. Testing Methods

Initial testing focused on characterizing the aggregates and bitumen to establish baseline properties. This phase also included the calculation of the optimum binder content (OBC) required to achieve 4% air voids. Standardized tests were conducted to determine the physical and mechanical properties of aggregates and bitumen, ensuring compliance with relevant ASTM and BS standards (shown in Table 4).

After basic tests, the consistency and performance of LDPE-modified bitumen for two grades, 60–70 and 80–100, were evaluated. Low-density polyethylene (LDPE) was incorporated into the bitumen at modification levels of 3%, 5%, and 7% by total binder weight. The blending process was carried out at 140–155 °C with a mixing rate of 500 rpm for 30 min to ensure uniform dispersion. The resulting modified bitumen was tested for consistency and other critical performance parameters.

The Marshal Mix design was conducted to determine the volumetric properties of both control and modified samples for the selected bitumen grades. The aggregate gradations and the optimum bitumen content data are presented in Table 5 and Table 6. The modification included the addition of LDPE at 3%, 5%, and 7%, as well as the partial replacement of coarse aggregate with steel slag (particle sizes ranging from 19 mm to 4.75 mm). The impact of these modifications on the volumetric properties was thoroughly evaluated, providing insights into the effects of LDPE and steel slag on the mix design.

Dynamic modulus testing was performed to assess the mechanical behavior of the specimens under varying temperature and frequency conditions. Specimens were cored from gyratory compacted samples with dimensions of 100 mm in diameter and 150 mm in height. The gyratory compaction process, carried out at a mixing temperature of 160–170 °C, involved 125 gyrations to simulate high-traffic conditions equivalent to ESALs ≥ 30 million [61]. The prepared specimens were tested at four temperature levels (4.4 °C, 21.1 °C, 37.8 °C, and 54.4 °C) and six frequency levels (25, 10, 5, 1, 0.5, and 0.1 Hz) [62]. The incorporation of LDPE and steel slag in the specified percentages was examined to determine their influence on the dynamic modulus.

4. Results and Discussion

4.1. Aggregate Testing

In this portion, steel slag and natural aggregates were tested for their physical properties to check and compare their abrasion, crushing strength, impact, specific gravity, and water absorption. The test result in Figure 3 shows that steel slag performs better in abrasion and crushing strength by 1%, while in the case of the impact, its value is reduced up to 3%, and the water absorption of steel slag is slightly more than natural aggregates by 0.3% with a specific gravity of 3.25. All the aggregate testing results were within the limits of standard specifications.

4.2. Bitumen Consistency Testing

This portion includes the test results on modified samples. The LDPE-modified samples were tested for penetration, softening point, ductility, and specific gravity. The bitumen was heated at a temperature of 100 to 100 °C, and then LDPE pellets were mixed in at a rate of 500 rpm for 30 min at 140 to 155 °C. After mixing, the required modified mixture was kept in an oven for 1 h at 110 °C to retain its morphology. LDPE in two bitumen grades (60–70 and 80–100) was taken into the mix at 3%, 5%, and 7% by weight of bitumen. The test results show that adding more LDPE to a sample makes the penetration grade and softening point values better in both bitumen grades. However, as the modification rate goes up, the ductility value goes down, as shown in Figure 4 and Figure 5. The 3% LDPE modifier mixed with bitumen in both grades had a lower ductility value but a higher softening point and penetration value. The softening point went up by 9 °C in grade 60–70 and by 8 °C in grade 80–100. The penetration value went down to 52 in grade 60–70 and to 61 in grade 80–100. The ductility value went down by 3 cm in grade 60–70 and by 4 cm in grade 80–100, as shown in

Table 7. Effect on bitumen consistency by adding LDPE.

Description of Test	Bitumen Grade 60–70				Bitumen Grade 80–100
	0%	3%	5%	7%	0%	3%	5%	7%
Penetration Value	65	52	38	32	96	61	49	33
Softening Point	47	56	62.25	68.4	45	53	57.35	67.1
Ductility Value	104	101	41	31	106	102	47	36
Specific Gravity	0.989	0.968	0.974	0.961	1.011	1.000	0.977	0.969

. The modification resulted in a reduction in the penetration value and an increase in the softening point value, indicating an increase in the stiffness and stability of the bitumen. However, the ductility decreased significantly in grades 5 and 7%, indicating a large surface energy value, which may not be desirable in certain situations, particularly in lower temperatures and slow movement conditions. All the bitumen test results were within the limits, except ductility in cases of 5% and 7% in both grades.

4.3. Marshall Stability and Volumetric Properties

The volumetric properties of two grades of bitumen, 60/70 and 80/100, were determined using an incremental percentage of bitumen starting from 3% to 5% with a 0.5 increment and a coaction rate of 75 blows on each side. An OBC of 4% was achieved after the Marshall Test on a virgin binder for both grades; this OBC was adopted in LDPE- and SS-LDPE modified HMA. The test results and volumetric properties are given in

Table 8. Marshall results of LDPE- and SS-LDPE-modified samples.

Samples No.	Description	Air Voids (%)	VFA (%)	Unit Weight of Mix (g/cc)	Stability (KN)	Flow (mm)	VMA (%)
Bitumen Grade 60–70
1	Control Sample	3.963	70.469	2.280	11.37	3.5	13.420
2	3% LDPE Modifier	3.952	70.970	2.242	13.14	3	13.614
3	5% LDPE Modifier	3.837	71.332	2.213	10.84	3.25	13.385
4	7% LDPE Modifier	3.795	71.732	2.215	11.04	3.25	13.427
5	3% LDPE Modifier with SS	4.143	70.199	2.236	14.96	3.5	13.903
Bitumen Grade 80–100
1	Control Sample	4.040	70.067	2.274	11.08	3	13.496
2	3% LDPE Modifier	3.910	70.693	2.238	12.54	3.5	13.340
3	5% LDPE Modifier	3.430	73.279	2.216	11.12	3.25	12.835
4	7% LDPE Modifier	3.192	74.656	2.207	11.24	3.5	12.593
5	3% LDPE Modifier with SS	4.263	69.349	2.245	14.60	3.5	13.907

. In cases of stability, 3% LDPE and 3% LDPE with SS show the highest values, with up to a 15% increase in the case of 3% LDPE and up to a 32% increase in the case of 3% SS-LDPE in both the bitumen grades, as shown in Figure 6. The flow values for all the modifications were within the limits of ASTM standard specifications in both the bitumen grades, as shown in Figure 7. The voids filled with asphalt (VFA) and voids in mineral aggregates (VMA) for 3% LDPE and 3% LDPE with SS have almost the same value or less deviation from the control sample in both the bitumen grade conditions and within the limits of standard specifications, as shown in Figure 8 and Figure 9. The greater value of stability is the indication of a more stable and compact mix that withstands higher loads with the best performance. The value of the Marshall quotient for 3% SS-LDPE in bitumen grade 60–70 is 436, and in the case of bitumen grade 80–100, this value is 425, which is the highest response among all the modifications being considered and is an indication of stiffer blends. For the performance of HMA pavements, VFA and VMA are of greater use and are indirectly an indication of a good mix with greater durability values and control of bleeding in the HMA. The control sample was selected according to optimum binder content, and the 3% LDPE and 3% SS-LDPE values of VFA and VMA closer to it show a durable mix.

4.4. Dynamic Modulus Using SPT

A dynamic modulus test was performed on the gyratory compacted specimens that were cored to 100 mm diameter and trimmed to 150 mm height under four temperature periods, i.e., 4.4 °C, 21.1 °C, 37.8 °C, and 54.4 °C, and under six frequency levels, i.e., 25 Hz, 10 Hz, 5 Hz, 1 Hz, 0.5 Hz, and 0.1 Hz. A total of thirty samples of modified HMA specimens were tested, including three samples each of LDPE-modified grades 60/70 and 80/100 bitumen with LDPE percentages of 0%, 3%, 5%, and 7%. Furthermore, 3% LDPE-modified bitumen grades 60/70 and 80/100 were used for SS-modified HMA samples. The dynamic modulus values for samples modified with LDPE are shown graphically in Figure 10 and Figure 11 at four different test temperatures and in Figure 11 and Figure 12 at six different load frequencies for grade 60–70 and 80–100 modified samples.

The SPT dynamic modulus (DM) (|E*|) of test samples is directly proportional to frequency and inversely proportional to test temperature, i.e., the dynamic modulus (DM) increases with increasing frequency and decreases with increasing temperature. Based on the dynamic modulus (DM) results, it is hypothesized that it is temperature-sensitive; in colder places, HMA pavements perform better with more stiffness as opposed to regions with higher temperatures, where there is more flexibility in HMA. The dynamic modulus increases with the addition of LDPE and SS, as shown in Figure 10 and Figure 11 for isochronal and isothermal curves, respectively, in grade 60–70 samples. HMA grade 60–70 modified samples perform better in all cases as compared to the control sample, and samples containing 7% LDPE and 3% SS-LDPE show the highest dynamic modulus at both lower and higher temperatures. Figure 12 and Figure 13 for isothermal and isochronal curves, respectively, demonstrate how the dynamic modulus increases with the addition of LDPE and SS in grade 80–100 samples.

In comparison to previous LDPE-modified samples, HMA grade 80–100 modified samples with 7% LDPE and 3% SS-LDPE exhibited the highest dynamic modulus in both lower and higher temperature situations; however, a slight increase in the dynamic modulus values of grade 80–100 control samples at 4 °C was observed as compared to grade 60–70 control samples, which showed less impact than modified grade 80–100 samples at lower temperatures. This is due to the higher penetration value of grade 80–100 samples as they perform better in low-temperature conditions. LDPE increases the stiffness of bitumen, and the SS improves the stability as steel slag is denser than Babuzai slag. The best response from the isothermal and isochronal curves was again taken at 3% SS-LDPE because the 7% LDPE in both the bitumen grades shows an abnormal result, with the highest dynamic modulus values being in higher temperature conditions, the average response being in low-temperature conditions, and the lowest response at being low frequency levels, which can further be seen from the master curves in Figure 14 for grade 60–70 and in Figure 15 for grade 80–100 HMA.

4.4.1. Master Curve Development from Dynamic Modulus Dataset

The temperature and time superposition rule at 21.1 °C was adopted for the dynamic modulus (DM) dataset to develop a master curve. These datasets have been obtained from different combinations of LDPE modification results that help in the design process of asphaltic pavement. While testing at −10 °C was not practicable with the AMPT apparatus in Pakistani environmental conditions, the master curve was designed from the procedure of NCHRP-Proj: 09-36. The NCHRP data-solver MS Excel add-in tool depicts the best-fit line which restricts SSE for creating master curves. Figure 14 and Figure 15 show master curves, which display the (DM) dataset responses with frequencies at levels of high–low for LDPE-modified HMA grades 60/70 and 80/100, respectively, where, accordingly, low and high frequencies correspond to high and low temperatures.

Master curves show that LDPE-modified mix samples have greater stiffness values at both low and high temperatures as compared to regular mix samples because of the mix’s similar performance at low temperatures. The 7% LDPE- and 3% SS-LDPE modified mixtures have the best results at high and low temperatures, according to Figure 14 and Figure 15 of the master curve. Also, a high frequency, like 25 Hz, depicts a high-speed movement of traffic, whereas a low frequency, like 0.1 Hz, shows a low-speed movement of traffic. Based on the master curve frequency distribution, at higher frequencies, LDPE-modified samples perform better, but when the frequency gradually reduces to 0.1 Hz and as the LDPE content increases from 3% to 7%, its dynamic modulus shows a weaker response. The findings corroborate the idea that LDPE-modified samples of 7% LDPE and 3% LDPE-SS have optimum (DM) values for high-speed cars at low as well as high temperatures, whereas 3% LDPE-modified samples have ideal (DM) values for low-speed cars at higher temperature ranges. The shift factor, reduced frequency, and inputs for the ME Pavement design guide are all included in the master solver worksheet so that each gradation can be used as a guide. The master curve findings demonstrate an increase in stiffness at both high and low temperatures for the LDPE- and SS-LDPE-modified HMA mixtures, particularly for the optimal 3% LDPE and 3% SS-LDPE combinations. This increase in stiffness indicates improved resistance to deformation under repetitive loading, which directly correlates to better fatigue resistance. Stiffer mixtures reduce the strain energy absorbed under loading cycles, thus slowing crack initiation and propagation in the pavement. R² of LDPE-modified and SS-LDPE-modified material for two bitumen grades is mentioned in

Table 9. R2 of LDPE- and SS-LDPE-modified HMA.

Description of a Modifier	R2	Se/Sy
Bitumen Grade 60/70
0% LDPE	0.9522	0.15
3% LDPE	0.9209	0.20
5% LDPE	0.9540	0.15
7% LDPE	0.9802	0.10
3% LDPE with SS	0.9538	0.15
Bitumen Grade 80/100
0% LDPE	0.9899	0.07
3% LDPE	0.9190	0.20
5% LDPE	0.7460	0.36
7% LDPE	0.8490	0.27
3% LDPE with SS	0.9028	0.22

.

4.4.2. Experimental Two-Level Factorial Design for SPT Dynamic Modulus

Experimentation probabilistic design using Minitab 21.1 software was performed on the data obtained by conducting dynamic modulus testing on LDPE-modified grade 60–70 and 80–100 bitumen. In this study, a two-level factorial mathematical design is selected, which means high and low levels of modifier, temperature, and frequency were taken as input variables and the output or response variable was dynamic modulus, as shown in

Table 10. Two-level factors for factorial design (SS-LDPE-modified HMA).

Factors	Measured Units	Low-Level Factors		High-Level Factors
		Grade 60/70	Grade 80/100	Grade 60/70	Grade 80/100
Temperature	Centigrade	4.4	4.4	54.4	54.4
SS-LDPE	%	0	0	3	3
Frequency	Hertz	0.1	0.1	25.0	25.0

.

Table 11. Effect estimates for dynamic modulus—grade 60–70 SS-LDPE-modified HMA.

	Main			Interactions
	A: Temp	B: Freq	C: LDPE	AB	BC	AC	ABC
Avg. High	1077	5355.75	3919	2059.25	3670.25	2832	3048.25
Avg. Low	5387.5	1108.75	2545.5	4405.25	2794.25	3632.5	3416.25
Effect	−4310.5	4247	1373.5	−2346	876	−800.5	−368
Coefficients	−2155.25	2123.5	686.75	−1173	438	−400.25	−184
p-Value	0.000	0.000	0.000	0.000	0.000	0.000	0.003

and

Table 12. Effect estimates for dynamic modulus—grade 80–100 SS-LDPE-modified HMA.

	Main			Interactions
	A: Temp	B: Freq	C: LDPE	AB	BC	AC	ABC
Avg. High	1052.75	6608.5	4137	2089	3916.25	3768.5	3950.75
Avg. Low	6722.5	1166.75	3638.25	5686.25	3859	4006.75	3824.5
Effect	−5669.75	5441.75	498.75	−3597.25	57.25	−238.25	126.25
Coefficients	−2834.87	2720.875	249.375	−1798.62	28.625	−119.12	63.125
p-Value	0.000	0.000	0.000	0.000	0.040	0.593	0.250

show the main and interaction effects of SS-LDPE-modified grades 60–70 and 80–100 HMA, respectively. The main effect is the individual average high and average low responses of factors within themselves, while the interaction effect is the average high and low response of each factor to another. Temperature is represented by A, frequency by B, and modifier by C, as indicated in the tables.

In Table 11 and Table 12, the main and interaction effects of frequency, modifier, and temperature on the dynamic modulus (DM) of SS-LDPE-modified grades 60–70 and 80–100 hot-mix asphalt are presented with R2 of 99.67% and 99.79%, respectively. Positive and negative signs of the effect indicate the nature of the relationship; a positive sign shows a direct relationship, while a negative sign is an indication of an inverse relationship. From this observation, it can be concluded that the dynamic modulus value rises with the rise in frequency and modifier and falls with the temperature rise. The design of this experiment was performed with a 95% confidence level. The p-value demonstrates the significance of each factor in this research work, apart from AC (two-way interaction) and ABC (three-way interaction), of the grade 80–100 SS-LDPE modifier. Therefore, the design of the experiment (DOE) excluded these factors for further processing.

Analysis of Variance (ANOVA) results are displayed in

Table 13. ANOVA for dynamic modulus—grade 60–70 SS-LDPE-modified HMA.

Source Description	DF	Adjusted SS	Adj MS	F-Value	p-Value
Given Model	9	274,123,233	30,458,137	472.4	0
Linear	3	231,009,162	77,003,054	1194.29	0
Temperature	1	111,473,841	111,473,841	1728.92	0
Frequency	1	108,213,560	108,213,560	1678.35	0
SS-LDPE	1	11,321,761	11,321,761	175.6	0
2-Way Interactions	3	41,458,467	13,819,489	214.34	0
Temperature*Frequency	1	33,012,913	33,012,913	512.02	0
Temperature*SS-LDPE	1	3,844,801	3,844,801	59.63	0
Frequency*SS-LDPE	1	4,600,753	4,600,753	71.36	0
3-Way Interactions	1	813,280	813,280	12.61	0.003
Temperature*Frequency*SS-LDPE	1	813,280	813,280	12.61	0.003
Error	14	902,664	64,476
Total	23	275,025,896

for grade 60–70 SS-LDPE-modified HMA and in

Table 14. ANOVA for dynamic modulus—grade 80–100 SS-LDPE-modified HMA.

Source Description	DF	Adjusted-SS	Adj MS	F-Value	p-Value
Given Model	7	450,194,120	64,313,446	984.24	0
Linear	3	372,044,758	124,014,919	1897.9	0
Temperature	1	192,876,390	192,876,390	2951.75	0
Frequency	1	177,675,858	177,675,858	2719.12	0
SS-LDPE	1	1,492,509	1,492,509	22.84	0
2-Way Interactions	2	77,982,777	38,991,389	596.72	0
Temperature*Frequency	1	77,641,245	77,641,245	1188.21	0
Temperature*SS-LDPE	1	341,532	341,532	5.23	0.036
Error	16	1,045,490	65,343
Total	23	451,239,610

for grade 80–100 SS-LDPE-modified HMA.

The p-values in the ANOVA tables indicate statistical significance, with values < 0.05 confirming that the effects of temperature, frequency, and SS-LDPE, as well as their significant interactions, are not due to random variation. The F-values quantify the relative influence of these factors, with higher values indicating stronger effects. For grade 60–70 HMA (Table 13), all main factors and interactions, including the three-way interaction, are significant, as shown by their p-values and high F-values. For grade 80–100 HMA (Table 14), the significant main factors and two-way interactions, particularly temperature and frequency, dominate the response, while the interaction term temperature*SS-LDPE is relatively less impactful but still significant (p = 0.036). These results confirm the robustness of the experimental design and validate the influence of the studied factors on the dynamic modulus. From Table 13, the degree of freedom (DOF) for the grade 60–70 model is 9:3 DOF for one-way interaction, 3 DOF for two-way interaction, and 1 DOF for three-way interaction. In this case, all the parameters are significant in the sense of F-value and p-value. From Table 14, the DOF for the 80–100 model is 7:3 DOF for one-way interaction, 2 DOF for two-way interaction, and no DOF for three-way interaction, as it was neglected previously from the design after the interaction effect’s insignificance.

The normal plot of standardized effects for grade 60–70 SS-LDPE-modified HMA is shown in Figure 16, and that for grade 80–100 SS-LDPE-modified HMA is shown in Figure 17. In both cases, the significance level was taken as 0.05 with three parameters, namely temperature, frequency, and modifier, along with main and interaction effects. All the parameters in the grade 60–70 and grade 80–100 models shown in red are significant, while the insignificant ones were removed from the grade 80–100 model after a higher p-value of the interaction effect. From the normalized graphs, it can be concluded that the bitumen modifier has a significant effect on the design model.

The main effect plots were based on the high and low levels of each parameter, namely temperature, frequency, and modifier, as shown in Figure 18 for |E*| DM of grade 60–70 SS-LDPE-modified HMA and in Figure 19 for DM of grade 80–100 SS-LDPE-modified HMA. The steepness of the lines drawn shows the correlation of the response variable with the considered parameters. From the main effect plots, the SS-LDPE modifier is responsible for bringing about a positive change in HMA behavior in both cases. The SS-LDPE modifier has a strong impact on HMA in grade 60–70 as compared to grade 80–100.

The interaction effect plots of each parameter with another parameter in the form of a two-way interaction and a three-way interaction were computed for their high and low levels for each factor, as shown in Figure 20 and Figure 21 for the dynamic modulus of grade 60–70 and 80–100 SS-LDPE-modified HMA, respectively. In interaction plots, parallel lines indicate the insignificance of the considered factors, while non-parallel lines show the interaction between those factors. It also has a considerable effect because of the steepness of the lines. From the plots, we can observe some non-parallel lines in grade 60–70 SS-LDPE-modified HMA; it can be inferred that there is an interaction effect between the variables with the response dynamic modulus, while in grade 80–100 SS-LDPE-modified HMA, the BC plot has parallel lines, which means there is no interaction effect between them.

Pareto charts for the dynamic modulus of grade 60–70 SS-LDPE-modified HMA are shown in Figure 22, and those for the dynamic modulus of grade 80–100 SS-LDPE-modified HMA are shown in Figure 23. Pareto charts show the mean response of the interaction and the importance of the components included in the design, together with the influence of relative importance. A level of significance of 0.05 was taken, which is indicated by the red dotted line in the charts. In the grade 60–70 SS-LDPE-modified HMA, all the factors in either a one-way interaction, two-way interaction, or three-way interaction are significant, while in the grade 80–100 SS-LDPE-modified sample, two-way interaction BC and three-way interaction ABC were non-significant, so they were removed from the model observations.

The cube plots for the dynamic modulus of grade 60–70 SS-LDPE-modified HMA are shown in Figure 24, and those for grade 80–100 SS-LDPE-modified HMA are shown in Figure 25. Cube plots clearly show that the highest response dynamic modulus values may be obtained at the lowest temperature and highest frequency and modifier when employed. The dynamic modulus values rise at lower temperatures because the material stiffens at that temperature, and they fall as the temperature rises because the bitumen softens with high strains, which has a detrimental effect on the dynamic modulus. On the other hand, lower frequency levels, which are indications of slow-moving vehicles, have negative impacts on the response factor. Hence, it is evident from the cube plot that the parameters, namely temperature, frequency, and modifier, are all significant factors affecting the response dynamic modulus.

The contour plots for the dynamic modulus of grade 60–70 SS-LDPE-modified HMA in terms of low levels of modifier are shown in Figure 26a, and those for high levels are shown in Figure 26b; similarly, the dynamic modulus of grade 80–100 SS-LDPE-modified HMA in terms of low levels of modifier is shown in Figure 26c, and that for high levels is shown in Figure 26d. The dynamic modulus has an inverse relationship with test temperature and a direct relationship with frequency and modifier. Response coefficients can be visualized as a topographical map since they are represented by contour plot lines. These line curves show how temperature and frequency interact at high and low levels. By drawing a line perpendicular to the vertical axis at each temperature point and maintaining a constant frequency, it is possible to determine the dynamic response of the model. At any frequency, keeping the temperature constant can also give us the response dynamic modulus; the dynamic modulus was anticipated at every frequency level.

Figure 27a displays the surface plot for the dynamic modulus of grade 60–70 SS-LDPE-modified HMA at a low level of modifier, while Figure 27b displays it at a high level. Similarly, Figure 27c displays the surface plot for the dynamic modulus of grade 80–100 SS-LDPE-modified HMA at a low level of modifier, while Figure 27d displays it at a high level. The surface plot provides a 3D view of the design, combining the response variable, dynamic modulus, with other factors like temperature and frequency. From the plots, at lower temperature and higher frequency levels, the response gives us the highest dynamic modulus, and at higher temperature and lower frequency levels, the response gives us the lowest dynamic modulus. These plots are used to obtain any response value by choosing random values of the given components by simply drawing a line straight from frequency and temperature to the axis of the |E*| (DM), which are not specified in the dataset of lab-evaluated test values.

4.5. Evolutionary Algorithm for Dynamic Modulus Using CI

Multi-expression programming was adopted in this study to examine genetic programming using a Python dataset to encode multiple parameters into the same chromosome in computational language, leading to an evolutionary computation algorithm. Performance is a measure of how the pavement responds to several factors, such as loading rate, pavement temperature, atmospheric condition, etc. Pavements need to perform as expected during the design phase to guarantee their long-term viability. The test temperature, loading frequency, and modifier content (such as low-density polyethylene and steel slag) all influence the dynamic modulus, a performance measure.

Dynamic Modulus = f (temperature, frequency, modifier)

(1)

The data from Equation (1) were utilized to create a model that can estimate the material’s dynamic modulus for values other than the test dataset. Based on the trial-and-error process, different models were produced, namely regression models using Minitab and MS Excel, Cobb–Douglas models, Witczak models, and Hirsch models. However, there was a drawback in each model, as it overpredicted the dynamic modulus values when model validation was performed on each. As a result, multi-expression programming was utilized, using computational regression intelligence models that best represented the data, with a training best error below 100 for accuracy and targeted output, as shown in Figure 28a,b for grade 60–70 LDPE and 80–100 LDPE-modified HMA, respectively.

Grade 60–70 bitumen has a relatively lower viscosity compared to grade 80–100, which may have led to inconsistent interaction between the LDPE modifier and the bitumen, especially at lower concentrations. This could have resulted in variations in the mixture’s rheological behavior and contributed to the observed error. Moreover, LDPE’s modifying effect is dependent on its interaction with the base bitumen. For grade 60–70 bitumen, which has a different chemical composition compared to grade 80–100, the modification may not have been as effective in improving the performance characteristics, leading to higher deviation from experimental data. Furthermore, the computational intelligence models may have shown increased sensitivity to changes in the properties of LDPE-modified HMA at the lower bitumen grade, as the interaction between the variables could have been more complex or nonlinear for grade 60–70.

For generating the model using MEP, different parameters were set, such as the number of subpopulations (10), subpopulation size (1000), code length (30), and number of generations (2000). The obtained result was encoded in Python, and it was decoded to obtain the models for grade 60–70 with a training error of 72.44 and those for grade 80–100 with a training error of 68.18.

The final version of the grade 60–70 LDPE-modified HMA model can be written as Equation (2):

$$\left|E^{*}\right|=2\times \left[132\sqrt{2F}+35M-16\sqrt{2TF}-4M\sqrt{T} \right]$$

(2)

The final version of the grade 80–100 LDPE-modified HMA model can be written as Equation (3):

$$\left|E^{*}\right|=\left|24F+12M-11T-{\displaystyle \frac{M}{F}}+258\right|-\left[2F^2+FM-62F-31M+11T+2{\displaystyle \frac{M}{F}}-246\right]$$

(3)

where |E*| = Projected Absolute Dynamic Modulus (ksi); M = Added Modifier (LDPE), % (0 to 7); F = loading frequency of SPT, Hz (0.1 to 25); and T = test temperature, °C (4.4 to 54.4).

Multi-Expression Programming Model Validation

Model validation was performed on LDPE-modified HMA models produced using the MEP computing algorithm. In our study, the mean absolute percentage error (MAPE) technique was adopted for checking model validation. MAPE is the difference between the lab-tested (actual) dynamic modulus and the calculated or projected (predicted) dynamic modules, divided mathematically by the actual dynamic modulus, then extracted as absolute values of all the differences, taking the mean of the overall sum, and finally converted into a percentage. Equation (4) for yielding the mean absolute percentage error is given below.

$$MAPE={\displaystyle \frac{100\%}{n}}{\sum }_{i=1}^n\left|{\displaystyle \frac{A_t-P_t}{A_t}}\right|$$

(4)

where A_t = Actual Dynamic Modulus and P_t = Predicted Dynamic Modulus.

The MAPE values for the grade 60–70 LDPE-modified HMA and the grade 80–100 LDPE-modified HMA models are 23.42% and 5.15%, respectively. The MEP models are valid through MAPE, but in the case of grade 60–70, the model shows some weak response but is still acceptable; in the case of grade 80–100, the model represents a strong relationship with the data. The model’s prediction shows a weak response when there is temperature involved in the affecting factors, which is true for this case study as well.

Table 15. Interpretation of MAPE score.

MAPE Score (%)	Model Accuracy (%)	Interpretation of Score
<10	>90	Excellent Model
10–20	80–90	Good Model
20–50	50–80	Relatively Good Model
>50	<50	Poor Model

shows the MAPE score and accuracy of the model.

Figure 29 shows a grade 60–70 LDPE-modified HMA validation plot, and Figure 30 demonstrates a grade 80–100 LDPE-modified HMA validation plot. The plots with 45-degree lines show the validation of the dataset; the closer the dataset points to the 45-degree line, the more accurately the model represents the data, so Figure 29 and Figure 30 indicate that the models are reliable and valid. However, the developed MEP model for the grade 60–70 LDPE-modified HMA shows a slight deviation from its 45-degree line, with dotted lines denoting the trend line below the solid 45-degree line.

4.6. KENPAVE Damage Analysis

An analysis of hot-mix asphalt (HMA) pavements using the mechanistic–empirical method with KENPAVE 1.0 software was conducted in this study. The primary objective was to evaluate the effect of various modifiers on the design life of flexible pavements. Dynamic modulus values were incorporated, accounting for seasonal variations in modulus at a constant frequency.

Two bitumen grades, 60–70 and 80–100, were analyzed with modifiers, including LDPE at 3%, 5%, and 7%, and 3%-SS for each grade. Dynamic modulus values for these grades, along with the modifiers, were evaluated across three temperature ranges (4.4 °C, 21.1 °C, and 54.4 °C) at a constant frequency of 5 Hz during KENLAYER analysis. The loading configuration included single-axle dual-tire trucks with an allowable load repetition of 50,000 cycles, a contact pressure of 100 psi, a dual spacing of 13 inches, and a contact radius of 4.2 inches.

For the damage analysis, two critical points were examined, i.e., at the bottom of the HMA layer and at the top of the subgrade.

Table 16. KENPAVE analysis of HMA samples.

Period of Year	Description	Percent of Modifier in Bitumen Grade 60–70					Percent of Modifier in Bitumen Grade 80–100
		0	3	5	7	3-SS	0	3	5	7	3-SS
Number 1 (Temp: 4.4 °C; Freq: 5 Hz)	Dynamic Modulus (ksi)	729	1023	1027	1157	1113	1090	993	996	1105	1257
	Allowable Load at Asphalt Bottom (Nf)	1,498,000	2,154,000	2,163,000	2,487,000	2,375,000	2,318,000	2,082,000	2,089,000	2,355,000	2,750,000
	Damage Ratio	0.03337	0.02322	0.02311	0.0201	0.02105	0.02157	0.02402	0.02393	0.02123	0.01818
	Allowable Load Repetition at Subgrade Top (Nd)	645,200	1,119,000	1,127,000	1,387,000	1,296,000	1,249,000	1,064,000	1,069,000	1,279,000	1,611,000
	Damage Ratio	0.07749	0.04467	0.04438	0.03604	0.03859	0.04003	0.047	0.04676	0.03908	0.03104
Number 2 (Temp: 21.1 °C; Freq: 5 Hz)	Dynamic Modulus (ksi)	417	494	504	651	598	574	575	543	714	601
	Allowable Load at Asphalt Bottom (Nf)	931,000	1,057,000	1,074,000	1,343,000	1,242,000	1,198,000	1,200,000	1,142,000	1,468,000	1,249,000
	Damage Ratio	0.0537	0.04731	0.04655	0.03723	0.04024	0.04173	0.04167	0.04377	0.03406	0.04003
	Allowable Load Repetition at Subgrade Top (Nd)	295,300	368,400	378,500	543,900	480,500	453,200	454,300	419,200	625,000	484,600
	Damage Ratio	0.1693	0.1357	0.1321	0.09193	0.1041	0.1103	0.1101	0.1193	0.08	0.1032
Number 3 (Temp: 54.4 °C; Freq: 5 Hz)	Dynamic Modulus (ksi)	105	123	146	227	184	107	151	169	267	175
	Allowable Load at Asphalt Bottom (Nf)	593,200	583,300	586,100	656,700	611,400	591,400	588,200	599,200	708,000	603,600
	Damage Ratio	0.08429	0.08572	0.08531	0.07613	0.08177	0.08455	0.085	0.08344	0.07062	0.08284
	Allowable Load Repetition at Subgrade Top (Nd)	74,330	84,560	97,870	148,200	120,700	75,460	100,800	111,600	175,500	115,000
	Damage Ratio	0.6727	0.5913	0.5109	0.3374	0.4142	0.6626	0.496	0.4482	0.2848	0.4347
Summary	Sum of Damage Ratio (Fatigue)	0.17136	0.15625	0.15497	0.13346	0.14306	0.14785	0.15069	0.15114	0.12591	0.14105
	Sum of Damage Ratio (Deformation)	0.91949	0.77167	0.68738	0.46537	0.55689	0.81293	0.6531	0.61426	0.40388	0.56894
	Summative Damage (F.C + P.D)	1.09085	0.92792	0.84235	0.59883	0.69995	0.96078	0.80379	0.7654	0.52979	0.70999
	Percentage Reduction in Damage Ratio	0	14.94	22.78	45.10	35.83	0	16.34	20.34	44.86	26.10
	Design Life (Years)	1.09	1.3	1.45	2.15	1.8	1.23	1.53	1.63	2.48	1.76
	Percentage Improvement	0	19.27	33.03	97.25	65.14	0	24.39	32.52	101.63	43.09

presents the results of the damage analysis, design life, and percentage improvement for bitumen grades 60–70 and 80–100 due to LDPE and SS modifications.

4.6.1. Reduction in Damage Ratios

The allowable load repetition at the bottom of the HMA (Nf) and the top of the subgrade (Nd) for each season at a single vehicle speed was used to compute the damage ratios. The damage ratios of modifiers with control samples were taken for both grades 60–70 and 80–100 HMA. The result in Table 16 shows that in the case of all modifiers, damage ratios were significantly reduced as compared to the control sample.

4.6.2. Percentage Improvement in Design Life

The design life of pavement in years was calculated from the highest damage ratio among the amounts of fatigue cracking at the bottom of the HMA and the permanent deformation at the top of the subgrade. From Table 16, the design life of a pavement can be significantly improved by adding LDPE and SS to the HMA.

5. Conclusions

This study evaluates the combined effects of low-density polyethylene (LDPE) at varying percentages (3%, 5%, and 7%) as a binder modifier and steel slag (SS) as a coarse aggregate replacement to enhance hot-mix asphalt (HMA) performance for durable and sustainable road construction. The following conclusions are drawn:

• LDPE and steel slag (SS) modifications significantly improved the performance of hot-mix asphalt (HMA) for both bitumen grades (60–70 and 80–100), with notable upgrades in physical and mechanical properties and Performance Grade (PG).
• SS exhibited superior abrasion resistance and crushing strength compared to conventional aggregates.
• LDPE-modified bitumen improved penetration and softening point values, enhancing high-temperature performance and upgrading the PG.
• The 3% LDPE yielded optimal results by balancing stiffness, stability, and ductility.
• Specimens with 3% LDPE showed improvements in Marshall Stability by 15.57% for grade 60–70 and 13.18% for grade 80–100, with minimal reductions in air voids and unit weight.
• Dynamic modulus values were higher across all frequencies and temperatures, reducing fatigue cracking and permanent deformation by 14.94% to 16.34%.
• Combining SS with 3% LDPE improved Marshall Stability by over 31%, reduced damage ratio by up to 35.83%, and significantly enhanced the design life of HMA.
• The design life of grade 60–70 HMA increased by 65.14%, while grade 80–100 saw a 43.09% improvement.
• Damage ratio reductions were observed as 35.83% for grade 60–70 and 26.10% for grade 80–100.
• Dynamic modulus improvements were observed across all temperature ranges, especially at extremely low and high temperatures.
• These findings confirm the potential of SS and LDPE modifications for durable and sustainable flexible pavement solutions.

6. Recommendations

This study recommends employing 3% LDPE as a bitumen modifier in both 60–70 and 80–100 grades. This target percentage of the LDPE modifier improved overall performance while slightly lowering some parameters compared to the control sample. Similarly, the same LDPE modifier percentage has good stability values, with the best dynamic modulus at a lower frequency and higher temperature ranges and ultimately increases the damage resistance of asphaltic pavements according to results from the KENPAVE software. Ultimately, the adoption of LDPE as a binder modifier and steel slag (SS) as a substitute for coarse aggregates in hot-mix asphalt enhanced its performance across all frequency levels and temperature ranges, particularly at the highest and lowest temperatures. This study strongly suggests using this approach for roads in Pakistan with higher temperature zones, aiming to harness the steel industry’s waste, reduce plastic waste accumulation, and prolong the pavement’s design life.

7. Limitations

The limitations of utilizing recycled steel slag (SS) and LDPE for pavement construction can be summarized in the following points:

• Waste Material Compatibility: The physical and chemical properties of LDPE and SS vary depending on the origin and nature of the materials and may be subject to volume expansion resulting from moisture and carbonation, causing volume instability that affects their bonding with the bitumen binder and may therefore cause cracks or deformation in the pavement.
• Cost Implications: Recycling materials may appear cost-effective, but the expenses of processing, shipping, transportation, and quality control might exceed the advantages in certain circumstances.
• Environmental Issues: SS may contain trace heavy metals, presenting possible leaching hazards. LDPE, being a plastic substance, may deteriorate into microplastics under certain circumstances, posing environmental concerns.
• Lab-Scale Testing: The findings are based on laboratory tests, and field validation under real-world conditions is needed for broader adoption.

8. Future Scope of Work

Future research on the integration of recycled steel slag (SS) and low-density polyethylene (LDPE) in hot-mix asphalt (HMA) for sustainable pavement solutions should focus on optimizing the blend composition for enhanced mechanical performance and long-term durability. Advanced experimental studies can investigate the microstructural interactions between SS and LDPE, aiming to improve stability and bonding. Life cycle assessment (LCA) and cost–benefit analysis should be conducted to establish the economic and environmental feasibility of scaling up this approach for widespread implementation. Furthermore, field-scale trials and performance monitoring under varying traffic loads and climatic conditions would validate their practical applicability. Exploring the use of other industrial by-products alongside SS and LDPE in HMA could further enhance sustainability and performance metrics. The following points are suggested for further studies:

• More research and cooperation with bitumen plants and road agencies are required to conserve industrial resources by replacing aggregates with different waste materials to achieve sustainability and lower costs.
• Further testing on HMA mixes, such as four-point beam fatigue, semi-circular bending, fuel leakage, Hamburg wheel truck, skid resistance, abrasion, and other features, should be carried out.
• Computational intelligence based on advanced computational methods and predictive models should be used to optimize pavement design further.