Dynamic compaction (DC) is a widely used ground-improvement technique due to its cost-effectiveness, low environmental impact, and high adaptability. Despite its simple implementation, compaction efficiency is governed by multiple interacting factors, including tamping energy and soil properties, which poses challenges to practical design. Although numerous investigations have been reported, a comprehensive review systematically linking the various aspects of the DC technique through multiple approaches remains lacking. This paper addresses this gap by integrating and critically evaluating findings from field studies, controlled laboratory experiments, analytical studies, and numerical modeling to establish an effective framework for dynamic compaction applications. In addition, the environmental performance of DC is critically assessed, demonstrating its relatively low environmental footprint compared to material-intensive ground-improvement techniques, as impacts are primarily governed by construction energy rather than material production, although vibration and noise remain key considerations. The findings indicate that DC performance is controlled by the combined effects of the tamper mass, drop height, and geometry, together with impact spacing, number of blows, and initial soil properties. Field studies show that densification depth and uniformity are influenced by the fines percentage, drainage conditions, and applied energy levels, often requiring appropriate tamping strategies to mitigate pore water effects. Laboratory investigations highlight the dominant role of tamper mass over drop height in stress transmission and penetration depth and demonstrate how the tamper shape and impact sequence govern crater formation and strain localization. Numerical models employing finite element, discrete element, smoothed particle hydrodynamics, and hybrid approaches provide insight into stress wave propagation, pore pressure evolution, and soil–structure interaction. However, limitations remain in simulating sequential tamping, boundary conditions, and coupled hydro-mechanical behavior. This review highlights the need for cross-validated modeling, advanced instrumentation, and machine learning integration to support predictive, site-responsive dynamic compaction design in complex geotechnical settings. 2. Research Methodology This review paper adopts a structured, integrated methodological approach, systematically synthesizing data from four main research methodologies: field studies, laboratory experiments, numerical modeling, and analytical studies. The goal is to consolidate scattered research insights and deliver a cohesive, comprehensive perspective on dynamic compaction mechanisms and practices. For the sake of brevity, past studies are not described individually in this section; however, details are provided in the Appendix of this paper. 2.1. Field Studies The reviewed field studies on dynamic compaction (DC), as summarized in Appendix A, followed a structured methodology involving site characterization, energy application using heavy tampers, and in situ testing to assess ground improvement. Most studies began with general subsurface descriptions based on field logs, identifying materials such as sandy silt, silty sand, clayey silt, loess, gravel, red fill, or soil–rock mixtures. While full laboratory classification was not consistently performed, the selected studies only reported basic index properties such as the natural moisture content, dry density, and void ratio. Other geotechnical parameters like Atterberg limits or gradation coefficients (C u, C c) were only occasionally included, resulting in an incomplete assessment. In the reported studies, DC was implemented using tampers weighing 4.5–26 tonnes, dropped from heights of 6–25 m, producing energy levels ranging from approximately 1 to 15 MN·m per drop. Impact points received 5–20 drops, typically across two to four passes, including primary, secondary, and ironing phases. Tamping grids ranged from 2 to 10 m in spacing, with some studies varying energy and spacing zone-wise. Tamper shapes were not always reported, but where noted, cylindrical or flat-bottomed designs were common. Post-compaction evaluation relied heavily on in situ testing. Standard Penetration Tests (SPTs) and Cone Penetration Tests (CPTs) were most frequently used, supported by Dynamic Penetration Tests (DPTs), Super Heavy DPT, and Plate Load Tests (PLTs). Surface wave methods such as Spectral Analysis of Surface Waves (SASW), Surface Wave Testing (SWT), and Rayleigh wave velocity tests were applied in several cases to monitor changes in stiffness. Ground response was assessed through crater depth measurements, surface settlement tracking, and, in selected studies, the use of vibration sensors or pore pressure transducers. Despite methodological variations, the studies reported in Appendix A reflect a consistent and field-appropriate approach to evaluating the performance of DC across diverse ground conditions. 2.2. Laboratory Studies The laboratory studies on dynamic compaction (DC), summarized in Appendix B, used controlled model-scale testing to examine the influence of tamper mass, geometry, drop height, energy, and soil type on densification behavior. Soils included red clay, silty clay, decomposed granite, clean and silty sands, sand–silt mixtures, and well-graded sands. Reported properties included the Atterberg limits, particle size distribution, specific gravity, maximum and minimum dry density, moisture content, relative density, and in some cases, cohesion and friction angle. Soil classification (e.g., SP, SW) was provided where applicable. For convenience, the size of tampers and soil domains were scaled down in most laboratory investigations. Tamper masses ranged from 0.2 kg to 122.6 kg, with drop heights of 100 mm to 3.42 m. Tamper bases varied in shape—flat, conical, convex, shell, and cylindrical—and diameters ranged from 30 to 100 mm. Drop counts ranged from 6 to 30 per point. Some studies also examined drop spacing (e.g., 3D T to 6D T) and drop sequences (consecutive vs. alternate). Compaction energy was applied via free-fall systems, electromagnetic release, or spring-actuated drop mechanisms, often with calibrated energy input. A wide range of laboratory boxes and containers were used. Small-scale Perspex-sided chambers (e.g., 350 × 150 × 750 mm) were designed for high-speed imaging, while larger steel or wooden boxes (up to 1000 × 600 × 600 mm) accommodated layered fill and crater development. Some studies used transparent sidewalls or colored paper layers to track internal soil movement. A 2 × 2 m surface model was also employed for large-area compaction testing. Soil was placed in controlled lifts or via air deposition to reach the target densities and moisture contents. Instrumentation included dynamic earth pressure cells, piezoelectric sensors, accelerometers, and imaging systems (e.g., GeoPIV, DIC) to record the crater depth, settlement, stress propagation, and strain localization. In a few cases, analytical or regression-based models (e.g., stochastic media theory, energy–depth relations) were proposed to predict improvement zones and validate laboratory outcomes against theoretical expectations. Despite differences in scale and configuration, the methodologies were systematic, repeatable, and well-suited to exploring DC behavior in varied soils. 2.3. Numerical Modeling Studies The numerical modeling studies on dynamic compaction (DC), summarized in Appendix C, applied a range of computational methods to simulate soil behavior under repeated impact loading. These included the Finite Element Method (FEM), Discrete Element Method (DEM), Smoothed Particle Hydrodynamics (SPH), Coupled Eulerian–Lagrangian (CEL), Arbitrary Lagrangian–Eulerian (ALE), and hybrid approaches. The models were used across various soil types, including dry and saturated sands, silty clays, loess, gravel fill, and collapsible deposits, under flat and sloped ground conditions. Soil behavior was defined using constitutive models such as Mohr–Coulomb, Drucker–Prager, cap plasticity, and Pastor–Zienkiewicz models. Typical input parameters included were density (14–26 kN/m 3), cohesion (0–70 kPa), internal friction angle (20–59°), Young’s modulus (5–75 MPa), and Poisson’s ratio (0.2–0.35), along with permeability, the damping ratio, the void ratio, and contact stiffness. In recent years, hydro-mechanical coupling was applied in studies involving saturated soils, facilitating advanced prediction of complex ground behavior under DC. Tampers were modeled as rigid or deformable bodies, with masses from 1.2 kg to 50 tonnes and drop heights of 0.2 m to 25 m. Tamper shapes included flat, cylindrical, conical, spherical, pyramidal, and hemispherical. The major objectives were often to evaluate the influence of drop energy, number of blows, and spacing on crater formation and the ground response. Model validation was conducted using laboratory tests, field data, or benchmark cases to ensure simulation reliability. Advanced numerical techniques were adopted to handle large deformation and nonlinear behavior. These included FEM–SPH coupling, large-strain Eulerian and ALE methods, and DEM with particle breakage (e.g., PFC3D). The SPH was also used in the mesh-free modeling of high-strain zones. Hybrid techniques, such as Lagrangian–SPH models, improved simulation efficiency by restricting mesh-free computation to impact regions. On the other hand, DEM has been used effectively to explore interparticle behavior such as particle rearrangement and energy transfer under DC. Several challenges were also reported. Lagrangian FEM suffered from mesh distortion during penetration. ALE and CEL methods addressed this but required longer computation times. SPH offered flexibility but was computationally intensive and limited to 3D. Most studies modeled only a single impact, while multi-pass effects were not well addressed due to strain accumulation complexity. Simulating wave propagation, energy dissipation, and pore pressure evolution, especially in layered or saturated cohesive soils, remains difficult due to nonlinear soil behavior. Furthermore, different constitutive and numerical modeling approaches exhibited varying capabilities in capturing the soil response under DC. The Mohr–Coulomb and Drucker–Prager models were computationally efficient and widely adopted, but had limited capability in simulating nonlinear stiffness variation, progressive densification, cyclic degradation, and pore pressure-related behavior. In contrast, critical-state and hypoplasticity models provided an improved representation of stress-dependent behavior, volumetric response, and strain localization, although they required more advanced calibration procedures. DEM contact models were particularly effective in simulating particle interaction, rearrangement, force-chain development, and large deformation behavior; however, they remained computationally demanding for large-scale engineering applications [ 108, 109]. Despite these limitations, the numerical models reviewed in Appendix C effectively captured the main mechanisms of dynamic compaction and have become valuable tools for performance evaluation, design optimization, and parametric analysis. 2.4. Analytical and Empirical Studies Analytical methods further extended these formulations by incorporating stress-based mechanics and idealized geometric assumptions. For example, Smits and De Quelerij [ 33] developed an equation that incorporates soil density, particle velocity, and vertical stress limits under the assumption of no lateral spread. Poran and Rodriguez [ 34] introduced a semi-spheroidal influence zone, with equations defining the horizontal and vertical extent of improvement as functions of energy and tamper geometry. These formulations are detailed in Section 1.3 and are conceptually linked to selected studies in Appendix A and Appendix C. The various equations for wave propagation and V p,max reported by various researchers are mentioned in Section 1.4. In laboratory work, Du et al. [ 110] in “ Appendix B” presented an analytical model based on stochastic media theory using crater shape parameters to estimate the internal densification zone. In field applications, Chow et al. [ 25], listed in Appendix A, applied a wave-equation-based analytical model using existing case histories to simulate the compaction influence and generate practical design guidance. These empirical and analytical approaches provide efficient tools for estimating the improvement depth, and wave propagation in the early stages of design or where numerical simulations are impractical. Their predictive capability depends heavily on accurate input parameters and calibration to specific site conditions and soil types, and they are typically best suited to uniform or simplified subsurface profiles. 3. Governing Factors of Dynamic Compaction The performance of dynamic compaction is governed by a combination of interdependent factors that influence the efficiency of energy transmission, the development of stress waves, and the degree of soil densification. These governing parameters include the tamper mass and drop height, tamper geometry, applied energy, grid spacing, number of blows per point, moisture content, etc. Their roles and interactions with varying soil types and compaction conditions directly affect the depth of improvement, lateral influence, and overall ground response. This section critically evaluates each parameter based on the synthesized findings from field investigations, laboratory experiments, numerical simulations, and analytical interpretations. The discussion highlights how these factors control key outcomes such as the crater depth, stress propagation, settlement behavior, and volumetric strain, providing a comprehensive understanding of their significance in dynamic compaction design and practice. 3.1. Impact of Tamping Energy Energy is a critical parameter in dynamic compaction, directly influencing the depth of improvement and the enhancement of soil strength. However, its effectiveness depends on the soil type, drainage conditions, and overall compaction behavior. A critical comparison of the provided studies reveals three consistent trends: (1) high energy is most effective in coarse dry soils; (2) moderate energy proves more efficient in collapsible and silty soils; and (3) lower to moderate energy is generally more appropriate for soft or saturated ground. The complete findings are presented in Figure 4. Importantly, the response of different soils to applied energy is not linear beyond an optimal level; increasing energy often results in diminishing returns or even adverse effects. Studies involving coarse-grained and well-drained soils demonstrate that high-energy DC (8–15 MN·m) can significantly increase the depth and uniformity of improvement. Feng et al. [ 111] applied 8 MN·m to very coarse gravel and reported improvement depths exceeding 10 m, with shear wave velocity increasing from 200–250 m/s to 350–450 m/s. Similarly, Wei et al. [ 112] applied 10 MN·m in soil rock mixtures and achieved a 14 m improvement in depth alongside a 93.2% increase in the deformation modulus. Extending this trend, Wei et al. [ 113] used up to 15 MN·m and observed 15 m of ground improvement and bearing capacities exceeding 900 kPa. Wang et al. [ 114] further confirmed this relationship by increasing energy from 4 to 15 MN·m in high-filled red soils, which resulted in depth increases from 7 to 13 m. However, they also noted that energy efficiency declined beyond 8 MN·m due to absorption by underlying weak layers. These studies confirm that high energy is essential for deep and adequate compaction in coarse soils, although, beyond a certain point, additional energy yields reduced returns. In contrast, a second group of studies shows that moderate energy levels (2–8 MN·m) are more efficient in collapsible, silty, or moderately compacted soils, where excessive energy can induce instability or unnecessary settlement. Zhang and Wang [ 115] applied 2–3.5 MN·m in coal gangue fill and achieved improvement depths between 5.38 and 7.92 m, with bearing capacity rising from 100 to 350 kPa. Feng et al. [ 116] evaluated 3, 8, and 12 MN·m in collapsible loess. They found 8 MN·m to be optimal, eliminating collapsibility up to 9 m, while 12 MN·m caused higher settlements with marginal additional benefit. Supporting these findings, Mei et al. [ 124] used ABAQUS to simulate DC in loess and showed that 2–6 MN·m energy effectively improved depths up to 9 m, increased dry density by up to 25%, and enhanced the modulus by 100–630%. In similar soil conditions, Feng et al. [ 117] reported that applying 8 MN·m to fine desert sands achieved a 12 m improvement in depth and bearing capacities over 450 kPa. Bo et al. [ 24], operating in loose granular sand using 2.94–5.63 MN·m, also fits within this group, having raised cone resistance to 18 MPa, though crater depth stabilized early, indicating that an optimal energy threshold had been reached. These studies support the fact that moderate energy levels are technically and economically efficient for a broad range of compact yet sensitive soils. Finally, in soft, fine-grained, or saturated soils, the third group of studies demonstrates that lower to moderate energy (1.5–6 MN·m) performs more reliably than higher-energy strategies, which often disturb the ground or prove ineffective. Miao et al. [ 118] found that 1.5 MN·m was more effective than 2.5 MN·m in soft interbedded soils, as the higher energy disrupted the weak layers and reduced improvement. Using numerical simulations, Zhou et al. [ 27] showed that 4 MN·m can improve dry sandy soil to 9 m but only reach 5.19 m in saturated conditions, reflecting the detrimental effect of excess pore pressure. Similarly, Alnaim et al. [ 119] introduced a behavior index (Ic), demonstrating that soils with Ic ≤ 2.6 responded well to DC, while those with Ic > 2.95 showed negligible improvement regardless of the energy applied. In reclaimed ground with sandy and clayey soils, Asaka [ 120] used 6.25 MN·m and reported significant SPT-N value gains in sandy layers, while cohesive zones improved gradually through pore pressure dissipation. Ismael and Al-Otaibi [ 121] used 2.1–2.4 MN·m in silty sand. They achieved up to a 7 m improvement depth, substantial increases in SPT, CPT, and modulus values, and bearing capacity above 300 kN/m 2. Together, these findings indicate that moderate energy applied with proper control measures in soft or saturated conditions is more reliable and effective than aggressive, high-energy compaction. Overall, the impact of energy in DC is closely tied to the soil type and condition. High energy (8–15 MN·m) enables deep and uniform improvement for coarse and free-draining soils. Moderate energy (2–8 MN·m) is sufficient and more cost-effective for collapsible or moderately compacted soils. Lower to moderate energy (1.5–6 MN·m) is generally more appropriate for soft or saturated soils, especially when applied with sufficient drainage time or monitoring. Therefore, energy should not be applied uniformly across projects but optimized based on site-specific conditions to ensure technical performance and economic efficiency. 3.2. Impact of Tamper Mass vs. Drop Height Although the total energy in dynamic compaction (E p = M·g·H) is traditionally considered to depend equally on the tamper mass and drop height, recent laboratory, numerical, and field studies show that tamper mass has a more dominant role in improving densification, stress propagation, and depth of improvement, particularly in cohesive and coarse-grained soils. A similar pattern was observed in modeling and field studies on coarse or heterogeneous soils. In dry rockfill, Ghassemi and Shah-ebrahimi [ 71] found that increasing the mass from 10 to 30 tonnes resulted in deeper craters than increasing the drop height from 5 to 25 m, with efficiency declining beyond 150 tm/drop. Li et al. [ 127] reinforced this through field and numerical studies on natural soils, including clay, gravel, basalt, and granite. Their results showed that heavier tampers produced deeper craters than lighter ones dropped from higher elevations, and the MH 0.2 index more accurately predicted the improvement depth than conventional energy equations. In both studies, mass governed the duration and penetration of stress, while height mostly affected surface stress peaks with limited depth influence. In layered or geometrically complex ground, the significance of mass also held. Wang et al. [ 65] showed that in gravel–sand fill over silty clay with cobbles, heavier tampers at lower drop heights achieved better compaction, while higher drops led to surface splash and instability. Conversely, Abdizadeh et al. [ 128] found that the drop height had a greater influence on sandy soils on steep slopes due to horizontal energy spread, enlarging the lateral influence zone. However, tamper mass again became the dominant factor on flat ground, indicating that terrain geometry can temporarily shift the effectiveness of mass vs. height. Taken together across red clay, dry sand, rockfill, and stratified soils, the evidence shows that tamper mass has a consistently greater impact than drop height in achieving more profound, uniform, and energy-efficient compaction. While increased height raises peak stress, it often causes shallow deformation and greater energy loss at the surface. The most effective results are achieved using heavier tampers dropped from moderate heights, aligned with MH 0.2–0.3-based models, which outperform traditional energy-based formulas in predicting real compaction behavior. However, it should be noted that several of these observations are derived from laboratory-scale experiments and numerical simulations, where scale effects, simplified boundary conditions, and controlled loading environments may influence the relative behavior of the tamper mass and drop height. Therefore, direct extrapolation of these findings to full-scale field applications should be conducted with caution, and additional large-scale field validation studies are still required to verify the applicability of MH-based relationships under practical engineering conditions. 3.3. Impact of Tamping Spacing Tamping spacing directly influences the uniformity and effectiveness of dynamic compaction (DC), as it governs the extent of energy overlap and distribution between impact points. Despite employing different research methodologies, Bonab and Zare [ 129] and Tashakori et al. [ 130] both recommend 4D T as the ideal tamping spacing for dry sandy soils. Bonab and Zare, through laboratory testing in fine dry sand (SP), found that 4D T spacing provided adequate overlap without unnecessary energy use. Although 3D T spacing resulted in stronger densification within the overlap zone, it was considered inefficient due to energy redundancy. Conversely, 5D T and 6D T spacings led to poor compaction between impact points and non-uniform improvement. In parallel, Tashakori et al. [ 130], using numerical modeling in dry sandy soil, confirmed that 4D T spacing achieved a uniform improvement profile with balanced energy efficiency. Their study further emphasized that spacings greater than the improvement depth, or more than 4D T, resulted in insufficient overlap and weak improvement zones. They also reported that a relative density increase of at least 10% was only achievable when spacing was 0.7 times the improvement depth or less. Additionally, Kundu and Viswanadhan [ 131] emphasized the role of tamper geometry, reporting that 3D T spacing was optimal for flat tampers, whereas 4D T spacing was better suited for conical tampers, making conical tampers more cost-effective due to their higher spacing requirement in dry sand. In contrast, Chow et al. [ 25] analyzed field data from thirteen DC projects across sites with loose sands, hydraulic fills, and interbedded silty clays. Rather than recommending a fixed spacing, they established S/D T = 2.1 as the maximum effective ratio for achieving full improvement at critical locations (center and side midpoints). Beyond this limit, compaction effects diminished significantly. They also observed that energy intensity alone was insufficient to ensure uniform soil improvement if spacing exceeded this threshold. Model predictions and field observations showed that lateral soil improvement extended to 3.5D T from each impact point, beyond which compaction effects were negligible. Figure 5 shows tamping spacing recommendations by Bonab and Zare [ 129], Tashakori et al. [ 130], Chow et al. [ 25], and Kundu and Viswanadham [ 131] for various soil types. Wider spacing applications, however, have yielded mixed outcomes. Wang et al. [ 114] applied both 6 m and 10 m spacing in high-filled red soil and found that the smaller spacing produced better strength and lower settlement. Despite high-energy input across both zones, the 6 m grid consistently performed more effectively. Similarly, Feng et al. [ 116], working in collapsible loess with 8–10 m spacing and energy levels up to 12 MN·m, reported an improvement depth of up to 11 m. However, inconsistencies between impact points were observed, particularly in the 8 and 12 MN·m zones. While these were primarily attributed to poor drainage, air cushions, and timing inefficiencies, the wide spacing likely contributed to uneven strength development between impact locations. Collectively, the findings from Bonab and Zare [ 129], Tashakori et al. [ 130], Kundu and Viswanadhan [ 131], and Chow et al. [ 25] provide a robust framework for optimal grid spacing in dry, sandy, and loose granular soils. These studies suggest that tamping spacing should ideally lie within the 3D T–4D T range and not exceed 4D T. Supporting field investigations reinforce this guidance, demonstrating closer spacing yields more reliable and uniform soil improvement. While higher energy input can partially compensate for wider spacing, it cannot entirely eliminate the risk of untreated or poorly compacted zones between impact points. Therefore, tamping spacing must be treated as a critical design parameter in DC, carefully adjusted to the site-specific conditions, soil type, and improvement depth requirements. 3.4. Impact of Number of Blows (Drops) The number of blows applied at each tamping point in dynamic compaction (DC) significantly influences ground improvement by affecting settlement, the improvement depth, and densification. However, the effectiveness of each additional blow is not constant. A consistent pattern across field, laboratory, and numerical studies shows that most improvement occurs during the early blows, while later impacts result in reduced effectiveness. Laboratory studies further confirm these trends. Bonab and Zare [ 129] and Jafarzadeh [ 132] reported that 5–10 drops were sufficient to reach the maximum relative density and stress response in dry sand. Li et al. [ 126] found that stress peaks stabilized after just four blows despite continuing to 20 drops in red clay. Similarly, Arslan et al. [ 70] showed that only 4–7 blows were required using conical tampers in dry sand to match the performance of 10 drops with flat tampers, demonstrating the efficiency of tamper geometry in reducing required blows. These combined results show that dynamic compaction is most effective during the early tamping stage, typically within the first 5 to 10 blows. Beyond this range, additional blows offer limited improvement and may lead to adverse effects, especially in fine-grained or saturated soils. Therefore, selecting the optimal number of blows based on the soil type, moisture conditions, and applied energy is essential for achieving safe, uniform, cost-effective ground improvement. 3.5. Impact of Tamper Geometry In the dynamic compaction process, the geometry of the tamper affects how stresses are distributed throughout the soil and its effect on the improvement depth, efficiency of energy transfer, and overall lateral densification. However, its effectiveness depends heavily on soil characteristics, including the gradation, density, fine content, and moisture percentage. Studies by Feng et al. [ 68], Arslan et al. [ 70], and Nazhat and Airey [ 64] consistently show that conical and convex tampers outperform flat shapes in loose or fine-containing soils due to their ability to direct energy vertically and achieve deeper densification ( Figure 6). Feng et al. [ 68] compared flat and 90° conical pounders in Mai-Liao sand (with 8% fines and contractive behavior) and Ottawa sand (clean and dilative). In Mai-Liao sand, the conical tampers created craters with 95% higher volume due to increased shear deformation, while the effect was negligible in Ottawa sand. This highlighted how fine content and contractive behavior enhance the effectiveness of conical shapes. Arslan et al. [ 70] evaluated conical tampers with different angles in well-graded sands and observed that they produced craters 38% deeper and up to 20% larger in volume than flat ones in loose conditions. However, the advantage declined as density increased, showing that shape benefits were reduced in stiff, dilative materials, which is consistent with the finding by Feng et al. [ 68]. Nazhat and Airey [ 64] extensively compared flat, conical, convex, and shell tampers in clean Sydney sand and a sand–silt mixture. In the silty mix, conical tampers performed best, while in clean sand, convex tampers produced the most improvement, showing up to 3.4 times the volumetric strain of conical tampers. Shell tampers underperformed in all conditions. They emphasized that crater depth and surface heave were unreliable indicators and that strain-based internal measurements provide a better understanding of actual densification. Numerical studies further support the above findings. Mehdipour and Hamidi [ 69] showed that the crater depth increased by 42% to 70.5% in dry sands of varying densities when the tamper cone angle matched the soil’s internal friction angle (α = φ). However, they also found that flat tampers resulted in better radial improvement, whereas conical tampers concentrated energy vertically. They suggested that conical tampers require closer spacing to achieve uniform improvement—a recommendation also supported by Arslan et al. [ 70] and Nazhat and Airey [ 64]. In contrast, Kundu and Viswanadham [ 131] found that conical tampers produced craters 16% deeper and 18% larger in volume than flat ones under the same energy level and improved soil vertically and horizontally as shown in Figure 6. Their GeoPIV displacement tracking showed that conical tampers extended influence zones up to 4D T, while flat tampers were limited to 3D T, challenging previous concerns that conical shapes lack lateral influence. Valdi et al. [ 66] numerically studied the water-entry behavior of cube, cylinder, sphere, pyramid, and cone tampers using the CEL method in Abaqus to evaluate displacement, velocity, and pinch-off characteristics. The cone showed the least velocity loss and deepest penetration, while the pyramid and sphere performed moderately, better than the cube and cylinder but less effective than the cone. Conical and convex tampers are highly effective for deep densification in loose, fine-containing, contractive soils but often require closer grid spacing due to their vertical energy focus. Flat and hemispherical tampers, though less aggressive in depth, are better suited for dense, cohesive, or saturated soils where uniform lateral improvement is essential. Shell tampers consistently underperformed in all studies. Therefore, tamper shape should be selected based on site-specific soil behavior and project goals. As emphasized by Nazhat and Airey [ 64], surface measurements alone are insufficient; internal strain-based evaluation is necessary to assess compaction effectiveness accurately. 3.6. Impact of Soil Type The effectiveness of dynamic compaction (DC) is strongly governed by the soil type, with distinct behaviors observed in granular, cohesive, saturated, and mixed soils. Dry granular soils, particularly loose sands and gravels, consistently deliver the most efficient and deep densification. In loose sands with 2.95 showed minimal improvement (0–1 MPa). Similarly, Feng et al. [ 68] reported liquefaction in wet Mai-Liao sand (Dr = 60%), limiting crater development and densification. These outcomes confirm that saturated soils reduce effective stress (σ′), requiring drainage or replacement for effective DC. Overall, clean dry sands, gravels, and well-graded rock fills show the highest DC effectiveness due to permeability and structural uniformity. Cohesive and collapsible soils are moderately responsive but highly sensitive to moisture and drainage. Saturated and silty soils show poor densification unless pre-treated. The fines content should ideally remain below 15%, with 6–10% proving optimal. Mixed fills and soil–rock combinations can perform well if energy and layering are controlled. Therefore, soil identification, fines quantification, and moisture assessment are prerequisites for successful DC application. 3.7. Impact of Moisture Content Moisture content plays a central role in determining the success of dynamic compaction. It influences how energy from the tamper is transmitted through the soil, how much energy is used for densification, and how pore pressure develops during the process. A critical literature review reveals that soils with optimum or moderate moisture content respond best to DC, while very dry or highly saturated soils result in poor compaction outcomes. Rollins et al. [ 137] demonstrated, in a field study, that as the moisture content increased from relatively dry conditions (6–10%) to near-optimum levels (15–18%), the depth of improvement rose from 3.25–4.2 m to 5–5.5 m. However, when moisture exceeded 18%, the improvement sharply declined due to pore water pressure build-up, which reduced effective stress. Crater depths also increased with moisture (from 0.48 m to 2.34 m), and tamper deceleration decreased, showing that wetter soils became softer and absorbed more energy. Zhang et al. [ 138] found similar results in silty clay, where compaction was more effective at 13% moisture than at 7%, even when high energy (8 MN·m) was applied. Efficiency dropped significantly at low moisture, and further increasing energy after a certain point had little benefit, confirming that moisture content is more critical than just applying more energy. These findings are supported by numerical models as well. Ghorbani et al. [ 67] showed, through simulations, that soils with intermediate saturation (around 50%) allowed better energy transmission and produced greater vertical displacement under the tamper. This indicates stronger densification compared to dry or fully saturated soils. Their later study [ 139] also confirmed that ground vibrations (V p,max) decreased as moisture increased because water absorbed more impact energy, leaving less energy for compaction and settlement. The air and water phases in the soil reduced its responsiveness to DC. On the other hand, studies that included saturated or low-permeability soils presented a different challenge. Zhou et al. [ 27] used advanced numerical modeling and found that the improvement depth dropped from 9 m in dry soil to just 5.19 m in saturated soil under the same energy level (4 MN·m). Saturated soils could not dissipate excess pore pressures effectively, which blocked deeper improvement. Furthermore, soils with higher permeability permit greater energy transmission compared to those with low permeability, even under wet conditions. Zhou et al. [ 27] introduced a relative enhancement index (R D) to assess the depth of improvement below the water table and to compare the performance of unsaturated and saturated soils, clearly demonstrating that the presence of water reduces the effectiveness of the compaction depth. These studies demonstrate that moisture content significantly influences dynamic compaction behavior; however, the response differs between cohesive and non-cohesive soils. In cohesive soils, moderate moisture conditions, often near the Proctor optimum moisture content (OMC), generally improve energy transfer and densification efficiency, whereas excessive moisture may increase pore water pressure generation and reduce compaction effectiveness. In contrast, non-cohesive soils are more strongly influenced by drainage behavior, particle rearrangement, and the degree of saturation, where fully saturated conditions may lead to inefficient energy dissipation and reduced improvement efficiency. Therefore, proper moisture control is essential in the design and execution of dynamic compaction projects. This includes adjusting moisture through pre-wetting, dewatering, or scheduling compaction during favorable moisture conditions to ensure maximum effectiveness and depth of improvement. Overall, the reviewed studies demonstrate that the effectiveness of dynamic compaction cannot be evaluated based on individual parameters alone, as the governing factors are strongly interdependent. The influence of tamping energy, tamper mass, drop height, spacing, number of blows, moisture content, and soil type is closely coupled and varies according to site conditions. For example, the effectiveness of high energy levels depends significantly on soil permeability and saturation conditions, while the influence of tamper geometry is strongly affected by the soil density and fines content. Similarly, the performance of tamping spacing and number of blows is governed by the depth of improvement, drainage behavior, and stress-wave interaction between impact points. Therefore, practical dynamic compaction design requires a multi-parameter integrated approach rather than isolated single-parameter evaluation to achieve reliable and efficient ground improvement. 4. Identified Research Gaps and Future Research Directions The current broad evaluation of field, laboratory, analytical, and numerical studies on dynamic compaction (DC) has revealed significant gaps as summarized in Table 3, which limit the full exploration and best use of this technique. Consequently, improved modeling systems, advanced instruments, broader soil characterization, and cross-disciplinary validation are needed. For every gap, future research directions are provided to address current limitations and support the further development of DC practices. 4.1. Mechanism and Prediction Models Most existing formulas and models are empirical and lack rigorous physics theories; they hence fail to incorporate all major governing parameters. These often provide generalized predictions that do not align with field behaviors across varied geotechnical conditions. Future research should focus on developing comprehensive, physics-based, or hybrid data-driven models and equations that incorporate multiple parameters and are validated against laboratory and field-scale experimental results. Existing studies have predominantly focused on a single parameter, for example, energy, drop height, or spacing. The combined outcomes of various factors, including moisture, the number of blows, the geometry of the tamper, the soil type, energy applied, and grid pattern, are seldom examined. The timing between impacts, which significantly affects pore pressure dissipation and stress recovery, remains unexamined. Further studies should conduct multi-variable tests and investigate the temporal sequencing of blows to reflect field-relevant loading conditions. Although tamping grid spacing and applied energy are commonly reported in the reviewed studies, many investigations primarily interpret compaction effectiveness based on total applied energy without systematically considering energy-distribution effects. In practical dynamic compaction applications, similar total energies applied under different tamping grid spacings may produce significantly different improvement responses due to variations in energy-transfer intensity. Therefore, the concept of energy per unit area or energy density should receive greater attention in future DC studies to improve comparison reliability and practical design interpretation. Many studies report basic soil index properties; however, only a limited number adopt well-defined classification systems such as USCS or AASHTO. Without standard soil symbols, the results cannot be generalized, and comparisons between studies or model applications become difficult. Problematic soils such as soft clays, collapsible loess, and expansive soils have also received limited attention in field experiments and numerical investigations. Future research should prioritize standardized classification protocols and extend testing to challenging soil types under varying compaction conditions. In addition, although the fines content is commonly reported in experiments, its direct influence on drainage behavior, the pore water pressure response, energy absorption, particle rearrangement under loading, and compaction efficiency remains insufficiently explored. Future work should therefore investigate the fines content more systematically in both experimental and numerical studies, with particular emphasis on threshold values influencing compaction performance. Numerous lab experiments rely on test boxes that are only one cubic meter or less, making it challenging to show wave propagation accurately. High-resolution approaches such as PIV, DIC, or advanced cameras are not generally used, limiting an understanding of the internal distribution of strain and the changes in soil structure during compaction. Moreover, many numerical simulations remain 2D and use simplified constitutive models (e.g., Mohr–Coulomb, Drucker–Prager) that do not capture essential soil behaviors such as anisotropy, strain softening, rate dependency, or particle breakage. Future research should advance toward large-scale experimental setups and adopt three-dimensional, high-fidelity numerical models using advanced soil constitutive frameworks. Certain simulation models incorporate basic damping parameters (e.g., Rayleigh damping and hysteretic damping) and define friction coefficients between the soil and tamper. Nevertheless, the impact interface is still simplified in most numerical approaches. Detailed contact mechanisms, including stress concentration zones, plastic deformation, and interface slip, are often represented using oversimplified penalty functions or linear spring models. Furthermore, the incomplete representation of energy dissipation caused by internal damping, rebound effects, interface friction, and insufficient calibration against experimental observations remains a significant limitation. This reduces the accuracy of stress prediction and limits an understanding of energy propagation within the soil mass. Future modeling efforts should incorporate more realistic contact formulations and validate energy-dissipation mechanisms through high-speed data acquisition and comparisons with controlled experimental results. 4.2. Field Monitoring and Parameter Verification Various laboratory and numerical studies have examined the influence of tamper geometry on soil behavior; however, most field investigations still rely on flat or cylindrical tampers. To date, the effect of tamper shape on energy transfer, crater formation, and improvement depth in cohesive, silty, or heterogeneous soils has not been adequately investigated under large-scale field conditions. Future investigations should therefore evaluate different tamper geometries to better assess their influence on energy efficiency and compaction performance. Saturation is widely recognized as a critical factor influencing pore pressure, effective stress, and densification behavior; however, the in situ monitoring of these responses in field studies remains limited. In addition, most numerical models employ uniform or simplified boundary conditions without validating their predictions through real-time hydro-mechanical response measurements. Future investigations may benefit from the use of piezometers and TDR probes for field data acquisition, together with hydro-mechanical numerical modeling, to improve prediction accuracy in saturated or low-permeability soils. Current investigations primarily focus on short-term performance indicators such as immediate settlement and CPT/SPT improvement. In contrast, long-term behaviors including stiffness evolution, secondary settlement, and structural resilience under operational loading remain insufficiently explored. This limits confidence in the long-term durability and life-cycle performance of DC-treated ground. Future research should therefore implement extended post-compaction monitoring programs and develop life-cycle-based performance models to evaluate the long-term reliability and durability of dynamically compacted ground. Advanced field instrumentation, including pore pressure transducers, accelerometers, strain gauges, and stress-wave sensors, is still rarely deployed during compaction operations, particularly at the field scale. Consequently, the lack of embedded monitoring restricts the validation of numerical models and limits understanding of internal soil behavior in real time. In addition, laboratory studies often lack internal strain-tracking systems such as Digital Image Correlation (DIC) or embedded sensors. Future efforts should therefore integrate advanced sensing technologies, including DIC and Particle Image Velocimetry (PIV), to obtain high-fidelity strain and displacement data for comprehensive analysis and model validation. Only a limited number of investigations compare numerical simulations, laboratory experiments, and field observations simultaneously. Most studies confine their findings to a single domain, reducing the broader applicability of the results. Consequently, the implementation of integrated validation approaches has become increasingly important. Future research should therefore incorporate field, laboratory, and numerical platforms together using standardized and clearly defined calibration and verification protocols. 4.3. Environmental Sustainability Quantification Dynamic Compaction is generally regarded as an environmentally friendly ground-improvement technique; however, existing studies primarily assess impacts such as vibration, noise, and energy consumption in isolation, with limited emphasis on their quantification within a unified framework. Critical sustainability indicators—including carbon emissions, resource efficiency, and soil ecosystem effects—are rarely quantified in an integrated manner. This lack of systematic quantification restricts robust comparisons with alternative ground-improvement methods and limits the incorporation of environmental criteria into design optimization. Future research should focus on developing integrated, quantitative frameworks that combine measurable environmental, technical, and economic indicators, supported by field data, life-cycle assessments, and data-driven approaches, to enable more reliable and sustainable DC design. 4.4. Intelligent Design Tools and Artificial Intelligence Integration Despite the rapid development of artificial intelligence (AI) in various engineering fields in recent years [ 140, 141], its applications to dynamic compaction have still been very limited. With data-driven strategies, design accuracy can be improved, risks will be lower, and decisions can become automated for DC applications. Future research should use AI algorithms for pattern recognition, live predictions, and decision-making, which could make DC project design and implementation perform better, more accurately, and flexibly. The development of a “Smart Tool” capable of capturing the numerous factors influencing the dynamic compaction process would offer significant benefits to practicing engineers. Such a tool should integrate outcomes from theoretical analyses, numerical simulations, field observations, and experimental investigations to provide a comprehensive assessment of site conditions. It should also evaluate the consistency and reliability of the input parameters and, where datasets are incomplete, recommend a plausible range of values for various soil types based on existing correlations and empirical evidence. The predictive model embedded within the tool must be user-friendly, allowing efficient data input and the rapid generation of informative
Sustainability, Vol. 18, Pages 5827: Dynamic Compaction for Ground Improvement: Mechanisms, Governing Parameters, Environmental Impacts, and Multiscale Research Approaches
