Static load tests on footings supported by rigid inclusions

A series of full-scale static loading tests on square footings supported by rigid inclusions was conducted as part of the French research project ASIRI+. The primary objective of these experiments was to evaluate the impact of a load transfer platform (LTP) between the shallow foundation and rigid inclusions on the performance of the footing and enhanced soil. To achieve this, various loading configurations were examined, including vertical loading with and without eccentricity, as well as horizontal loading, across different structural scenarios—shallow foundations on soil, shallow foundations on reinforced soil with or without a load transfer platform, and rigid inclusions. The comprehensive data obtained from these experiments contribute to a deeper understanding of the load transfer mechanisms in reinforced soil, elucidating the function of each component and facilitating the calibration of numerical models.

1 Introduction

The mechanical resistance of soil often necessitates the use of deep foundations to effectively transfer structural loads to deeper soil layers capable of accommodating such demands. In certain soil conditions and under specific load scenarios, footing piles may be substituted with rigid inclusions. This method of soil reinforcement is frequently more cost-effective and environmentally friendly. The French guidelines outlined in ASIRI (2013) advocate for the implementation of a load transfer platform between the footing and rigid inclusions. However, the inclusion of this platform incurs additional costs associated with the movement of the earth and the provision of gravel for its construction. Therefore, many construction firms have developed specifications that allow for footings to be placed directly on the rigid inclusions. To standardize this approach, the French research project ASIRI+ (2019–2025) aims to evaluate the performance of footings on rigid inclusions across various configurations. Full-scale experiments were conducted at three sites in France during the ASIRI+ project. Comprehensive geotechnical investigations were performed at each location, along with preliminary load tests on isolated rigid inclusions to evaluate their behavior under compression. Various loading configurations were examined, including vertical loading with and without eccentricity, horizontal loading, and footings on both non-reinforced soil and on reinforced soil with or without a load transfer platform. Specialized instrumentation was employed to measure settlement and load transfer, facilitating a thorough analysis of footing behavior under diverse loading conditions.

2 Background

Soil reinforcement with rigid inclusions under footing is not well documented or not studied. Many papers on rafts on piles exist but the mechanisms are very different. Some papers present centrifuge tests of non-connected piled raft system (El Sawwa, 2010; Fioravante, 2011; Rasouli et al., 2015). El Sawwa (2010) investigated the effectiveness of using short piles either connected or unconnected to the raft on the behavior of an eccentrically loaded raft. He specifically studied the pile arrangement, the number of piles but the comparison between cases with and without platform between the raft and the piles were not studied. Fioravante (2011) studied the behavior of a raft on a connected single pile and of a raft on a single pile with an interposed layer. This study focused on the load mechanisms with or without negative friction along the pile. Due to the size of a raft in comparison with a footing and then number of piles, the results of studies on piled raft are non-adapted to a case of footing on rigid inclusions.

The first guideline on the application of footing on rigid inclusions was found in the ASIRI guideline (2013). As a part of this project, Blanc et al. (2004) investigated, with small-scale centrifuge models, the behavior of a footing above a pile supported earth platform. The case without platform was not investigated. From their results, the ASIRI guideline recommend to add systematically a load transfer platform between the footing and the rigid inclusions. As a part of the ASIRI+ project (2019–2026), Thorel et al. (2024) investigated, with small-scale centrifuge models, the load eccentricity effect of a square foundation resting on 4 rigid inclusions. The results show that for a settlement of 10% of the width of the foundation, the rotation reaches a value of about 2°, and that the front RIs catch three times more load than the rear RIs. During the loading of each foundation, two mechanisms seem to appear: (1) load transfer mainly in the LTP; (2) a load transfer more active towards the RIs. Unfortunately, results without load transfer platform were not presented.

Pham et al. (2019) investigated the behavior of footing on rigid piles without load transfer platform, but no comparison was done with load transfer platform. They observed that:

• The stress on the inclusion is significantly higher than the stress on the soft soil,

• The pressure on the inclusion increases linearly with the vertical loading,

• The eccentricity of the vertical loading also creates a difference in the vertical pressure on the inclusions.

3 Materials and methods

3.1 Experimental Site 1

3.1.1 Configurations and geotechnical context

Two isolated piles and three footings were tested at Site 1. The geological and geotechnical conditions were relatively straightforward, characterized by a single layer of silts. All tests were conducted on a uniform precast reinforced concrete footing, measuring 1.8 m × 1.8 m × 0.5 m. The rigid inclusions implemented at this site consisted of full-displacement piles, each 4.5 m in length and 0.27 m in diameter. The piles were arranged in a square grid beneath the footings, with a side length of 1.2 m.

Cone penetration tests (CPTs) conducted in each section revealed substratum depths of 5.6 m in Section 1, 5 m in Section 2 and 5.2 m in Section 3, indicating that the piles were insufficiently long to be classified as anchored (Figure 1). The soil profile was established based on the results of the CPT and pressuremeter tests, as shown in Figure 1. At ground level, there exists a 0.7 m thick embankment layer, followed by a silt layer of 2.5 m (soft soil). Beneath this is a 2.5 m layer of green sandy clay, underlain by a gray sandy clay layer with superior properties. Groundwater level was monitored at a depth of 0.3 m.

Figure 1

Figure 1. Configuration of the three sections of the footings in Site 1.

The various test configurations are schematically represented in Figure 1, along with the outcomes of the penetrometer and pressuremeter tests, summarized as follows:

• Section 0: Two isolated inclusions.

• Section 1: A footing supported by four rigid inclusions without a load transfer platform.

• Section 2: A footing supported by four rigid inclusions with a load transfer platform of 30 cm composed of granular material.

• Section 3: A footing supported by four rigid inclusions with a 10-layer of leveling sand.

3.1.2 Loading procedures

In each section with footing, five tests were conducted to investigate the behavior of footings over soil improved using rigid inclusions without a granular platform (Table 1). Tests T1 and T5 assess the system responses under central vertical loads. The structure was subjected to eccentric vertical loads at T2 and T3. Combined vertical/horizontal loading was applied in test T4.

Table 1

Table 1. Loading applied in each section for Site 1.

Vertical loads were applied using a reaction frame that mirrors the design typically employed for pile testing. For horizontal loads, a concrete block was utilized, supported four large piles to achieve the necessary reaction.

3.1.3 Instrumentation

The load transfer, footing displacement, pile strain, and settlement were measured during the tests. To measure the load transferred to the inclusions and subsoil, four Earth pressure cells (0.27 m in diameter) were placed on the pile caps, with a cell positioned between the piles on the subsoil. Four displacement sensors were installed at each corner of the footing, along with one at the footing center, facilitating comprehensive monitoring of footing displacement during the tests. Settlement sensors were also positioned on each Earth pressure cell. Additionally, a force sensor was connected to the actuator to control the applied load. The measurement instruments used are shown in Figure 2.

Figure 2

Figure 2. Details of the instrumentation for Site 1.

3.2 Experimental Site 2

3.2.1 Configurations and geotechnical context

At Site 2, testing was conducted on an isolated pile and three footings. The geotechnical profile is shown in Figure 4, which indicates a stratigraphy comprising 2.5 m of backfill, 3.1 m of alluvium, 6.4 m of fragmented chalk, and a substratum composed of compact chalk. The rigid inclusions installed at this site were continuous flight auger inclusions, each measuring 12 m in length and 0.4 m in diameter.

The various test configurations are schematically represented in Figure 3, alongside the results from the penetrometer and pressuremeter tests, summarized as follows:

• Section 0: An isolated inclusion.

• Section 1: A footing on non-reinforced soil.

• Section 2: A footing supported by four rigid inclusions directly installed on the piles.

• Section 3: A footing supported by four rigid inclusions with a 30 cm load transfer platform.

Figure 3

Figure 3. Configuration of the sections in Site 2.

All tests utilized the same precast reinforced concrete footing, with dimension of 2.8 m × 2.8 m × 0.5 m. In Sections 2, 3, the piles were arranged in a square mesh configuration with a side length of 1.6 m, positioned side by side. Both the isolated pile and footing were subjected to central vertical loading.

3.2.2 Instrumentation

For all sections, a load cell was positioned between the jack and beam, and displacement sensors were installed on the tested components to facilitate the plotting of force–displacement curves.

In Sections 2, 3, testing was performed on the same four piles, with a layer of granular material introduced between the piles and footing for the test in Section 3. The instrumentation employed in both sections included (Figure 4):

• A force sensor was situated between two metallic plates on each pile.

• A settlement sensor was placed between two metallic plates on each pile and one on the soil at the center of the grid.

• An earth pressure cell was located at the center of the grid.

Figure 4

Figure 4. Details of the instrumentation for Site 2.

3.3 Experimental Site 3

3.3.1 Configurations and geotechnical context

Two isolated piles and three footings were tested at Site 3. The geotechnical profile is shown in Figure 5, which indicates a 1 m layer of backfill, 1.6 m of clay, and 4 m of schistous clay resting on a schistous substratum. The rigid inclusions installed at this site were continuous flight auger inclusions, measuring 4.5 m in length and 0.32 m in diameter.

Figure 5

Figure 5. Configuration of the sections in Site 3.

The various test configurations are schematically represented in Figure 5, along with the results of the penetrometer and pressuremeter tests. These configurations are summarized as follows:

• Section 1: A footing placed on non-reinforced soil.

• Section 2: Two isolated inclusions.

• Section 3: Four rigid inclusions set up directly on piles.

• Section 4: Four rigid inclusions set up directly on piles.

In each section, a reinforced concrete footing with dimensions of 2 m × 2 m × 0.5 m was cast in place, directly on the soil for Section 1 and on the piles for Sections 3, 4. In Sections 3, 4, the piles were arranged in a square mesh configuration with a side length of 1.2 m, positioned side by side beneath the footings.

3.3.2 Loading procedures

Vertical loads were applied through the solicitation of a reaction frame identical to that used for pile testing. For the horizontal solicitations applied in Section 4, a large concrete slab was utilized as a reaction mechanism, positioned adjacent to the section. In Section 3, three loading procedures were implemented on the footing, whereas in Section 4, a vertical load of 560 kN was directly applied to the footing before applying the horizontal loading. The loadings applied to each section are listed in Table 2.

Table 2

Table 2. Loading applied in each section for Site 3.

3.3.3 Instrumentation

For all sections, a load cell was positioned between the jack and beam, complemented by displacement sensors installed on the tested components to facilitate the plotting of the force–displacement curves.

In Section 3, the same instrumentation than those installed in site 2 was used:

• A force sensor, positioned between two metallic plates on each pile.

• A settlement sensor, placed between two metallic plates on each pile and one on the soil at the center of the grid.

• An earth pressure cell, positioned at the center of the grid.

In Section 4, the following sensors were installed (Figure 6):

• Two displacement sensors to measure the horizontal displacement of the footing.

• Two displacement sensors to measure the horizontal displacement of the piles. Two holes were formed within the footing to facilitate the installation of the metal rod inside the pile and connection of the displacement sensor.

Figure 6

Figure 6. Details of the instrumentation in Section 4 of Site 3.

4 Results

4.1 Site 1

4.1.1 Load tests on piles

Two static load tests were conducted on the two piles to evaluate their end-bearing and shaft friction. A theoretical resistance of 260 kN was estimated from pressuremeter tests. The second pile was subjected to ten unloading cycles–reloading between 0 and 100 kN, with instantaneous unloads and reloads and bearings maintained for 2 min . Pile 1 failed under an axial load equal to 220 kN, with 80% attributed to shaft friction and 20% to end-bearing. Pile 2 failed under an axial load of 180 kN, with 65% from shaft friction and 35% from end-bearing. The observed difference of 40 kN in failure loads between Pile 1 and Pile 2, this difference can at least partly be explained by the fatigue and displacement accumulation caused by the 10 cycles undergone by the pile 2 The analysis of the creep curves indicates a creep resistance of 110 kN for Pile 1 and 100 kN for Pile 2.

4.1.2 Distribution of load on piles below the footing

The measured stresses on each pile in Sections 1, 3 are shown in Figure 7. In Section 1 (Figure 7A), during the initial four load stages, the stress was relatively evenly distributed among the four piles. However, from the fifth stage (Q = 520 kN), a significant increase in stress was observed on Rigid Inclusion RI2, whereas the stress on Rigid Inclusion RI4 remained constant. In the final stage, the stress distribution among the four piles was no longer uniform. Similarly, in Section 3 (Figure 7B), a non-uniform transfer was evident across the four piles, despite the introduction of a thin layer of sand between the pile and footing. From the outset of the test, Rigid Inclusion RI4 experienced lower loading, with stress being more evenly distributed among the other three piles.

Figure 7

Figure 7. Site 1: Stress distribution on piles for (A) Section 1, (B) Section 3.

The curves of the mean stress values on the piles for Sections 1, 3 is shown in Figure 8A. Initially, the load transfer on the piles resulting from the weight of the footing is consistent across both tests, with an initial efficiency of 94%. Efficiency is defined as the load transferred on the pile head divided by the applied load. During loading, the load transfer to the piles was greater in Section 3 compared with that in Section 1. At the end of the test, with an applied load of Q = 650 + Q_footing = 690 kN, the efficiency decreased to 51% for Test 3 and 46% for Test 1.

Figure 8

Figure 8. Site 1: (A) Mean stress on piles. (B) Increase of stress on piles for each loading stage. (C) Stress measured at the head of the inclusion for the three sections.

In this instance, the presence of a thin layer of sand facilitated enhanced load transfer to the pile heads and a more stable distribution of load across the four piles. This enhanced performance may not solely be attributed to the sand layer; the anchorage of the piles in Section 3 (Figure 1) is also superior. Notably, the increase in stress on the piles was higher in Section 3 during the initial loading stages (Figure 8B). However, at the final loading stage at 690 kN, the stress increase became comparable for both sections. In full-scale experimentation, identifying the influence of each component on overall behaviour can be challenging. This difficulty arises because the sections are seldom perfectly identical, and even minor differences can significantly impact the results.

The stress increases measured on the rigid inclusions for the three sections at the end of the test for the centered vertical loading are shown in Figure 8C. Across all three configurations, the footing never rested evenly on the four inclusions. This observation aligns with findings from a similar experimental study (Pham et al., 2019). However, Section 3, which incorporates a leveling sand layer, seemed to promote a more effective distribution at the heads of the inclusions. In contrast, Section 2, featuring a granular platform, results in lower stress levels at the heads of the inclusions.

4.1.3 Influence of eccentricity

Owing to a malfunction with the data logger, stress measurements for Section 2 were unavailable for Tests T2 and T3. The stress increases measured on the rigid inclusions of Section 1 and 3 at the end of the test for central vertical loading (Test 1), eccentric vertical loading at 30 cm (Test 2), and eccentric vertical loading at 45 cm (Test 3) are shown in Figure 9. To evaluate the impact of eccentricity, average values were provided for two inclusions that were perpendicular to the direction of eccentricity. An imbalance of 60%–40% was observed in the central vertical test, even with the presence of a leveling sand layer between the footing and inclusions. At an eccentricity of 30 cm, a slight tilting was observed toward the two inclusions located on the eccentricity side. This tilting became more significant at an eccentricity of 45 cm, where the two inclusions accounted for 77% and 95% of the load applied to Sections 1, 3, respectively. This disparity can be attributed to the fact that, in Section 1 (Figure 9A), the eccentricity was applied in the opposite direction to the two inclusions experiencing the highest loads, whereas in Section 3 (Figure 9B), the eccentricity was applied in the same direction as these heavily loaded inclusions. The impact of eccentricity on load transfer is challenging to evaluate in this full-scale experiment, as the vertical load applied during the central test is not evenly distributed across the four piles.

Figure 9

Figure 9. Site 1: Impact of loading eccentricity for (A) Section 1, (B) Section 3.

4.1.4 Horizontal loading

In Section 1, a vertical load of up to 650 kN was applied and subsequently maintained. Under this vertical load, a horizontal load was introduced to the footing in increments of 50 kN, each held for 30 min. Notably, at 200 kN, the footing demonstrated a displacement of 2 mm, which increased to 6 mm at 250 kN. For safety considerations, testing was halted at this point. Stress measurements obtained at the heads of the four inclusions indicated a tilting of the footing toward the inclusions opposite the direction of the horizontal load.

In Section 2, a similar vertical load of 650 kN was applied and held constant. Under this vertical load, a horizontal load was applied to the footing in increments of 50 kN and held for 15 min. Continuous manual measurements revealed a displacement of the footing at a horizontal load of 150 kN, prompting the cessation of the test for safety reasons. The stress and vertical displacement measurements did not indicate any tilting of the footing.

In Section 3, test measurements were not recorded owing to a malfunction of the measurement station. However, the test was conducted up to a load of 250 kN, resulting in significant displacement that necessitated the termination of the test for safety reasons.

4.1.5 Ultimate bearing capacity

Test 4 evaluated the failure points of each configuration. The creep load was estimated based on the settlement of the piles at 900, 950, and 850 kN for Configurations 1, 2, and 3, respectively (Figure 10A). We consider that the presence or absence of an LTP between the piles and footing has no effect on the creep-load level. Notably, the settlement of the footing was most pronounced owing to the deformation of the LTP.

Figure 10

Figure 10. Site 1: Analysis of the rupture test. (A) Evaluation of the creep load. (B) Mean stress on piles.

The mean stress measured for the four piles as a function of the applied load is shown in Figure 10B. The stress observed in Section 2 was lower than that in Sections 1, 3, attributable to the presence of the LTP. Additionally, the stress measured in Section 1 was lower than that recorded in Section 3, which can be attributed to the floating piles in Section 1. Notably, the behaviors of Sections 1 and 3 were similar for loadings between 0 and 760 kN after the footing of Section 1.

At the final load of 1,660 kN, the load distribution across the four piles was 46%, 40%, and 79% for Sections 1–3, respectively.

4.2 Site 2

4.2.1 Vertical static load tests on pile (Section 0) and on footing on non-reinforced soil (Section 1)

The load test on the pile was halted upon reaching a settlement of 1.8 mm owing to an issue with the beam, preventing the determination of the creep load from this test. The footing was subjected to a load of 3,000 kN; however, the test could not proceed to the 200 mm settlement required to satisfy the failure criterion owing to the inclination of the footing. However, the creep load was exceeded and estimated at 1,800 kN.

4.2.2 Loading test on the footing on piles (Sections 2 and 3)

The same four piles were utilized for both Sections 2, 3, with the objectives of each test being to apply loadings below the creep force, rather than to induce failure. This approach allowed for a detailed analysis of load transfer within the piles and resulting settlement. In Section 2, a load test incorporating an unloading-reloading cycle was conducted, whereas Section 3 involved a single, straightforward loading application.

To analyze the behavior of both configurations, Figure 11 presents:

• The load applied to the piles for Section 2 (Figure 11A) and 3 (Figure 11B)

• Stress measured on the soil at the pile head level for Section 2 (Figure 11C) and 3 (Figure 11D)

• Settlement at the pile head level for Section 2 (Figure 11E) and 3 (Figure 11F).

Figure 11

Figure 11. Site 2: Analysis of the footing loading tests. (A) Load applied to the piles for Section 2. (B) Load applied to the piles for Section 3. (C) Stress measured on the soil at the pile head level for Section 2. (D) Stress measured on the soil at the pile head level for Section 3. (E) Settlement at the pile head level for Section 2. (F) Settlement at the pile head level for Section 3. (G) Settlement of the footing for Section 3.

In Section 2, the curve representing the displacement of the footing as a function of the applied load indicates that the creep load was not exceeded. Although the load application capacity was 5,000 kN, the test was halted at 3,375 kN owing to the uneven load distribution across the four inclusions (Figure 11A), with the F3 sensor going out of range at an applied load of 3,000 kN. During the initial loading phase, a load was applied to inclusions RI1 and RI3, whereas inclusions RI2 and RI4 experienced minimal loading, only slightly engaged during the final two steps. During the second loading phase, the load was initially detected by the inclusion of sensors RI1 and RI3. However, after a few steps, sensors F2 and F4 began to register significant loads. At a loading of 2,625 kN, a notable decrease was observed in the F1 sensor, accompanied by a sharp increase in sensors F2 and F4. The F3 sensor was overrange at this point, preventing us from capturing the data for this stage. It appears that the footing experienced a toppling effect during the loading at 2,625 kN.

This toppling phenomenon is further corroborated by the settlement sensors located on the inclusions (Figure 11E), with a sharp increase in the SS1 sensor, suggesting that inclusion RI1 may have suddenly become embedded. Additionally, sensors SS2 and SS4, as along with the SS_soil sensor positioned on the ground, displayed an inflection point at 2,625 kN. In the first loading (Figure 11E), sensor SS3, situated on the RI3 inclusion, recorded the most significant settlement, indicating it was the most heavily loaded (Figure 11A). This toppling is also observed from the settlement of the footing (Figure 11G).

The force and settlement measurements on the inclusions seemed consistent and effectively illustrated the footing behavior throughout both loading phases. The measurement of the ground stress between the four inclusions (Figure 11C) shows a very low load during the first loading and a higher stress during the second loading.

In Section 3, the curve representing the displacement of the footing as a function of the applied load reveals that the creep load has not been exceeded. This test could accommodate a higher load, as the measured values of the force sensors were significantly lower than those in Section 3, indicating they were well below their full scale. The load was primarily distributed across three inclusions (Figure 11B). The soil experienced greater loading owing to the presence of the load transfer platform (Figure 11D), which experienced significantly greater settlement compared with the inclusions (Figure 11F).

4.3 Site 3

4.3.1 Vertical static load tests on the piles (Section 2) and on the footing on non-reinforced soil (Section 1)

The static load test conducted on the first pile successfully reached failure at a displacement of 10% of the pile section (32 mm). This displacement was achieved at a load of 600 kN. However, for the second pile, failure could not be achieved owing to issues with the vertical alignment of the actuators.

In the static load test on the footing situated on non-reinforced soil, the test was halted at a vertical displacement of less than 200 mm, corresponding to 10% of the footing section. Based on Chin’s method (Chin, 1970; Chin, 1972), the failure force was determined to be 1,315 kN.

4.3.2 Vertical static load tests on footing on piles (Section 3)

A comparison of the three curves (Figure 12A) reveals that the initial loading resulted in a greater settlement for a given load. In the final test, failure was not reached (corresponding to a displacement of 200 mm); the test was terminated at 2,800 kN, resulting in a displacement of 123 mm. The estimated failure force was 3,585 kN (Figure 12B).

Figure 12

Figure 12. (A) Site 3: Static loading tests on the footing in Section 3 (B) Zoom on the beginning of the curve.

In the first test described in Section 3 (centric vertical loading SLS), the load distribution among the four piles was uneven (Figure 13A). Pile RI3 experienced the highest demand and demonstrated the most significant settlement (Figure 13C). The surrounding soil also experienced loading (Figure 13E) and settled more than four piles. At the final stage of loading, set at 780 kN, the cumulative measured loads on the piles totaled 348 kN, accounting for 44% of the applied load.

Figure 13

Figure 13. Site 3: Analysis of the centric and eccentric load tests SLS on the footing. (A) Test 1-Load applied to the piles. (B) Test 2-Load applied to the piles. (C) Test 1-Settlement at the pile head level. (D) Test 2-Settlement at the pile head level. (E) Test 1-Stress measured on the soil at the pile head level. (F) Test 2-Stress measured on the soil at the pile head level.

In the second test described in Section 3, which focused on eccentric vertical loading under serviceability limit state (SLS) conditions, the load was distributed between the two piles adjacent to the loading. However, pile RI3 consistently experienced the highest load (Figure 13B). The settlement observed in all four piles was lower than that recorded in the first test, attributable to the reduced load and inherent characteristics of subsequent tests on the same footing. Notably, while the soil was subjected to loading (Figure 13F), it did not demonstrate any settlement.

In the final test of Section 3, which examined centric vertical loading under ultimate limit state (ULS) conditions, the load distribution improved compared with the first test. However, piles RI3 and RI1 remained the most heavily loaded prior to unloading (Figure 14A). During the second loading phase, the footing on piles RI3 and RI2 was toppled, and the soil was also loaded (Figure 14C). At the final loading stage of 2,800 kN, the cumulative loads recorded on the piles totaled 1,200 kN, representing 43% of the applied load.

Figure 14

Figure 14. Site 3: Analysis of the centric load test ULS on the footing. (A) Load applied to the piles. (B) Settlement at the pile head level. (C) Stress measured on the soil at the pile head level.

4.3.3 Horizontal static load tests on footing on piles (Section 4)

In Section 4, a vertical load was applied with an overload of 560 kN, followed by the incremental application of a horizontal load to the footing, with increments of 25 kN after a first loading of 50 kN, each sustained for 30 min. Notably, the footing slid, despite being directly concreted on the piles (Figure 15). Similar tests were performed by Pham et al. (2019) on footing directly on four piles. They also obtained a sliding of the footing with low horizontal displacement of piles. This behavior is not taken into account in the design methods or not observed with numerical simulations. Others tests have to be done to understand these mechanisms of sliding and be used to calibrate numerical models.

Figure 15

Figure 15. Site 3: Horizontal static loading tests on the footing in Section 4.

5 Discussion

These three experimental studies were conducted at distinct sites, each characterized by varying experimental conditions, including:

• Different pile lengths

• Continuous flight auger inclusions or full-displacement piles

• Precast versus cast-in-place footings

Despite these variations, consistent observations emerged across all three sites.

In the vertical load tests, the load distribution among the four piles was uneven. Depending on the specific test conditions, two or three piles experienced greater loading. Notably, the load transfer dynamics could change during loading; if one heavily loaded pile settled more than the others, the footing risked toppling, particularly in the absence of LTP. The load transfer at the pile head was significantly higher in scenarios without LTP. Furthermore, the cumulative loads recorded at the pile heads did not equal the total applied load on the footing, even in cases lacking LTP (Table 3). Assuming that the stress measured in the soil between the four piles is uniform across the entire soil surface beneath the footing, the sum of the loads recorded at the pile heads and the load measured on the soil is less than the load applied to the footing (Table 3). This discrepancy between the applied and measured loads beneath the footing suggests that the load distribution is not uniform; instead, the soil surrounding the piles is subjected to higher stresses.

Table 3

Table 3. Load distribution on piles for vertical load tests.

For all tests conducted without LTP, soil solicitation was observed, and measurable settlement occurred at the pile head level between the four piles. This finding challenges the assumption of an extremely rigid footing utilized in the numerical modeling.

Overall, these tests demonstrate that the inclusion of LTP enhanced the distribution of the four piles and contributed to enhanced stability under substantial loading conditions.

The experiments only cover square footings and specific pile types (full-displacement piles and continuous flight auger inclusions). The results cannot be extended to other footing shapes (e.g., rectangular, circular) or inclusion types (e.g., driven pipe piles). For that, it is recommended to calibrate numerical models with the tests of these three sites and numerically explore others cases. This work will be done as a part of the project ASIRI+ and the results will be edited in the ASIRI+ guideline.

6 Conclusion

As part of the French research initiative ASIRI+ (2019–2026), three full-scale experiments were conducted to investigate the behavior of footings on rigid inclusions. These tests were performed at three sites with different pile lengths, types of rigid inclusions, and footing conditions. The experimental studies revealed the following:

• The load was not evenly distributed among the four piles, regardless of the presence of load transfer platforms.

• Even in the absence of a load transfer platform, the soil accounted for a significant portion of the load applied to the footings.

• The load applied to the soil was non-uniform.

• Under horizontal loading conditions, the footing experienced sliding while the heads of the rigid inclusions remained stationary.

The experimental results enabled the ASIRI+ project partners and other researchers to calibrate their numerical models and evaluate their design methodologies.

Data availability statement

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Author contributions

LB: Writing – original draft, Writing – review and editing. FS: Writing – review and editing.

Funding

The authors declare that financial support was received for the research and/or publication of this article. This work was supported by National project ASIRI+.

Acknowledgements

The authors extend their gratitude to all project members for their invaluable support.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The authors declare that no Generative AI was used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

ASIRi (2013). Recommendations for the design, construction and control of rigid inclusions ground improvements. Paris: Presses des ponts.

Google Scholar

Blanc, M., Thorel, L., Macé, D., Néel, A., and Rault, G. (2004). “Raft above rigid inclusions—centrifuge investigation of complex loading. Physical modelling in geotechnics,” London: Taylor and Francis Group, 591–596.

Google Scholar

Chin, F. K. (1970). “Estimation of the ultimate load of piles from tests not carried to failure,” in Proceedings of the second Southeast Asian conference on soil engineering, 83–91.

Google Scholar

Chin, F. K. (1972). “The inverse slope as a prediction of ultimate bearing capacity of piles,” in Proceedings 3rd Southeast Asian conference on soil engineering, 83–91.

Google Scholar

El Sawwaf, M. (2010). Experimental study of eccentrically loaded raft with connected and unconnected short piles. J. Geotechnical Geoenvironmental Eng. 136 (10), 1394–1402. doi:10.1061/(asce)gt.1943-5606.0000341

CrossRef Full Text | Google Scholar

Fioravante, V. (2011). Load transfer from a raft to a pile with an interposed layer. Geotechnique 61 (2), 121–132. doi:10.1680/geot.7.00187

CrossRef Full Text | Google Scholar

Pham, V. H., Briançon, L., Dias, D., and Racinais, J. (2019). Footings over rigid inclusion-reinforced soft soil. Experimental and numerical approaches. Can. Geotechnical J. 56 (12), 1940–1952. doi:10.1139/cgj-2018-0495

CrossRef Full Text | Google Scholar

Rasouli, H., Saeedi Azizkandi, A., Baziar, M. H., Modarresi, M., and Shahnazari, H. (2015). Centrifuge modeling of non -connected piled raft system. Int. J. Civ. Eng. 13 (2), 114–123.

Google Scholar

Thorel, L., Blanc, M., Neel, A., Jagu, A., Audrain, Ph., Gaudicheau, P., et al. (2024). “Centrifuge eccentric loading of square foundation over rigid inclusions with the 4 degree of freedom robot,” in Proceedings of the 5th European conference on physical modelling in geotechnics, 2-4 October 2024. Delft.

Google Scholar