Densification of Granular Soil

views updated

Chapter 15
Densification of Granular Soil


The reclamation of new land with hydraulic fill results in a loose profile of granular soil mass. This loose granular soil will contribute to high elastic immediate settlement as well as liquefaction upon dynamic forces. In addition, the bearing capacity of a granular foundation is mainly dependent upon shear characteristics such as the friction angle of the soil. The compressibility is in turn dependent upon the elastic modulus of the soil.

To increase the friction and the elastic modulus of granular soil, it has to be improved by a densification method. If reclamation is carried out by landfill operation, granular soil mass can be densified by roller compaction with a certain lift and a specified moisture. However, for existing land or land reclaimed by hydraulic filling, such method may not be feasible and hence one has to rely on deep compaction methods.

Before carrying out deep compaction, the first thing that needs to be done is to ensure that the type of soil is densifiable with the deep compaction method. Generally, granular soil with less than 10% of fine can be densified with this method. Figure 15.1 shows a range of grain sizes of soils that can be densified by the vibro compaction method.

There are a few methods of deep compaction. Among these, (i) dynamic compaction, (ii) vibroflotation, and (iii) Muller resonance compaction are the methods most commonly used in densification of granular soils. Before carrying out the deep compaction works, the extent of densification required must first be decided. This required degree of densification is based on the bearing capacity and tolerable settlement of the soil.


15.1.1 Degree of densification

The degree of densification required may be decided based on the magnitude of bearing capacity required and the extent of tolerable settlement. The bearing capacity of a shallow foundation is usually dependent upon the geometry of the foundation and shear strength parameters, especially the frictional angle of the soil, whereas the magnitude of settlement is inversely proportional to the modulus of the soil. Therefore, in order to increase the bearing capacity, the friction angle needs to be increased whereas to minimize settlement, the modulus of granular soil needs to be increased. However, in practice, it is technically impossible to measure the in-situ friction angle and also difficult and time consuming to measure the modulus of soil at the various levels. Some projects still specify an increase in modulus and this is measured by a pressuremeter. This aspect will be discussed later. Therefore, the required degree of densification is generally given with the relative density, which can correlate well with both the friction angle and modulus of soil. Firstly, an explanation is given below on how to arrive at the required friction angle and modulus to increase the bearing capacity and to minimize settlement.

Generally, relative density is correlated to friction angle, as shown in Figure 15.2, as proposed by Hilf (1991). Therefore, if the required friction angle for certain bearing capacity and type of fill material is known, the

required relative density can be determined. In order to determine the relative density of the soil mass, three values, such as current dry density, and maximum and minimum dry densities are required to be obtained. Although maximum and minimum dry densities can be determined in the laboratory, it is impossible to obtain the in-situ dry density at great depths. It is thus common practice to determine the other in-situ parameters which can be correlated to the relative density. Such in-situ tests include standard penetration test (SPT), and cone penetration test (CPT). Both correlate well with relative density. Correlation between SPT and relative density, proposed in the Bureau of Reclamation’s Earth Manual and Coffman (1960) are shown in Figures 15.3a and 15.3b.

The correlation between CPT and relative density was proposed by Schmertmann (1988), as shown in Figure 15.4. However, Schmertmann has pointed out that the measured cone resistance is highly dependent upon lateral stress in the soil, and cone resistance increases with lateral stress increase. Lateral stress also increases with the over-consolidation ratio of granular soil. Therefore, in order to correctly obtain the equivalent cone resistance values of certain relative densities after compaction, correlation, as proposed by Schmertmann, for over-consolidated granular soil needs to be used. Based on the above proposed correlation, the required CPT value is decided to obtain the necessary relative density (Figure 15.5).

15.1.2 Required depth of compaction

The second step is to determine the depth or thickness of the profile that needs to be densified. If there are no expected seismic or dynamic forces that can cause liquefaction in the future, the granular soil mass is densified to increase the bearing capacity and to reduce the elastic settlement. The densification depth should also be determined based on pressure bulb calculation for particular types and geometry of foundation (Figure 15.6). The soil mass beyond the stress influence zone may not need to be compacted. If a liquefaction problem exists, the whole profile of granular soil needs to be compacted.


Dynamic compaction (DC) is a technique for improving the mechanical properties of soil to relatively greater depths by repeatedly lifting and dropping a heavy weight (pounder) onto the ground surface. It generates repeated impacts at short intervals when the heavy weight hits the ground surface. The force of the impact rearranges the soil particles into a denser state. The selection of spacing of the impacts and the number of drops per print point is important to achieve the specified density of DC.

The tamping process is usually repeated in several phases until the requisite post-treatment in-situ strengths have been achieved. The initial spacing of the impacts should usually be equal to the thickness of the densifiable layer in order to compact the lower part of the layer to be improved. Subsequent phases tend to have progressively closer spacing. After each phase, the craters created by the dropping pounder are usually backfilled with the surrounding materials before the next phase. Finally, an “ironing” phase with low energy is carried out to compact the surface layer. There is no further benefit from continued tamping on the same spot after closure of the soil voids.

The basic mechanism underlying dynamic compaction in granular soils is relatively well understood. When the pounder impacts the ground surface, the force of the impact is transformed into seismic radiation transmitting to the underlying soil mass. At the moment of impact by the pounder, the force is transformed into body waves that consist of compression waves and shear waves. Surface waves are also created in the soil. Whilst the waves propagate radially outward from the source along a hemispherical wave front, as shown in Figure 15.7, surface waves propagate horizontally along the surface. The role of these shock waves is dependent on the soil types and the degree of saturation. For dry deposits, the impact induces compressive and shear waves to overcome the interlocking stresses within the loose strata, resulting in a reduction of the voids. The mechanism of densification for saturated granular deposits is quite different. The compressive stresses induced by the DC impact result in a sudden increase in pore water pressure, thereby forcing the soil into a state of liquefaction. The shear waves and Rayleigh waves, which are slower, travel through the solid skeleton. The combination of temporary loss of contact stresses and dynamic oscillation causes the soil particles to be rearranged into a dense state.

The pounders are usually square or circular in shape and made of steel or concrete. Their weights normally range from 5 to 40 tons and drop

heights of up to 25 m have been used frequently. Owing to the lack of proper understanding of the mechanism of dynamic compaction, most dynamic compaction projects are carried out by specialist contractors with the help of a trial compaction. Presently, a rational or well-established design procedure is not available. The common understanding is that the degree of improvement increases with the energy applied and influence depth increases with increases in pounder weight and drop height.

15.2.1 Influence depth

Menard and Broise (1975) initiated and proposed a formula which estimates the influence depth with dynamic compaction as follows:

where w is the weight of the pounder in ton, and h is the height of drop in meter. D is energy per drop in ton-meters. A more appropriate and accepted equation is given as:

where n is an empirical coefficient factor varying from 0.3 to 1.0.

The effectiveness of dynamic compaction is strongly affected by the soil condition as well as by the energy configuration.

The results of the numerous tests in Changi suggest that the “n” factors for energy, weight and drop height vary from 0.33 to 0.44. It is worth noting that the depth of influence is also dependent upon the size and shape of the

pounder. The same weight of pounder with the same energy can give different influence depths if the geometry of the pounder is different. It can be seen in Table 15.1 that different “n” values are obtained from the tests using different pounders.

Van Impe (1992) also pointed out that the depth of influence depends upon the surface area and the shape of the pounder. Lukas (1986) has stated that multi-tamping only improves the zone of influence and not the depth of influence. Mayne et al. (1984) has suggested that the degree of soil improvement by dynamic compaction peaks at a critical depth, which is roughly one half of the maximum depth of influence. The numerous test data from Changi support the above concept (Na 2002).

From suggested correlations, an estimation of influence depth can be made and the required pounder weight and height of crane can be selected to meet the required depth of compaction. The effectiveness of dynamic compaction is dependent on the combination of the weight, geometry of pounder, height of drop, spacing, number of drops, and total compactive energy applied. Details of the equipment and the energy produced, together with the achieved densification in dynamic compaction work carried out at Changi East, are presented in Table 15.2 (Na 2002).

Table 15.1 “n” value for various pounders (after Na 2002).
Pounder Weight (tons)15142323
Drop Height (m)202012.525
Pounder Surface Area (m2)3.872.255.55.5
Energy (ton.m)300280287.5575
Influence Depth (m7.5768
“n” Factor0.4330.4180.3540.334
Table 15.2 Details of dynamic compaction (after Na 2002).
Pounder Weight (t)23151818
Drop Height (m)25202424
No. of Drops per pass5101012
Energy per drop (t)575300432432
Spacing at each pass (m x m)6 x 66 x 68.5 x 8.510 x 10
No. of passes2222
Effective surface area (m2)5.53.873.43.4
Energy per m2 (ton-m/m2)160166120105
Compacted depth (m7777
Cone resistance achieved (MPa)15151212

15.2.2 Shape of pounder

Pounders are usually square in shape but some pounders are hexagonal. The thickness of the pounder also varies. Some pounders have foot studs which use bolts and nuts to hold the steel plates together. Pounders are usually made up of steel plates, but a few are made up of concrete blocks. Figure 15.8 shows various types of pounders used in dynamic compaction works. The pounder with a smaller base area will penetrate deeper than a pounder with a bigger base area. This creates depth and settlement, which will be discussed later.

15.2.3 Lifting and dropping mechanism

Lifting is usually achieved by a winch system. A high capacity crane with various boom lengths are used in DC works. However, if the pounder is too heavy, a tripod is used instead of a crane. Drop points are just in front of the crane, and if a tripod is used, the drop points are at the center of the tripod. Figure 15.9 shows various types of cranes and tripod used in dynamic compaction works. In most cases, when the pounder is released, the cable follows the pounder. Therefore, there is significant friction between the pulley and the cable. As such, there is some energy loss due to the friction. This is taken into account in the empirical coefficient described in Equation 15.2.

There is a system which can drop the pounder in a free fall without energy loss due to friction. This is a clip holder, as shown in Figure 15.10. Figure 15.11 shows various types of pounders used in dynamic compaction works. However, even with free fall there will still be energy loss due to friction caused by air.

15.2.4 Field measurement during dynamic compaction

Several measurements were made at the Changi reclamation project to check the energy loss, displacement, stress and pore pressure due to dynamic compaction. Figure 15.12 shows a comparison of theoretical velocity and measured velocity of a pounder at the time when the pounder touches the surface. It can be seen that the measured velocity is only about 80% of the theoretical velocity. The impact of an 18-ton pounder dropped from a 10-meter height is about 1200 kPa. By integrating the calculated velocity, the displacement can be estimated (Figure 15.12). Figure 15.13 shows a comparison of measured and estimated pounder displacement, and they are found to be comparable (Na et al. 1997).

This mean displacement of the pounder—in other words, the crater depth—can be pre-estimated. A typical relationship between the peak of impact deceleration and drop number is shown in Figure 15.14, for an 18-ton pounder dropped from a 10-meter height for various numbers of drops. It was found that at the peak, three drops were made and after that deceleration reduced it to six drops, which then remained constant. This means after six drops the crater depth almost no longer increases. Pore water pressure in the ground mass at 2 meters and 3 meters away from the pounding point and at 5 meter depth was measured at the trial test. It was found that a piezometer at 2 meters away measured excess pore pressure of about 140 kPa, and 3 meters away measured only 60 kPa (Figure 15.15).

This pore pressure is almost tenfold less than the stress occurring at the surface. The excess pore pressure peaks at only about 0.2 seconds and within 2 – 3 minutes all pore pressure dissipates. Therefore, for dynamic compaction in granular soil, excess pore pressure may not be a major issue. The excess pore pressure was measured after every drop of the pounder and it was found that excess pore pressure increased with the number of drops. This could be due to an increase in the densification of the soil after each drop.

15.2.5 Degree of improvement

To design densification work, the selection of spacing and the number of drops per point is crucial to achieve the specified density requirement. In other words, the selection of total energy per unit surface area with the type of pounder and drop height is important to achieve the specified density. Leonards et al. (1980) has suggested that the degree of compaction correlates best with the product of the total energy applied per unit surface area times the energy per drop. The test results at Changi East shown in Figure 15.16, support his concept but the upper limit of the maximum attainable cone resistance may be about qc~ 180 kg/cm2 (18 MPa) for the soil type, as shown in Figure 15.17.

In the Changi East project, to attain maximum cone resistance of 18 MPa (180 kg/cm2) and 12 MPa (120 kg/cm2), the energy per unit area multiplied by the energy per drop of 92,000 and 48,900 square tons respectively were applied. From the above correlation, the effective spacing of the pounding point and the number of drops per point can be designed for particular pounders and cranes, as shown in Tables 15.3 and 15.4. To achieve the selected effective spacing, the sequence of pounding can be arranged into two or more phases to allow for pore water pressure dissipation between phases. Normally, spacing can be around 5 – 7 meters. In the Changi East project, two phases of 6 by 6 meters square spacing using two different energies per drop, with 5 and 10 drops, were applied to achieve the required cone resistance of 15 and 18 MPa respectively. The resulting cone resistance after compaction is shown together with pre-CPT in Figures 15.18 and 15.19 for methods A and B respectively.

Table 15.3 Calculation of required spacing.
1Required CPT qc (MPa)1815
2Pounder Weight (tons)2315
3Drop Height (m)2520
4 (2*3)Available Energy per Drop (ton-meter)575300
5 (Figure 1)Energy/m2*Energy/Drop (ton2)92,00048,900
6 (5/4)Required Energy/m2 (ton-m/m2)160163
7 (4/6)Effective Area/Drop (m2)3.591.84
Table 15.4 Calculation of required number of drops.
1First Pass Effective (m2)3636
2Second Pass Effective (m2)3636
3Net Effective Area (m2)17.9717.97
4 (Table 2)Effective Area/Drop (m2)3.591.84
5 (3/4)Required No. of Drops/Phase510

15.2.6 Normalized crater depth

Mayne et al. (1984) also suggested another useful correlation between normalized crater depth, Dc/ (wh)1/2 and the number of drops.

The authors have also carried out such correlations on various types of pounders and drop heights under various soil conditions. It is noted that for the same pounder and the same initial soil condition, the trend of normalized crater depth versus number of drops is the same although the drop height varies. This can be seen in Figures 15.20 and 15.21.

However, for the same drop weight with the same pounder and same drop height, the trend of normalized crater depth can vary if the initial soil

condition is different. The different trends of normalized crater depth measured after the first and second phases are shown in Figure 15.22. The crater depths are shallower with the same number of blows in the second phase since the soil has been densified to a certain degree in the first phase of pounding.

If the geometry of the pounder is different, the trend of normalized crater depth can vary even if a similar weight of pounder and the same drop height are applied in the same soil type. This can be seen in Figure 15.23. This information is very useful for site supervision since the trend of normalized crater depths is the same for the same pounder.

As such, supervision can be kept to a minimum and the crater depth can be measured after pounding. If a specified number of drops are applied, the same crater depth will be measured for the same pounder with the same drop height. As such, the physical counting of the number of drops and the observation of drop heights may not be required after trial tests have been carried out on the particular type of soil.

15.2.7 Most compacted point

In most specifications, the test location for acceptance of compaction works is specified at the centroid point based on the assumption that the centroid location is the least compacted point. However, if the correct spacing is used, it has been found that the centroid point is in fact the most compacted point, and the location under the pounder is actually the least compacted point.

The authors have carried out a large number of cone penetration tests around and under the pounder, for 6 meter by 6 meter square grid spacing for two phases of pounding and 7 meter by 7 meter square grid spacing for two phases of pounding. The comparative CPT data can be seen in Figures 15.24 and 15.25. It was confirmed that the location under the pounder was the least compacted although at about 2 – 3 meters depth of sandfill there was a thin layer that was highly compacted. The centroid point was found to be the most homogeneously compacted location. Therefore, it is suggested that the location of the post-CPT point for acceptance of compaction work

would be under the pounder for dynamic compaction works. Na (2002) has produced a contour of densification under the print and centroid point from several tests he carried out (Figure 15.26). It can be seen that for a single pounding the degree of densification reduced with the distance from the pounding point. When pounding is done with the correct grid pattern, the centroid point is more compacted.

15.2.8 Aging effect

Dynamic compaction is carried out in phases to allow for pore pressure dissipation during the pause period. However, granular soil is highly permeable and dissipation of excess pore pressure is quite rapid, as explained in the earlier section. In addition to this, owing to the soil fissuring during pounding, pore water pressure dissipation is very rapid. Therefore, no significant aging can be seen after compaction. Negligible increases in cone resistance values after compaction may be due to the slow redeposition of soluble silica at grain contact, which acts as natural cementation. A comparison of CPT testing results immediately and three months after compaction can be seen in Figure 15.27.


MRC does not require water for penetration. In this method, a steady-state vibrator is used to densify the soil. As a result of vibration, the friction between the soil particles is temporarily reduced. This facilitates the rearrangement of particles, resulting in densification of the soil. A specially designed steel probe is attached to a vibrator which has variable operating frequencies. The frequency is adjusted to the resonance frequency of the soil, resulting in strongly amplified ground vibrations and thereby efficient soil densification is achieved.

Generally, a higher capacity vibrator (MS-200) requires wider spacing, whereas a lower capacity vibrator (MS-100) requires narrower spacing. The achieved cone resistance at various distances from the probe point in the MRC is shown in Figure 15.28 (Choa et al. 1997a). As can be seen in the figure, the densification achieved is significant in the bottom part of the profile. However, the top part of the profile seems to have been densified by seepage forces before compaction.

15.3.1 Equipment

Two main types of MRC vibrators used for Muller resonance compaction are the MS-100 and MS-200 vibrators. The MS-100 vibrator has a maximum static moment of 1000Nm while the MS-200 vibrator has a maximum static moment of 1900 Nm. Details of the dimensions and specifications are shown in Table 15.5. The probe profile is a “wing” of double Y-shaped flexible plates with openings. The usual shape of the probe is shown in Figure 15.29. The length of the “wing” as well as the size of the opening can vary depending upon the soil condition.

15.3.2 Procedure of compaction

The procedure of compaction is such that the probe is inserted into the ground at a high frequency in order to reduce the soil resistance along the shaft and the toe. Usually during penetration, a frequency of 23 to 25 Hz is used. When the probe reaches the required depth, the frequency is adjusted to the resonance frequency of the soil layers, thereby amplifying the ground response. Normally, the natural frequency of uncompacted soil ranges between 12 and 15 Hz. The natural frequency of the soil can be found by spectral analysis (Massarsch 1991). This analysis was carried out at the Changi site and the soil natural frequency was found to be about 12 Hz for uncompacted sand. This is shown in Figure 15.30.

Table 15.5 Dimensions and specifications of the Muller vibrator.
Muller VibratorMS-100HFMS-200H
Max. Centrifugal Force (kN)25004000
Max. Static Moment (Nm)10001900
High Frequency Step (Nm)480
Max. Oscillation Frequency (rpm)21561500
Total Weight without grip (kg)1090015500
Dyn. Weight without grip (kg)770011750
Oscillation Amplitude (mm)2634
Max. Pulling Power (kN)600800
Vibrator Height (m)3.2353.655
Vibrator Width (m)0.661.352
Static Mass Width (m)2.412.3
Static Mass Thickness (m)0.60.75

The MRC probe is executed in a vertical direction and the vibration energy is transmitted to the surrounding soil along the entire length of the probe. When resonance is achieved, the whole soil layer will oscillate simultaneously and this is an important advantage compared to other vibratory methods. The compaction duration depends on the soil properties and the required extent of densification to be achieved.

Normally the speed of vibration required is two minutes per meter. Compaction is usually carried out in a square grid pattern of two or more phases. The square grid spacing typically ranges between 3 meters and 5.5 meters. In subsequent phases, the compaction is carried out between the compaction points in the first phase. Figure 15.31 shows MRC in progress. Figure 15.32 shows the type of MRC probe.

15.3.3 Monitoring of performance

The MRC system has a comprehensive monitoring unit. From the vibrator and rig, the vibrator frequency, oil pressure and depth of probe can be recorded throughout the operation. In addition, triaxial geophones are used to measure radial, vertical, and tangential ground vibration velocity. An example of monitoring data is shown in Figure 15.33. Quality control of the compaction can be based on the monitoring data recorded by an automatic recording system for each and every compaction point.

15.3.4 Most compacted point

Like vibroflotation, the degree of compaction for the MRC type of system is also largely dependent upon the vibrator, the spacing of the probe, the duration of compaction, and the applied frequency. With closer spacing, better ground densification can be achieved. Compared to the vibroflotation method, the MRC method compacts the soil mass more homogeneously and the variation of cone resistance with distance from the probe does not vary significantly. CPT test results carried out after compaction along the axis of the two diagonally compacted points are shown in Figure 15.28.

15.3.5 Aging effect

There is an aging effect after compaction but this is not as significant as vibroflotation compaction because of the residual excess pore pressure soon after compaction. This is shown in Figure 15.34.


Another method used for densification of granular soil is vibroflotation. It is a technique designed to induce compaction of granular materials at depth. The basic principle behind the process is that particles of non-cohesive soils will be rearranged into a denser configuration by means of horizontal vibrations induced by the depth vibrator. For non-cohesive soils with natural dry densities less than the maximum dry density, the influence of vibrations will result in a rearrangement of their grain structure. A schematic of the vibroflotation technique is shown in Figure 15.35.

As a result of the vibroflotation process, the void ratio and compressibility of the treated soil will decrease and the angle of shearing resistance increase. The treated compacted soil is capable of sustaining higher bearing pressures compared to untreated soil.

15.4.1 Onshore vibroflotation method

The essential equipment for the vibroflotation process is a vibrator, a long heavy tube enclosed with eccentric weight and either electrically or hydraulically driven. The motion of eccentric weight inside the vibrator induces effective horizontal vibrations. The motion of eccentric weight and the resulting horizontal vibrations are schematically represented in Figure 15.36.

The vibrator is connected to a power source and a high-pressure water pump. Extension tubes are added as necessary, depending on the treatment depth, and the whole assembly is suspended from a crane of suitable capacity. With the power source and water supply switched on, the vibrator is lowered into the ground. The combination of vibration and high-pressure water jetting causes liquefaction of the soils surrounding the vibrator, which assists in the penetration process. When the required depth is reached, the water pressure is reduced and the vibrator pulled back in short steps. With the inter-particle friction temporarily reduced, the surrounding soils then fall back below the vibrator and, assisted by vibration, are rearranged into a denser state. This process is repeated up to the ground level, leaving on

completion, a column of well compacted dense material surrounded by material of enhanced density. The whole process is schematically shown in Figure 15.37.

The degree of improvement in compaction achieved depends on the soil being treated (the grain shape and size, composition and percentage of fine soil), the amount of time spent at each stage of compaction, the distance

from the probe point, and the effect of vibration. Typically, the zone of influence will have a diameter between 3 and 4 meters. The spacing of the probes is designed to ensure that the zones of influence overlap sufficiently to achieve minimum requirements throughout the treated area. Generally, the effect of the compaction becomes visible at the ground surface in the form of a cone-shaped depression, as shown in Figure 15.38. The depression formed around the vibrator or the extension tubes is continually infilled with granular materials, which is either imported or obtained from the natural granular deposits at the site.

15.4.2 Offshore vibroflotation method

In the offshore method, a single or multi-vibro set-up is used to compact the sandkey formation under marine conditions. A barge or pontoon is required to serve as a working platform on which a crawler crane of sufficient capacity is mounted to support the vibro string assembly. The whole method is the same as the onshore vibroflotation. A schematic diagram of a typical set-up for offshore vibroflotation is shown in Figure 15.39. After shifting the barge to the treatment zone, the exact positioning of the vibrator to the each probe point is done by a crane using the data constantly provided by a GPS (global positioning system) receiver mounted at the top of the vibro string.

15.4.3 Equipment

Several types of vibroflotation equipment with slightly different specifications can be used, driven either electrically or hydraulically. The specifications of the different types of vibroflotation equipment used are shown in Table 15.6. Figure 15.40 shows a photograph of the equipment. Among the equipment listed in the table, the power rating of vibroflotation is lower and that of pennine is higher. The V23 and V32 types of vibroflotation equipment use the same power rating but different centrifugal forces. Amplitudes range between 23 and 32 mm, and the model numbers signify their amplitudes.

Keller (S-300) uses a higher power rating, but low centrifugal force and amplitude. Pennine has a high centrifugal force and amplitude and its dimensions are all the largest of the three plant types. In general, the equipment is 3 to 3.5 meters in length and has a 350 to 400 mm diameter

Table 15.6 Vibroflotation equipment specifications.
Power Rating (kW)130130130150
Speed of Rotation (rpm)1800180018001775
Rated Current (Amp)300300300300
Centrifugal Force (kN)280330450290
Amplitude (mm)23283225
Vibrator Diameter (mm)350360350400
Vibrator Length (m)
Vibrator Weight (kN)22252524.5

vibrating poker with a vibrating electric motor inside. The power rating of the vibrator ranges between 130 and 150 kW. The vibroflotation equipment can compact up to 30 meters in depth.

15.4.4 Procedure of compaction

The procedure of compaction is to use a vibrator probe to penetrate the ground with the aid of tip-water jetting. On reaching the required depth, the tip-water jets are shut off and vibration started with side water jetting. Usually the duration of vibration varies from 30 seconds to one minute, and the current required varies from 120 to 260 amperes, both of which depend upon the required density, initial soil condition, and type of vibrator used. The vibrator is raised up about 0.5 meters when the criterion is satisfied, and the process repeated. During the compaction process, additional granular soil on the surface and the collapsed soil layer caused by side water jetting backfills the cavity in the cylinder. Based on the author’s experience, to achieve 10, 12 and 15 MPa of cone resistance with V32 type of vibrator, a corresponding amperage of 160 amp, 240 amp and 260 amp is required respectively.

Sometimes difficulties arise in compacting deep portions of the soil if there is an existing dense layer in the upper part of the soil profile. In such cases, side water jetting may not be able to loosen the upper part of the dense soil layer and hence the bottom part of the hole will be left unfilled by the additional soil. As such, the bottom part of the soil may not be well compacted. Compaction in such a soil profile can be carried out by increasing the capacity of side water jetting or by repeated side water jetting onto the wall of the hole by up and down penetrations to loosen the hard layer.

15.4.5 Quality control

The monitoring of each probe point is performed by an automatic recording device. This instrument yields a computer record of the installation process in a continuous graphical mode, plotting depth versus time, and power consumption versus time. The information provided by the recording device also includes the probe reference number, date and period of execution, maximum depth, and maximum power consumption. These records are the main quality control tools during the compaction process. One such typical record is shown in Figure 15.41.

15.4.6 Design procedure

The purpose of vibroflotation is the densification of the existing soil. The feasibility of the technique depends mainly on the grain size distribution of the soil. The range of soil types treatable by vibroflotation is shown in Figure 15.1.

The degree of improvement will depend on many factors including soil conditions, type of equipment, procedures adopted, and skill of the site staff. Such variables do not permit an optimum design to be established in advance but rather require the exercise of experience and judgment for

their successful resolution. For large projects, it is preferable and advisable to conduct trials with varying probe spacing in order to determine the optimal spacing of probes to achieve the minimum required specifications at the probable weakest points.

15.4.7 Monitoring of performance

To assess the performance achieved by vibroflotation, post-cone penetration tests are conducted at the weakest points (typically at the centroid of the grid pattern) to check the achieved tip resistance with depth in comparison to the specifications. In general, these tests are conducted seven days after the vibroflotation works to allow the dissipation of excess pore water pressures developed during compaction works. As a result of densification by vibroflotation, the expected settlements are in the range of about 10% of treatment depth.

15.4.8 Achievable cone resistance and most compacted point

The spacing of probe points will differ for different types of equipment to achieve the same densification requirements, as shown in Table 15.7. It can be seen in the table that, for the same vibrator, wider spacing produces lower cone resistance. A higher capacity vibrator can achieve the same degree of densification with wider spacing compared with a smaller capacity vibrator. The variation of cone resistance with distance from the probe point after compaction is shown in Figure 15.42. It can be seen in the figure that cone resistance is highest at the probe point and lowest near the centroid point of the triangle (CPT point 2 and 4). However, cone resistance at the centroid of the four compaction points is higher than that at points 2 and 4.

The degree of compaction is largely dependent upon the type of equipment, spacing of probes, duration of compaction, and the magnitude

Table 15.7 Details of vibroflotation equipment and spacing, together with achieved cone resistance.
Serial No.Type of EquipmentModelSpacingCone Resistance Achieved (MPa)
1VibroflotationV233 plus roller compaction10

of amperage achieved. The closer the spacing, the greater the possibility of densifying the whole mass of soil. If spacing is wider than required, some loose profile can be found at the centroid point. Based on the author’s experience with type of soil shown in Figure 15.12, a triangular grid spacing of 2.5 to 3.0 meters is required to achieve a cone resistance of 15 MPa by a V32 type of vibroflotation equipment. To achieve a cone resistance of 12 MPa, a triangular grid spacing of 3.0 to 3.2 meters is required with the V32 vibrator. For the S300 vibrator, a triangular grid spacing of between 2.4 to 2.6 meters is required to achieve a cone resistance of 15 MPa. However, this is also dependent upon the initial soil condition.

CPT testing results after compaction carried out along the axis of two far end vibro probe locations are shown in Figure 15.42 for the V32 vibrator. It was found that the cone resistance decreases with distance from the probe point.

15.4.9 Aging effect

No fissuring occurs during vibroflotation, but because of the application of additional water pressure and pore water pressure, dissipation takes a longer period than for the dynamic compaction process. Therefore, the aging effect is quite significant for vibroflotation. The significant increase in cone resistance four months after compaction can be seen in Figure 15.43.

15.4.10 Applications

Typical applications of vibroflotation include individual footings and large reclamation areas. Typical plan views for these applications with an appropriate grid pattern are schematically shown in Figure 15.44.


The vibroflotation technique has been proven to be an effective and economic method of improving the loose granular non-cohesive soils for a wide range of applications. By compacting these soils using the vibroflotation technique, the density will increase and improve the bearing capacity of the ground. The technique also helps in improving the shear strength and compressibility characteristics of granular soils and result in reduced total and differential foundation settlements.

About this article

Densification of Granular Soil

Updated About content Print Article