**4. Unconfined compressive strength tests**

Strength tests have been performed using Sydney sand mixed with two cementing media, gypsum and the bacterial mixture. **Figure 1** shows the responses of the UCS tests of cemented sand performed with gypsum contents in the range of 5–20%. All samples fail at small strains with brittle responses. A clear trend of increasing strength with gypsum content is evident. However, interpretation of these tests requires caution as strength is affected by density, which tends to increase as fines are added; the limited water retention in clean sand, which limits the water available for hydration; and the presence of suctions in the tested samples (dried in laboratory conditions), which tends to increase the strength. All these factors tend to enhance the effectiveness of the cement as more cement is added, at least initially.

Bio-cemented samples of Sydney sand and its UCS strength are shown in **Figure 2**. As expected, as the amount of calcite precipitation increases, the UCS readings also increases. Similarly to the gypsum-cemented samples, the bio-cemented sand samples show generally stiff and brittle behavior, although with more ductility than for the gypsum cement. The results accord with other

**113**

**Figure 2.**

**Test No/ID**

**Table 2.**

*UCS responses for bio-cemented specimens.*

**Test type Average** 

**calcite (%)**

*Variance of calcite distributions in UCS and triaxial test samples.*

**Top (%)**

**Middle (%)**

3%B UCS 1.33 1.52 1.14 1.34 0.19 3.6 B14 Triaxial 1.54 1.49 1.51 1.61 0.10 1.1 5%B UCS 2.73 2.73 2.91 2.68 0.12 1.5 B10 Triaxial 2.61 2.94 2.79 2.79 0.16 2.7 10%B UCS 5.33 5.52 5.21 5.26 0.15 2.3 B13 Triaxial 4.26 4.48 4.17 3.62 0.16 2.5 15%B UCS 6.23 6.13 6.25 6.31 0.15 2.3 B16 Triaxial 6.98 7.09 6.90 6.91 0.11 1.1

**Bottom (%)**

**Standard variance (±%)**

**Sample variance (%)**

*Geomechanical Behavior of Bio-Cemented Sand for Foundation Works*

studies [4, 7, 19], which have reported bio-cemented sand responds similarly to naturally and artificially cemented sand at low confining pressure. Comparison of **Figures 1** and **2** shows that samples with calcite contents less than 3.33% have lower stiffnesses than gypsum-cemented samples of the same strength. With low calcite contents, it is possible that sample heterogeneity influences the results, as reported in other studies [24] where the bio-solution has been pumped into the samples. However, as shown in **Table 2**, the mixing procedure used in this study has produced uniform calcite distributions through the samples, with less than 5% variance in different sample sections, for all calcite contents. End effects may also have affected the apparent stiffness as the sample ends were not prepared perfectly square and there was a tendency for water and bio-solution to flow out of the samples, because of the low water retention, during sample preparation. Nevertheless, the consistent trend with calcite content suggests that the results

**Figure 3** shows a comparison between the UCS responses of gypsum-cemented samples and bio-cemented samples of approximately similar strengths. It is evident that based on strength, also shown in **Figure 4**, the calcite produced by MICP is a much more effective cementing agent than gypsum, with about 0.5% of calcite

reasonably represent bio-cemented sand performance.

*DOI: http://dx.doi.org/10.5772/intechopen.88159*

**Figure 1.** *UCS test responses from gypsum-cemented specimens.*

#### *Geomechanical Behavior of Bio-Cemented Sand for Foundation Works DOI: http://dx.doi.org/10.5772/intechopen.88159*

**Figure 2.** *UCS responses for bio-cemented specimens.*


#### **Table 2.**

*Sandy Materials in Civil Engineering - Usage and Management*

**4. Unconfined compressive strength tests**

to the bio-cemented samples.

before being filled in the molds and gently tamped each time.

1 min before being placed in cylindrical molds. To prepare uniform and reproducible samples with a consistent density, the mixture was divided into five portions

The bio-cemented samples have been compared with gypsum-cemented samples which have been prepared by combining the dry ingredients (sand and unhydrated gypsum) followed by mixing with water and placing in a mold similarly

The preparation technique produced cylindrical samples with 55 x 110 mm in dimensions. After extraction and curing, the samples were either placed directly in a compression machine to perform UCS tests or in a fully computerized triaxial testing apparatus to perform geomechanical tests with elevated confining stresses, which was also fitted with bender elements to monitor the secondary (shear) wave pulse. Once the UCS and triaxial shearing tests were completed, bio-cemented samples were extracted and analyzed to determine the amount and distribution of the calcite precipitated. Further details of methods and procedures are provided by Duraisamy [2].

Strength tests have been performed using Sydney sand mixed with two cementing media, gypsum and the bacterial mixture. **Figure 1** shows the responses of the UCS tests of cemented sand performed with gypsum contents in the range of 5–20%. All samples fail at small strains with brittle responses. A clear trend of increasing strength with gypsum content is evident. However, interpretation of these tests requires caution as strength is affected by density, which tends to increase as fines are added; the limited water retention in clean sand, which limits the water available for hydration; and the presence of suctions in the tested samples (dried in laboratory conditions), which tends to increase the strength. All these factors tend to enhance

the effectiveness of the cement as more cement is added, at least initially.

Bio-cemented samples of Sydney sand and its UCS strength are shown in **Figure 2**. As expected, as the amount of calcite precipitation increases, the UCS readings also increases. Similarly to the gypsum-cemented samples, the bio-cemented sand samples show generally stiff and brittle behavior, although with more ductility than for the gypsum cement. The results accord with other

**112**

**Figure 1.**

*UCS test responses from gypsum-cemented specimens.*

*Variance of calcite distributions in UCS and triaxial test samples.*

studies [4, 7, 19], which have reported bio-cemented sand responds similarly to naturally and artificially cemented sand at low confining pressure. Comparison of **Figures 1** and **2** shows that samples with calcite contents less than 3.33% have lower stiffnesses than gypsum-cemented samples of the same strength. With low calcite contents, it is possible that sample heterogeneity influences the results, as reported in other studies [24] where the bio-solution has been pumped into the samples. However, as shown in **Table 2**, the mixing procedure used in this study has produced uniform calcite distributions through the samples, with less than 5% variance in different sample sections, for all calcite contents. End effects may also have affected the apparent stiffness as the sample ends were not prepared perfectly square and there was a tendency for water and bio-solution to flow out of the samples, because of the low water retention, during sample preparation. Nevertheless, the consistent trend with calcite content suggests that the results reasonably represent bio-cemented sand performance.

**Figure 3** shows a comparison between the UCS responses of gypsum-cemented samples and bio-cemented samples of approximately similar strengths. It is evident that based on strength, also shown in **Figure 4**, the calcite produced by MICP is a much more effective cementing agent than gypsum, with about 0.5% of calcite

#### **Figure 3.**

*Comparison of UCS responses of gypsum and bio-cemented specimens.*

**Figure 4.** *Effects of cement content on UCS strength.*

equivalent to about 5% of gypsum and 9% of calcite equivalent to 20% of gypsum. As noted above the stiffness and ductility of the bio-cemented samples are lower than for gypsum when cement contents are low, and this can in part be explained by the very low amounts of calcite required, and as shown in other studies [13, 27], much of this acts as space filler and does not actively contribute to the strength.

The influence of gypsum and bio-cement content on the strength improvement is shown in **Figure 4**. Previous report [24] also indicates similar UCS and calcite content relationship. However, the results from the current study in which mixing was used all lie above the previous research [24] in which the bio-cement solution was pumped into the samples. It was also reported [24] that preparing samples with low calcite contents by pumping in the solution was problematic as samples tended to have poor homogeneity and these weakly cemented samples tended to deform locally, giving low shear strength, and on occasion to collapse immediately upon loading. In contrast, the low calcite content bio-cemented specimens prepared by mixing in this study all showed significant improvements in resistance. As sands with similar gradings were used and samples were prepared to similar densities in both studies, the different responses point to the sample preparation method as

**115**

**Figure 5.**

*Geomechanical Behavior of Bio-Cemented Sand for Foundation Works*

being the key difference. Several studies [13–27] have shown that regardless of the injection process, obtaining uniform calcite precipitation is difficult, and it is also difficult to control, especially for small amounts of cement. The results from the mixing method of preparation show that this can lead to more homogenous cementation at low calcite contents and mixing can achieve calcite contents of nearly 10%. Nevertheless, the process of injecting bio-cement has advantages. It has been shown to be practical at field scale, and by using a series of injection phases, the process is

The strength and stiffness produced by the mixing technique varies depending on the soil and cement. Even though gypsum was mixed in with the soil, much more gypsum was required to produce the same strength and stiffness as the calcite cement. Because of its acicular particles, gypsum does not easily bridge between the sand particles, and it tends to fill the void spaces. Many studies have shown that in adding silt-sized fines, such as gypsum, fines fill the voids up to a transition fines content of approximately 25% after which the fines have increasing influence on the behavior. Once enough gypsum is present, it will fill the voids and form a strongly cemented matrix. Gypsum contents >15% appear to be needed to achieve this effect

The progress of the cementation process during curing was monitored using bender elements by recording the shear wave velocity change over time. This was achieved during the preparation of samples for triaxial testing. Split mold using PVC material fitted with a rubber membrane inside was designed to produce identical cylindrical samples with length of two times the diameter. Bender elements in the end platens transmitted waves vertically through the samples, and the waveforms and travel times were monitored using a semiautomated procedure [28]. The typical responses in **Figure 5** happen during the curing of gypsum-cemented samples. The hardening process occurs rapidly for gypsum contents above 10% and that curing is essentially complete after 1 h. This is consistent with the setting time reported by the gypsum supplier of 55 minutes. However, for low gypsum contents, there is some variability and longer setting times have been recorded. This is believed to be because

*DOI: http://dx.doi.org/10.5772/intechopen.88159*

capable of achieving very high strengths [27].

**5. Curing process and bender elements**

*Shear wave velocity changes during curing of gypsum-cemented specimens.*

as illustrated in **Figure 4**.

#### *Geomechanical Behavior of Bio-Cemented Sand for Foundation Works DOI: http://dx.doi.org/10.5772/intechopen.88159*

*Sandy Materials in Civil Engineering - Usage and Management*

*Comparison of UCS responses of gypsum and bio-cemented specimens.*

equivalent to about 5% of gypsum and 9% of calcite equivalent to 20% of gypsum. As noted above the stiffness and ductility of the bio-cemented samples are lower than for gypsum when cement contents are low, and this can in part be explained by the very low amounts of calcite required, and as shown in other studies [13, 27], much of this acts as space filler and does not actively contribute to the strength. The influence of gypsum and bio-cement content on the strength improvement is shown in **Figure 4**. Previous report [24] also indicates similar UCS and calcite content relationship. However, the results from the current study in which mixing was used all lie above the previous research [24] in which the bio-cement solution was pumped into the samples. It was also reported [24] that preparing samples with low calcite contents by pumping in the solution was problematic as samples tended to have poor homogeneity and these weakly cemented samples tended to deform locally, giving low shear strength, and on occasion to collapse immediately upon loading. In contrast, the low calcite content bio-cemented specimens prepared by mixing in this study all showed significant improvements in resistance. As sands with similar gradings were used and samples were prepared to similar densities in both studies, the different responses point to the sample preparation method as

**114**

**Figure 3.**

**Figure 4.**

*Effects of cement content on UCS strength.*

being the key difference. Several studies [13–27] have shown that regardless of the injection process, obtaining uniform calcite precipitation is difficult, and it is also difficult to control, especially for small amounts of cement. The results from the mixing method of preparation show that this can lead to more homogenous cementation at low calcite contents and mixing can achieve calcite contents of nearly 10%. Nevertheless, the process of injecting bio-cement has advantages. It has been shown to be practical at field scale, and by using a series of injection phases, the process is capable of achieving very high strengths [27].

The strength and stiffness produced by the mixing technique varies depending on the soil and cement. Even though gypsum was mixed in with the soil, much more gypsum was required to produce the same strength and stiffness as the calcite cement. Because of its acicular particles, gypsum does not easily bridge between the sand particles, and it tends to fill the void spaces. Many studies have shown that in adding silt-sized fines, such as gypsum, fines fill the voids up to a transition fines content of approximately 25% after which the fines have increasing influence on the behavior. Once enough gypsum is present, it will fill the voids and form a strongly cemented matrix. Gypsum contents >15% appear to be needed to achieve this effect as illustrated in **Figure 4**.
