CHAPTER 5: SETTLEMENT ANALYSIS AND RESULTS
5.2 Settlement analyses using nonlinear models
5.2.3 Analyses for PF groups at SAMSE project
Axial load (kPa)
0 100 200 300 400 500 600
Settlement (mm)
0
5
10
15
PF group (Equivalent model) CDM group (Equivalent model) Static load test
Single PF column
64
Table 5.4 Material model and parameter used for SAMSE factory project.
Parameter Unit Fill layer Soft soil
clay Medium
clay PF/CDM Steel plate
Model MC SS SS MC Elastic
Material behavior Drained Undrained Undrained Undrained d=0.15 m
Elastic modulus, Es’ kPa 30000 390,000 1012
Unit weigh, kN/m3 18.0 17.82 17.82 23.0 78.5
v’ - 0.25 - - 0.25 0.2
Cc - - 0.15 0.15 - -
Cs - - 0.03 0.03 - -
einit - - 0.86 0.86 - -
POP kPa - 30 0 - -
E. cohesion, c’ kPa 5 10 10 1300 -
E. friction angle, ’ degree 30 27 27 - -
Note: MC= Mohr-Coulomb; SS= soft soil
Within the improved zone, PF column have three parts: Head of PF, cone of PF and tail of PF. Similarly, the improved zone have three parts: Head of the improve zone, cone of the improved zone and tail of the improved zone. Using equation (3.5), the parameters of stiffness (Es), effective cohesion (c’), effective friction angle (’) change by depth (m). In this case, each footing has three columns, thus it is quite complicated for equivalent method. In the model, the size of of the plate B L = 2.4 2.4 m was used to fully cover the columns. Table 5.4 shows material models and input parameters for Samse project.
It is important to note that settlement of PF group 1 (PF length = 8.5 m) is not presented in this section. This is because the measured settlement values of the group are too small to match with any constitutive model used. More time is needed to investigate the settlement of this group. Thus, in the following sections, comparisons are made for the group 2 and group 3 at the site.
Settlement analysis
In this cases, the PF columns and soil under the footing are modelled using two approaches:
(i) An equivalent material; (ii) A true 3D model of PF columns and soil.
(i) An equivalent model:
65
The effectiveness when the treated zone under the footing is modelled as an equivalent material (with secant modulus for PF/CDM columns: E50=150qu). Figure 5.15 (a) and (b) show load-settlement curves of the simulated footings on PF columns, CDM columns and from the static load test. Two distinct points from the comparison are:
Settlement of the footing on PF columns and CDM columns is almost equal.
Generally, the difference in settlement of footings on PF columns and static load test is not much (less than 10%), this value is accepted for numerical analysis.
(a) Group 02 (LPF=6 m)
(b) Group 03 (LPF=4 m)
Figure 5.15 Load settlement curves from Numerical method for PF columns and CDM columns using equivalent material (E50=150qu)
Axial load (kPa)
0 100 200 300 400 500 600 700 800
Settlement (mm)
0 20 40 60 80 100 120
PF group (Equivalent model) CDM group (Equivalent model) Static load test
Samse project group 02
Axial load (kPa)
0 100 200 300 400 500 600
Settlement (mm)
0 20 40 60 80 100 120
PF group (Equivalent model) CDM group (Equivalent model) Static load test
Samse project group 03
66
(a) Group 02 (LPF=6 m)
(b) Group 03 (LPF=4 m)
Figure 5.16 Load settlement curves from Numerical method for PF groups and CDM groups using true 3D model of PF columns and soil
(ii) A true 3D model of PF columns and soil.
Figure 5.16 (a) and (b) show a comparison of load- settlement curves otbained from simulated footings on PF and CDM columns using both equivalent material and true 3D approaches (with secant modulus for PF/CDM columns: E50 =150qu kPa) and from the Some key points from the results are as follows:
Axial load (kPa)
0 100 200 300 400 500 600 700 800
Settlement (mm)
0 20 40 60 80 100 120
PF group (Equivalent model) PF group (3D model) CDM group (Equivalent model) CDM group (3D model) Static load test
Samse project group 02
Axial load (kPa)
0 100 200 300 400 500 600
Settlement (mm)
0 20 40 60 80 100 120
PF group (Equivalent model) PF group (3D model) CDM group (Equivalent model) CDM group (3D model) Static load test
Samse project group 03
67
Generally, settlement of simulated footing on PF columns is also close to settlement of the footing from static load test. This result can be acceptable.
Settlement of footings on PF (or CDM) columns using equivalent model is slightly smaller than settlement of footings on PF (or CDM) columns using the true 3D model.
The difference might be attributed to some reasons such as shape of the columns, foundation size (steel plate), soil characteristics, etc. However, the settlement discrepancy is unnoticeable when the load-settlement curves in the relatively linear range. This means that under service loading conditions, in which settlement curve is almost linear, both models result in rather similar settlement values
Settlement of the footings on CDM columns is smaller than settlement of footings on PF columns for both equivalent material and true 3D models.
For these two experimental PF groups, why is the settlement of footings on CDM columns smaller than settlement of footings on PF columns? This result is against previous finds that the settlement of footings on PF columns is generally smaller settlement of footings on CDM columns.
Possible reasons and proof
As discussed above, there are 2 possible reasons that make settlement of the footings on PF columns smaller than settlement of the footings on CDM columns:
Reason 1: The PF columns shape is not the optimal shape. In the ideal case and JEF case, there exist a minimum settlement value at certain value in the range of 0.4 to 0.5, but in SAMSE project: (i) =Lh/L=1/6=0.16 < 0.4 for PF columns of group 02 (LPF=6 m); (ii)
= Lh/L = 1/4 =0.25 < 0.4 for PF columns of group 03 (LPF = 4 m). These values do not make a difference in settlement of shallow footings on PF columns and CDM columns.
Reason 2: The settlement consists parts: Settlement of improved zone and settlement of layers under improved zone. The main difference in stiffness between columns and the soil is so high, therefore, the settlement of improved zone is very small compared with the total settlement.
68
Figs. 5.17(a) and 5.17(b) show the settlement profiles with depth of the footings on PF groups 2 and 3. As shown, for each group, settlement profiles from footings on PF and CDM columns are almost identical.
(a) Group 02 (b) Group 03 (c) Optimal shape Figure 5.17 Settlement profiles with depth of footings on PF and CDM columns from
numerical method using equivalent material model (q = 800 kPa)
Figure 5.18 Load settlement curves from Numerical method for PF columns and conventional CDM columns (Optimal shape design for PF columns)
Settlement (mm) 0 20 40 60 80 100
Depth (m)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
columnsCDM columnsPF Samse project (Group 02)
Settlement (mm) 0 50 100 150 200
Depth (m)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
CDM columns PF columns Samse project (Group 03)
Settlement (mm) 0 50 100 150 200
Depth (m)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
CDM columns PF columns Samse project (Optimal shape)
Axial load (kPa)
0 100 200 300 400 500 600 700 800
Settlement (mm)
0
50
100
150
200
PF group CDM group
Samse project
Optimal shape design for PF group
69
(a) PF group and CDM group in Samse project
(b) Group 02
(c) Group 03
Figure 5.19 Variation of settlement () and effective vertical stress (v) at the toe of CDM and PF columns obtained from numerical analysis using true 3D model
'v,PF (v,PF) 'v,CDM (v,CDM)
q (kPa)
Settlement (mm)
0 20 40 60 80
Axial load (kPa)
0 100 200 300 400 500 600 700 800
PF group CDM group
Shallow footing Group 02
Effective vertical stress at toe (kPa)
0 100 200 300 400
Axial load (kPa)
0 100 200 300 400 500 600 700 800
PF group CDM group
Shallow footing Group 02
Settlement (mm)
0 20 40 60 80 100
Axial load (kPa)
0 100 200 300 400 500 600
PF group CDM group
Shallow footing Group 03
Effective vertical stress at toe (kPa)
0 100 200 300 400 500 600
Axial load (kPa)
0 100 200 300 400 500 600
PF group CDM group
Shallow footing Group 03
70
Fig. 5.17(c) shows the settlement profiles of footings on PF group 3 with two parameters changed: (1) the PF columns were extended from Lh = 1.0 m, Lc = 1.0 m, and Lt = 2.0 m (i.e., = 1/4 = 0.25) to have Lh = 4.0 m, Lc = 1.0 m, and Lt = 1.0 m (i.e., = 4.0/6.0 = 0.66); (2) the stiffness ratio of the columns to soil was changed from Ec/Es = 150 to Ec/Es
= 65 (an average value taking into account the variation of soil stiffness with depth). As shown, the settlement of the footing on PF columns is smaller than that on CDM columns.
The results indicate two important points: (i) the head length should be long enough to cover the upper soft soil layer: (ii) If the stiffness ratio is too high, the settlement of the footings come mainly from the settlement of the soft layers underneath the treated zone.
Note that common stiffness ratio of 10 to 20 are found from many case studies (Kitazume and Terashi 2013).
Fig. 5.18 shows load-settlement curves of footings on PF group 3 extended discussed in the previous graph. It is very clear that the settlement of the footing on PF columns is smaller than that of CDM columns. The results support findings from previous cases.
Settlement () and effective vertical stress (v) at the toe of CDM and PF columns obtained from numerical analysis using true 3D model (Discussed in Fig. 5.16; E50 = 150qu for the PF/CDM columns; using soft soil model) are shown in Fig. 5.19.
As shown, for group 02 (Fig. 5.19(b)) the applied pressure-toe settlement curves as well as applied pressure- toe pressures from PF columns and CDM columns are almost identical.
However, the group 3, the curve show different trend: both settlement and effective vertical stress at toe PF columns’ toe are larger than the values at CDM columns’ toe. This indicates a very important point that soft soil layers under the floating columns play very important in the total settlement of the footing.
71