You are on page 1of 14
Boston embankment using SSC and drains Boston Trial Embankment C An application of the Soft Soil Creep model with drains Based on the original excercise made by Pieter Vermeer Martino Leoni Institute for Geotechnical Engincering University of Stutigart Germany C ‘Computational Geotechnies 1 Boston embankment using SSC and drains 2 Computational Geotechnies Boston embankment using SSC and drains INTRODUCTION The construction of a 3.9 km road extension north of Boston started in 1965. An 11 m high embankment was constructed on top of a 41 m thick overconsolidated Boston Blue Clay deposit. The embankment is 28 m wide at the top and founded on a 3 m thick sand layer. The geometry of the embankment is given in Figure 1. It is, recommended to set points A.B and C in the middle of the Boston Blue Clay layers since they will be used for the output. You are asked to simulate the construction of the embankment and calculate the final settlement of the embankment. Figure 1: Cross section of the trial embankment In order to increase stability and decrease construction time drains will be used during the construction stages. In order to model a 3D drain field in a 2D plane strain calculation there are 2 possibilities that. we will consider in this exercise: ‘* Use of Plaxis drain elements in combination with an equivalent horizontal permeability '* Use of equivalent vertical permeability without drain elements. The exercise is a continuation of the Boston Embankment using the Soft Soil Creep model and as such the geometry of that exercise should be used. The easiest way to do this is to read, for every alternative calculation, the project belonging to the Boston Embankment using the Soft Soil Creep model in Plaxis Input and save it under a different: name after which the necessary changes ean be made. gop? ALTERNATIVE 1: DRAIN ELEMENTS Tn order to decrease the calculation time by limiting the development of excess pore pressures strip drains will be applied in a regular grid with 1.5 m in between drains. The stri 0 meters deep. On the left in figure 2 a top view of the layout of the drain field is shown while on the right the top view of the plane strain model is given. In the 2D plane strain model the drains are no longer individual strip drains but form draining walls in the out-of-plane (2-)direction. What is more important is that im the 8D drain field points located on the axis of symmetry between rows of drains have different distances to the closest drain, in figure 2 indicated as drainage lengths D1, D2 and D3. However, in the 2D plane strain model the same points have equal distances to the closest drain: the drainage length D*. This difference in 3D reality and 2D model can be taken into account by using a so-called equivalent horizontal permeability. The horizontal permeability of the soil is chosen so that the results of the 2D plane strain model match the results of the real 3D model. Unfortunately it is not possible to find one equivalent horizontal permeability in order to always match results from the 2D calculation with 3D reality, but only to match results Computational Geotechnies 3 wydte sor» Avan kh pasa Kh tqurcvabaade (Apéodis a) Bie ciea Ts tos # kya Ky Oxnyinad do selo fs Boston embankment using SSC and drains gde Curbe uno afin ko (7 Kay Apa a) € fo) 1 1 | ' ' 1 * D* z I a D2 @ 1 De 1D, 1D 1 x i | 1 7 1 t i ! I ° ' ° ! ( 3D drain field 2D plane strain model Figure 2: Real 3D drain field versus 2D plane strain model for a specific degree of consolidation, Furthermore the equivalent horizontal permeability chosen depends on the ratio of the distance between drains in reality and the distance between the drains in the plane strain model. Best results are obtained if in the plane strain model the distance between the drains is approximately the same as the distance between the drains in reality, but for practical reasons like meshing this may not be feasible. In this project we will put the drains in the plane strain model at an intermediate distance of 4.5 meters. Assuming we're interested in the prediction at the end of consolidation this results in an equivalent horizontal permeability that is 2.7 times larger than the original permeability. In Appendix A the determination of this value is given in ‘more detail. Changing the model ‘The following 3 changes have to be made to the model: Define material sets with equivalent horizontal permeabilities + Define a zone in which these equivalent horizontal permeabilities are used ¢ Introduce drains. ‘The most efficient order is to first define the zone with equivalent horizontal permeabilities and then add the drains. Start the new model by opening the existing geometry used for the Boston Embankment exercise with the use of the Soft Soil Creep model in Plaxis Input and save the project under another name before making any changes. Material sets ‘The drains will pass through the 3 Boston Blue Clay layers as well as the Sand layer that covers the Boston Blue Clay. Hence, for these 4 layers altemative material sets have to be made. These alternative material sets are exact copies from the original material sets but with a different horizontal permeability. Table 1 shows the permeabilities for these soil layers. re 4 Computational Geotechnies Boston embankment using SSC and drains [om/dey] Sand 010 Boston blue day — layer 1 Tis] Boston blue clay - layer 2 5.8010" Boston blue day ~ layer 8 | 157107 | 5.8010-* In order to add the material sets follow these steps © Open the material sets database + For all layers mentioned in table 1 do the following: = Right-click on the material set, a popup-menu appears = Choose the option Copy. A new material set window opens with all parameters filled in according to the original material set, but no nate is specified. ¢ — Give the copy of the material set an appropriate name, for instance the name of the original with “equiv. kh” added to it. — Change the horizontal permeability according to the first column of table 1, Note that the vertical permeability remains unchanged! © Close material sets database. Definition of the zone with equivalent horizontal permeabilities It is assumed that the influence zone of the drains extends up to the intermediate distance of the drains below the toes of the drains as well as beyond the last drain. To make the changes follow these steps: + Draw a geometry line from (x,y)=(0, 81.5) to (61, -31.5) and to (61, 1.5). ‘This borders the zone in which the soil has the equivalent horizontal permeability. © Within the zone formed assign the newly made material sets to the appropriate layers. Henco, the Sand with equivalent horizontal permeability to the soil layer between Om and -2.5m, the Boston Blue Clay 1 layer with equivalent horizontal permeability below that layer and so on. (Drains ‘The main change in the geometry implies the introduction of drain elements under the full width of the embank- rent with 4.5 m interval; this involves the introduction of 11 drains. Hence, drains have to be introduced between coordinates (4.5, 0) and (4.5, -30), (9.0, 0) and (8.0, -30) ete. until the last drain between (49.5, 0) and (49.5, 30). The fastest. way to do this is by manual input of the coordinates: ee * Choose the Drain button (75°) «# ‘Type the coordinates of the top end of the dain with a space in between x- and y-coordinate. Hence, type “4.5 0” without the quotes and press on the keyboard » Type the coordinates of the bottom end of the drain with a space in between x- and y-coordinate. Hence, type “4.5 -20” without the quotes and press on the keyboard. The drain chain is now inserted © Right-click the mouse to end drawing of this drain chain » Continue with step 2. again, but now for the drain with x-coordinate x=0.0. Do this for all drains. Note that no drain is applied at the axis of symmetry of the model; drainage on the axis of symmetry will be (_s0¥ved by using an open flow boundary Computational Geotechnies 5 Boston embankment using SSC and drains Mesh generation The distance in between the drains is relatively small compared to the mesh element sizes and therefore the mesh between the drains will not be very good. The mesh in the area in which the drains are located, which is the area that has the material sets with the equivalent horizontal permeabilities, must be refined. The easiest way to do this is by following these steps ‘© Reset all present mesh refinements by selecting the Reset all option from the Mesh mena ‘© Also from the Mesh menu, select the option Global Coarseness and set it to Fine. However, don't generate the mesh yet but close the window by pressing the OK’ button. © With the key pressed multi-select with the mouse all clusters of Boston Blue Clay material that fall within the zone of equivalent horizontal permeabilities and the 3 clusters that form the embankment. With all these clusters selected, choose the option Refine cluster from the Mesh menu. Figure 3: Generated mesh C 6 "Computational Geotechnies Boston embankment using SSC and drains CALCULATION Continue to the PLAXIS Calculation program, the phase list is the same as for the Boston Embanlanent without drains. However, the drains have been added and are still inactive in all phases. For all phases the drains have to be activated, this can be done by: '» Select phase 1 (Replace peat for fill material) from the phase list Press the Define button to define the staged construction settings # Go to the Water mode of the program # Click on every de dain is grey. (hence 1 times) to activate the drain. An active drain is light blue while an inactive On the axis of symmetry (left side) remove the closed flow boundary between y = 0.0m and y = -31.5m. ‘This can be done by clicking on the line (no change of colour) and then press the button on the Keyboard. When asked, indicate that the Closed boundary should be deleted. + Press Update to save the changes and return to the calculation program. Tn Plaxis offers the possibility for every phase to veset clthen the staged construction settings (or the water conditions 80 that they become equal to the settings and conditions of the phase before. ‘This can be done by right-clicking on the phase one wants to be reset s0 that a popiip menu occurs, From this popup memu choose either Reset staged construction _ to reset. all staged construction settings to the settings of the preceding phase or choose Reset. water conditions to make all water conditions equal to the water conditions of the preceding phase ‘The drains must be activated for all phases. Since drains are part of the water conditions and water conditions are the same for all phases one can use the option to reset water conditions in order to make the water conditions of all successive phases equal to phase 1 and therefore activate the drains in all those phases. This is done by right-clicking every phase one-by-one starting from phase 2 and select the option Reset water conditions. Note ‘that for phase that have not explicitly defined water conditions, typically phases 3, 5, 7 and & that involve only consolidation, this option is not. available. Finally, select a node at the axis of symmetry just below the fill for making a load-displacement curve after the caleulation and press Calculate to start the calculation. OUTPUT Check output results for the End of Construction phase (Phase 6) and the End of Primary Consolidation (Phas 7). Give special attention to the development of excess pore pressures. Further discussion of the results is given after the second alternative calculation Computational Geotechnies 7 Boston embankment using SSC and drains ALTERNATIVE 2: EQUIVALENT VERTICAL PERMEABILITY Instead of using individual drains it is also possible to use a simplified model that implies the definition of a zone of soil with “distributed drains", that is a zone the size of the influence zone of the real drain field with soil that has an adjusted permeability. This adjusted permeability, referred to as the equivalent permeability, is chosen in such a way that the results match the results of a real drain field Jn Alternative 1 the horizontal permeability had to be adjusted to take into account the differences between the real 3D drains and the 2D plane strain model as the main flow direction is horizontal, that is towards the drains. However, in this Alternative we're not going to use any drains and adjust the soil properties in such a way that the modified soil will behave the same in terms of consolidation as the real soil body with drains. In this situation the main flow direction will be vertical (towards adjacent. draining soil layers) and so this approach requires an ‘equivalent vertical permeability. The equivalent vertical permeability, similar to the method with the horizontal equivalent permeability, can be found for just one specific degree of consolidation. Another similarity is that it depends on the ratio of the distance between drains in reality and the distance between the drains in the plane strain model. Additionally, the equivalent vertical permeability depends on the layer thickness of the layer in which it is applied. In Appendix B it is described how this equivalent vertical permeability is exactly derived, but in this exercise the equivalent vertical permeability can be caleulated by Ki = hy + 0.57H?- hy where H is the layer thickness. Note that the equivalent vertical permeability depends on the horizontal perme- ability! For the Boston Embankment problem this leads to the permeabilities given in table 2: ‘Table 2: Permeabilities and equivalent vertical permeability per layer Tayer i Bs Ey ey {on} _|_fm/day]_| [m/day]_|_fim/day] Sand ~ [30 [010 | o.10 061 Boston blue day —layer 1] 9.0_[ 119-10" | T1340" | 539-10" Boston blue clay -layer 2 | 9.5 | 6.80-10°* | 5.8010" Boston bine clay layer 3 | 10.0) 58010" | 80407 Changing the model Only 2 changes have to be made to the model: + Define material sets with equivalent horizontal permeabilities ‘+ Define a zone in which these equivalent horizontal permeabilities are used ‘To do so, start the new model by opening the existing geometry used for the Boston Embankment exercise with the use of the Soft Soil Creep model in Plaxis Input and save the project under another name before making any changes. Material sets In reality the drains will pass through the 3 Boston Blue Clay layers as well as the Sand layer that covers the Boston Blue Clay. Hence, for these 4 layers alternative material sels have to be made. These alternative material sets are exact copies of the original material sets but with a different vertical permeability as given in table 2 above. Follow the same steps as mentioned in Alternative 1 to create copies of the original material sets, but this time change the vertical permeabilities and not the horizontal permeabilities. 8 Computational Geotechnies Boston embankment using SSC and drains Definition of the zone with equivalent horizontal permeabilities ‘The influence zone of the drains is in principle assumed the same as for Alternative 1. Hence, draw the boundary of the zone with equivalent vertical permeabilities following these steps: # Draw a geometry line from (x,y)=(0, -31.5) to (61, -91.5) and to (51, 1.5). This borders the zone in which the soil has the equivalent horizontal permeability. ‘© Within the zone formed assign the newly made material sets to the appropriate layers. Hence, the Sand with equivalent horizontal permeability to the soil layer between Om and -2.5m, the Boston Blue Clay 1 layer with equivalent horizontal permeability below that layer and so on. Mesh generation Reset all mesh refinements and set the Global coarseness to Medium. Make a local refinement by refining all clusters in both the zone with equivalent permeabilities as well as the embankment using the Refine cluster option from the Mesk menu. CALCULATION Continue to the PLAXIS Calculation program, the phase list is the same as for the Boston Embankment with Soft Soil Creep model, The mesh regeneration has already updated all phases according to the changes made in the geometry, hence there is no need to redefine the calculation phases. Mako sure to select @ node on the axis, of symmetry just below the embankment in order to draw a load-displacement curve after the calculation. Press Calculate to start the calculation. OUTPUT Check output results for the End of Construction phase (Phase 6) and the End of Primary Consolidation (Phase 7) Again, give special attention to the development of excess pore pressures and compare the results with Alternative 1 Urstee0: Quando nko divas os duos —» 1 95 saralesle (soem Opsecnoe, pecstbclilale — coud oy Tevenne danse) # Kl > kh ouperul do cabo, thy kesuwrabule (yiudir 6) Computational Geotechnies 9 Boston embankment using SSC and drains COMPARISON OF THE RESULTS With the Curves module the settlement of the middle of the embankment as a function of time can be plotted as in figure 4 | ceulvalent 8 Excers pore pressures a) 200 | | 350 oy 7 See | 0 100 1000 10000 00000 *ac0000 Figure 4: Settlements of the embankment - Drains vs. Equivalent vertical permeability Tt can be seen that during the whole consolidation process the alternative with equivalent vertical permeability has loss vertical settlements compared to the calculation using drains, Only after creep behaviour becomes dominant over consolidation (t > 2000 days) settlements are almost equal increasing from 2.5 meters after 2000 days to 3.1 meters after approx. 14000 days (almost 40 years) In figure 5 a comparison is made between the maximum excess pore pressures that, occur during the calculation for both alternatives. Tt can be seen that with the use of equivalent vertical permeabilities the maximum excess pore pressure is higher than with the use of drains during almost the entire calculation, Figure 6 shows an important reason for the difference in settlement between the two altemative calculations. In ‘the case of drains there are hardly no excess pore pressures in the area where drains have been applied while for the alternative with equivalent vertical permeabilities there ave still significant excess pore pressures in that same area, Main reason for this is that the individual layers cannot free drain above and below, hence the drainage path is very long. The equivalent vertical permeabilities are dependent on the drainage path which for more complex stratigraphy is hard to determine. 10 Computational Geotechnies Boston embankment using SSC and drains “Drains + Fh eauvalon iy equal g 2 [Excess pore pressures (KPa) 5 5 Figure 5: Development of excess pore pressures - Drains vs. Equivalent vertical permeability Figure 6: Excess pore pressure after the last construction phase in case of using drains (left) or equivalent vertical permeability (right) Computational Geotechnies ul Boston embankment using SSC and drains APPENDIX A: EQUIVALENT HORIZONTAL PERMEABILITY According to CUR publication 191 of the Dutch Centre for Civil Engineering Research and Codes one can model a SD drain field with 2D plane strain drain field taking into account 3D effects. These 3D effects of flow towards local drains are taken into account with an adjustment of the horizontal permeability: c R= autor ke a= azainivipo2 azaleas = sS2z [inlay - +m (-ae)), In this equation © U = degree of consolidation « D = equivalent drain distance: € For a triangular pattern of drains D = 1.05 times the real drain distance For a rectangular pattern of drains D = 1.13 times the real drain distance ¢ d= equivalent drain diameter: For circular drains and sand drains d ~ real drain diameter = For strip drains d = (2b-+2h)/ where b and h are the width and thickness of the strip drain. ‘+ B = half the horizontal distance between the drains in the 2D plane strain model. For this exercise we're mostly interested in the final situation, hence the degree of saturation is 100%. For practical purposes it’s better to chose U = 0.99 though. The chosen parameters are now: u = 0.99, s.24lMe2O802 yg = dx (29:15420 02 C Filling in gives: ‘The distance between drains in the plane strain model is 4.5m, hence B=2.25m, so that finally: hy, = agbrky = 3.1 BE hy kh = 2.7 ky ¢ 2 Computational Geotechnies Boston embankment using SSC and drains APPENDIX B: EQUIVALENT VERTICAL PERMEABILITY According to CUR publication 191 of the Dutch Centre for Civil Engineering Research and Codes one can model a 3D drain field with a continuum soil body using an equivalent vertical permeability to properly predict the consolidation process: =k + BIER, as w= yl) 3+ eae Im this equation ‘© U = degree of consolidation © D = equivalent drain distance: ~ For a triangular pattern of drains D = 1.05 times the real drain distance ~ For a rectangular pattern of drains D = 1.13 times the real drain distance © d= equivalent drain diameter: wn Bt cules deaias ~ For strip drains d = (2b+2h)/7 where b and h are the width and thickness of the strip drain, © © H = the maximum vertical drainage path in the soil layer: — For a layer with perfect drainage above and below: H=half the layer thickness ~ For a layer with only perfect drainage on one side: H=layer thickness Since for the Boston embankment drainage will occur at both the top and bottom of the Boston Blue Clay layers it would be tempting to use half the layer thickness for H. However, on both sides drainage is not perfect and so His assumed to be the full layer thickness. For this exercise we're again mostly interested in the final situation, hence the degree of saturation is 100%, For practical purposes it’s better to chose U = 0.99 though. The chosen, parameters are now: wy ia of I Drains in rectangular pattern: D = 1.1941.5 ~ 1.7m Filling in gives: n= = 152 4-198 And for the equivalent vertical permeability: Ki = hy + 2 bah Ki, = ky + O.5THPy, Computational Geotechnies 3

You might also like