6.5 Time Variation of Pile Capacity (Pile Setup)
6.5.1 Reported Results from Field Studies
Tomlinson (1994) reports the results of pile load tests carried out on 200×215 mm piles into soft clay at different times after installation. Figure 6.13(a)shows the measured and estimated (Skov and Denver, 1989) capacity gain ratio for one pile;Figure 6.13(b)shows the measured and estimated (Svinkin and Skov, 2002; Chun et al., 1999) capacity gain ratio for the same pile.
It is evident fromFigures 6.13(a) that for large lapsed times after EOID the relationship proposed by Chun et al. (1999) predicts the capacity gain ratio over the entire time duration better than the method proposed by Svinkin et al. (2002) for the case studies considered above.
Two material parameters,B andC,are needed for the Chun method of capacity prediction whereB andCare the long-term capacity gain ratio and a material constant, respectively. The relevant value of B is obtained by considering the time capacity variation of the measured capacity gain ratio whereas parameterCis obtained by matching the measured with the predicted values of the capacity gain ratio. The values of parameters BandCestimated by Thilakasiri et al. (2003) and available values obtained from the literature are shown inTable 6.8.
Also indicated inTable 6.8are the times taken to develop 90% of the long-term capacity and the percentage of the long-term capacity developed 1 week after the EOID.Table 6.8 shows that the 1-week wait period from the EOID is sufficient for piles in sand whereas the 1- week period is not enough for piles driven into clay deposits for which a minimum wait period of 2 to 3 weeks may be required.
More recently, Bullock et al. (2005) published their test findings on a Florida test pile program. The Florida DOT commonly uses 457mm (18 in.), square, prestressed, concrete piles to support low-level bridges. Bullock et al. (2005) provided the instrumentation and installed dedicated test piles of this type at four bridge construction sites in northern Florida.
Each pile included an O-cell cast into the tip, strain gauges at soil layer boundaries, and total stress cells and pore pressure cells centered in one pile face between adjacent strain gauge elevations. They calculated the shear force and average shear stress acting on the face of the pile from the difference in load between adjacent strain gauge levels. The strain gauges defined a total of 28 side shear segments, of which 18 also included pore pressure and total
data to investigate the Side Shear Setup (SSS) for each segment. Bullock et al. (2005) expressed Equation (6.25a) as
Page 257
FIGURE 6.13
(a) Measured and estimated capacity from Svinkin et al. (2002) and Chun et al. (1999) for the pile I of Tomlinson (1994). (From Svinkin, M.R. and Skov, R., 2002, Setup effect of cohesive soils in pile capacity, 2002,http://www.vulcanhammer.net/svinkin/set.htm. With permission.) (b) Side shear setup results from the Florida pile testing program. (From Bullock, P.J., Schmertmann, J.H., Mcvay, M.C., and Townsend, F., 2005,Journal of Geotechnical and Geoenvironmental Engineering,131(3): 292–300. With permission.)
TABLE 6.8
Available Values of ParametersBandCfor Chun (1999) Method
Source Soil
Type
ParameterC ParameterB Time for 90%
ofRα(days)
Capacity Gain After 7 Days (% ofRα) Tomlinson
(1994)—pile I
Soft clay
3.01 8.2 16 72
Tomlinson (1994)—pile II
Soft clay
3.58 8.2 16 69
Chun et al.
(1999)
Clay 5.34 1.64 3 99
Chun et al.
(1999)
Clay 6.13 9.29 20 60
Chun et al.
(1999)
Stiff clay
2.8 1.7 3 99
Chun et al.
(1999)
Sand 2.08 0.57 1 100
Chun et al.
(1999)
Sand 1.41 4.33 5 94
Chun et al.
(1999)
Sand 2.8 1.7 3 100
Chun et al.
(1999)
Sand 1.6 0.21 1 100
Source:From Thilakasiri, H.S., Abeyasinghe, R.M., and Tennakoon, B.L., 2003, A study of strength gain of driven piles,Proceedings of the 9th Annual Symposium,Engineering Research Unit, University of Moratuwa, Sri Lanka. With permission.
(6.25b) where Qis the capacity of the entire pile in subsequent segmental analysis, Q0is the capacity at initial reference time t0, t is the time since EOID,t0is the reference time since EOID, andA is the dimensionless setup factor.
Bullock et al. (2005) presented the following relationship between the segmental side shear setup factors and the side shear setup factor for the entire pile:
(6.25c) wherefs0iis the unit side shear stress at time t0for segmenti, Liis the length of segmenti,and Aiis the side shear setup factor for segmenti.
Values of Aobtained by Bullock et al. (2005) for staged and unstaged tests are shown in Figure 6.13(b). Bullock et al. (2005) recommend a reduction factor ofCst=(Aunstaged/
Astaged)=0.4 for all soil types to correct setupA factors measured using staged field tests, including repeated dynamic re-strikes, repeated static tests, or repeated SPT-Ts.
Based on the relevant literature and the above study, Bullock et al. (2005) reach the following general conclusions and recommendations:
1. Using staged tests of unloaded piles, and an accurate measurement of side shear obtained by the O-cell test method, this research demonstrated SSS similar to that observed by others in prior research.
2. All pile segments showed setup, with similar average magnitudes in all soils and at all depths, continuing long after the dissipation of pore pressures, and with postdissipation setup due to aging effects at approximately constant horizontal effective stress. The pile tests (all soil types) and the SPT-T predictor tests (cohesive soils only) confirm the approximately semi-log-linear time setup behavior previously observed by others.
3. For soils similar to those tested in this research or known to exhibit SSS, a defaultA=0.1 is recommended without performing predictor tests, and higher values when supported by dynamic or static testing of whole piles, or staged SPT-Ts in clay and mixed soils. Reduce A-values measured during staged tests (pile or SPT-T) by the factorCst=0.4. Reduce pile segmentAi and SPTAby the factorCpile=0.5 for movement compatibility with whole-pile side shear capacity (if unknown). If the SPT-T Astaged≤0.5, use the defaultAi=0.2 andA=1.
4. A conservative method is proposed for including SSS in pile capacity design. The appendix in Bullock et al. (2005) provides some idealized, but realistic, examples to show the
methods recommended for including SSS in design. Depending on the percentage of capacity due to side shear, the final design time, and the applicable setup factors, SSS may significantly increase design pile capacity.
5. Dynamic tests during initial driving and subsequent re-strikes provide a method, after applying the 0.4 reduction factor for stage testing, by which to check the designAvalue.
Repeated re-strikes also allow SSS behavior to occur at the increased rate of staged testing, and may permit the acceptance of a pile that initially does not demonstrate adequate
capacity.
The research program by Bullock et al. (2005b) confirms the approximate semi-loglinear time relationship of SSS and extends it to instrumented pile segments as well as the entire pile.
Short-term dynamic tests and long-term static tests produced similar SSS behavior (Bullock et al., 2005a), apparently with no significant change before and after the dissipation of excess pore pressure. The measured side shear and horizontal effective stresses seemed reasonable, with negligible adhesion at the pile-soil interface and an increase in the interface friction coefficient (tanδ) of 40% during SSS. All depths had about the same range of pile segmentA1 values, with a minimumA1=0.2 and with no apparent depth dependency. These findings apply to all of the soil types tested, ranging from plastic (plasticity index≤60%) clays to shelly sands.