Missouri University of Science and Technology Scholars' Mine
International Conferences on Recent Advances 2010 - Fifth International Conference on Recent in Geotechnical Earthquake Engineering and Advances in Geotechnical Earthquake Soil Dynamics Engineering and Soil Dynamics
26 May 2010, 4:45 pm - 6:45 pm
Method to Reduce Variability of S-Wave Profiles in Seismic Cone Penetration Tests
Lou Areias Belgian Nuclear Research Centre/ Vrije Universiteit Brussel, Belgium
Follow this and additional works at: https://scholarsmine.mst.edu/icrageesd
Part of the Geotechnical Engineering Commons
Recommended Citation Areias, Lou, "Method to Reduce Variability of S-Wave Profiles in Seismic Cone enetrP ation Tests" (2010). International Conferences on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics. 17. https://scholarsmine.mst.edu/icrageesd/05icrageesd/session01b/17
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.
This Article - Conference proceedings is brought to you for free and open access by Scholars' Mine. It has been accepted for inclusion in International Conferences on Recent Advances in Geotechnical Earthquake Engineering and Soil Dynamics by an authorized administrator of Scholars' Mine. This work is protected by U. S. Copyright Law. Unauthorized use including reproduction for redistribution requires the permission of the copyright holder. For more information, please contact [email protected].
METHOD TO REDUCE VARIABILITY OF S-WAVE PROFILES IN SEISMIC CONE PENETRATION TESTS
Lou Areias ESV Euridice/SCK•CEN Belgian Nuclear Research Centre and Vrije Universiteit Brussel 2400 Mol, Belgium, [email protected]
ABSTRACT
The pseudo method used to calculate shear wave velocity (Vs) in seismic cone penetration (SCP) tests often generates high variability of Vs values at shallow depths. This occurs when travel paths are small and signal variability large to allow accurate arrival time differentiation between successive signals. The offset distance between the source and receivers has the largest influence on signal variability. A method described in this paper shows good results in reducing Vs variability of SCP tests during post processing. The method consists of increasing the sampling interval to calculate Vs and then regrouping the data to provide its original test-depth profile. The method is illustrated with a case study.
INTRODUCTION
There are three main methods to calculate and display SCPT As a result, lower signals, which normally have longer travel test data: the pseudo-interval; the true-interval; and the times than upper signals, can appear to have shorter travel assumed travel path (ATP) methods. In each of the methods, times. In other cases, wave-velocity profiles will display the main objective is to generate seismic velocity and moduli seemingly abnormal velocity variations when the differences profiles in soil. The steps taken to accomplish this include: in successive travel paths are small, and signal variability identifying arrival times; calculating travel-path length, large, to allow proper differentiation between signals. velocity and moduli; and plotting the data using an appropriate method. Often this arises when both the test interval and offset distance are made small. Choosing small test intervals may be The pseudo-method generates high variability in shear Vs and important to increase measurement resolution. On the other compression Vp velocity profiles at shallow depths when hand, offset distance usually depends on the type of equipment relatively small differences in ATP values exist in this zone. used and is usually fixed. The discussion in this paper focuses Small ATP differences result when selected test-depth primarily on these two parameters. The objective is to intervals and/or offset distances are incompatible with both demonstrate how to reduce apparent signal variability by wave-velocity and accuracy of the data acquisition (DAQ) changing the test-depth interval during post-processing. system. In theses cases, signal variability (Areias & Van Impe, 2005) may be high enough to influence the outcome of arrival-time measurements. PSEUDO-INTERVAL METHOD
This happens, for example, when the statistical range The pseudo-interval method (Patel, 1981 and Rice, 1984) (difference between statistical maximum and minimum) of converts ATPs into their vertical-equivalent ray-path travel measured arrival times, which depends on setup and DAQ distances values. ATPs are straight paths between the source system characteristics, approaches the difference in travel-time and receivers. This is not always correct (Areias & Van Impe, between signals from succeeding depths. 2006), although it holds approximately in most cases (Areias, 2007).
Paper No. 1.20b 1
The pseudo-interval method needs only two geophones, one oriented horizontally to measure shear (S) waves and another in the vertical direction to detect compression (P) waves. Jacobs & Butcher (1996) refer to this setup as incremental SCPT testing. A schematic illustrating the pseudo-interval method appears in Fig. 1.
The assumed (straight) ray-travel path of ray n (ATPn) in Fig. 1 is the Pythagorean length of the hypotenuse formed by the right triangle with offset distance (X) and vertical length Zn, expressed as: 2 2 ATPn Zn X (1) where: X = horizontal offset distance between source and SCPT cone rods; and Z Zn = vertical distance between surface and receiver of ray n. T
The corrected total travel time (Tcorr) for a given ray n becomes: Z n (2) Tcorr Tmeas ATPn where: Fig. 1. Ray-path geometry for pseudo-interval method Tmeas = total travel time measured from SCPT test.
The values of ∆Z and ∆T are then: z Zn1 Z n is the test-interval depth (3) INFLUENCE OF OFFSET DISTANCE AND TEST INTERVAL T T T (4) corr, n1 corr, n with Zn+1, Tcorr, n and Tcorr, n+1 as defined in Figure 1. The offset distance (Fig. 1) has the largest effect on signal variability when combining large offset distances with small Substituting gives the pseudo-interval velocity Vs, p of S or P ∆Z test intervals, as illustrated by the solid-line curves in waves: Fig. 2. The figure plots changes in ATP distance between two z Z Z V n1 n (5) successive test depths for two cases of offset distances of s, p 1.0 m and 4.0 m and two test-depth intervals of 0.5 and 1.0 m. T Tcorr,n 1 Tcorr,n Layering also influences ray-path distance but its influence is The terms Tcorr,n and Tcorr,n+1 in Equation 5 are the total travel generally small and ignored in the calculations. It depends times for signals n and n+1, respectively. Therefore, velocities mainly on velocity contrast between soil interfaces calculated by Equation 5 are average velocities for test interval (Areias, 2007). ∆Z. The corrected test depth (Zcorr,n) corresponding to these velocities is then: It is evident from Fig. 2 that a setup represented by the solid- Z -line curves is preferable to the one described by the dashed Z Z (6) corr,n n 2 lines. It shows that the solid-line curves, which represent a This is the test depth reported when plotting the SCPT test source with an offset of 1.0 m and two different test intervals data. of 0.5 and 1.0 m, reach a maximum difference in ATP length at a depth of 5.0 m, approximately. These differences in ATP Alternatively, one can express Equation 5 in terms of total length correspond to the respective ∆Z values of 0.5 m and ATPs and travel times to give: 1.0 m. ATP ATP n1 n The solid lines give the greatest travel-time difference Vs, p (7) Tmeas,n1 Tmeas,n between signals when compared with the setups represented by the dashed lines. In the first case, the maximum change in ATP reaches its maximum value at a depth of approximately This method provides V(s,p) values directly without first having to convert arrival times to their vertical equivalent, as is the 2.5 m, as shown. case for the first method.
Paper No. 1.20b 2
Change in ATP [m] Table 1. Proposed method to change ∆Z values during n post-processing 0.00 0.20 0.40 0.60 0.80 1.00
0.0 Field Regrouped signals (original) 1.0-m intervals signals 5.0 Non- sampled at Integral integral 0.5-m depths depths intervals [m] 10.0 [m] [m] 0.0 0.0 - 15.0 0.5 - 0.5 1.0 1.0 - Depth [m] Depth 1.5 - 1.5 20.0 2.0 2.0 - 2.5 - 2.5 3.0 3.0 - 25.0 1.0m offset & 0.5m interval 3.5 - 3.5 4.0m offset & 0.5m interval 1.0m offset & 1.0m interval . . .
4.0m offset & 1.0m interval . . . 30.0 . . . Fig. 2. Changes in ATPn with depth for 1.0 and 4.0 m offsets . i.0 . at 0.5 and 1.0 m test-depth intervals i.5 i.5
By contrast, the setups shown by the dashed lines, which are Each of the re-sampled subgroups, therefore, gives velocity for a 4.0-m offset, require a depth of at least 15 m to reach profiles equivalent to those obtained using 1.0-m-depth ATP differences close to their respective test-interval values intervals in the field. Similarly, combining the results from of 0.5 m and 1.0 m. It is evident that these setups will both subgroups gives velocity profiles for the original depth potentially result in greater signal variability than the first two interval of 0.5 m. An illustration of this method appears in cases because they represent shorter differences in ATP Fig. 3, which shows velocity profiles for two SCPT tests. length. V [m/s] Vs [m/s] s Test-depth interval ∆Z thus plays an important role in 0.00 100.00 200.00 300.00 400.00 0.00 100.00 200.00 300.00 400.0 0.0 0.0 determining differences in ATP length between SCPT tests. For example, setups with a test-interval of 1.0 m lead to 2.0 2.0 considerably larger ATP differences than the ones having only 4.0 4.0 0.5 m (left-hand curves) intervals. Since larger ATP differences allow better differentiation between arrival times, 6.0 6.0 this suggests that changing ∆Z improves the method of 8.0 8.0 Depth [m] calculating wave velocity and reduces unwanted signal [m] Depth variability from velocity profiles. 10.0 10.0
12.0 12.0
14.0 0.5m original interval 14.0 0.5m original interval PROPOSED METHOD TO CHANGE ∆Z VALUES 0.5m regrouped 0.5m regrouped DURING POST-PROCESSING 16.0 16.0
As described below, it is possible to select different ∆Z values Fig. 3. S-wave profiles for original and regrouped during post-processing by re-grouping signals into alternative 0.5-m-depth intervals depth intervals, as illustrated in Table 1. The signals obtained Figure 3 shows Vs profiles for two SCPTs obtained using both at 0.5-m-depth intervals in the field are re-grouped into two the original test-depth interval of 0.5 m and a 1.0-m-depth subgroups using a new interval of 1.0 m. The subgroups interval, with the latter results combined to give velocity consist of integral and non-integral depth intervals, namely: values for each half-meter depth using the above technique. one group with signals from 1.0, 2.0, 3.0,…, i.0 m and another The offset distance was 4.0 m for both tests. As expected, the with 0.5, 1.5, 2.5,…, i.5 m depths, where i is an integer regrouped data give an improved velocity profile, with number representing test depth. markedly less variability in the uppermost 2.5 m, than the original profile.
Paper No. 1.20b 3
For comparison, the Vs profiles analysed using both the REFERENCES original depth interval and the re-sampled 1.0-m-integral-depth subgroup appear in Fig. 4. The results Areias, L (2007). A study of the SCPT test. University of from the re-sampled data, therefore, give values every Gent, PhD thesis. one-meter depth. The results show a similar reduction in velocity variability as in the previous figure, even though they Areias, L. & W.F. Van Impe (2005). Selecting a seismic contain half as many values. This loss of definition, however, source for the SCPT test. 16th International Conference on does not significantly affect the results, as suggested by the Soil Mechanics and Geotechnical Engineering. September 12- plots. 16, Osaka, Japan
V [m/s] s Vs [m/s] Areias, L. & W.F. Van Impe (2006). Effect of layering on 0.00 100.00 200.00 300.00 400.00 0.00 100.00 200.00 300.00 400.00 travel path rays in the SCPT test method. In situ methods for 0.0 0.0 site characterization. GeoShanghai International Conference.
2.0 2.0 Shanghai, China, June 6-8
4.0 4.0 Jacobs, P.A. & A.P. Butcher (1996). The development of the 6.0 6.0 seismic cone penetration test and its use in geotechnical engineering. Advances in site investigation practice. Thomas 8.0 8.0 Telford, London. Depth [m] Depth Depth [m] 10.0 10.0 Patel, N.S. (1981). Generation and attenuation of seismic 12.0 12.0 waves in downhole testing. M.Sc.E thesis, Dept. Civil 14.0 0.5m interval 14.0 0.5m interval Engineering, University of Texas, Austin, TX. 1.0m interval 1.0m interval 16.0 16.0 Rice, A. (1984). The seismic cone penetrometer. M.A.Sc. Fig. 4. S-wave velocity profiles from original and regrouped Thesis, Department of Civil Engineering, University of British 1.0-m-integral-depth intervals Columbia, Vancouver, Canada.
CONCLUSIONS
The pseudo-interval method generates unwanted V(s,p) variability when choices of offset-distance and test-depth intervals are incompatible with DAQ system properties. The Offset distance has the largest effect on signal variability when combined with small test-depth ∆Z intervals.
The method proposed to reduce V(s,p) variability provides a simple and efficient way of post-processing data to improve velocity profiles for cases using large offset distances together with short test-depth intervals. This method consists of regrouping signals using a larger depth interval than originally used during testing and then combining them into their original depths to provide the original depth-interval profile.
Paper No. 1.20b 4