Esee00 4

ESEE Report No.
00-4 August 2000
Engineering Seismology and Earthquake Engineering
Reappraisal of the effect of vertical ground motions on response
N. Ambraseys & J. Douglas
Imperial College of Science, Technology and Medicine Civil Engineering Department London SW7 2BU
1. INTRODUCTION
The work contained within this report was funded by EPRSC Grant No. GR/L87385 on Importance of the vertical strong-motion ground-motion for design of structures. It updates and extends earlier work by Ambraseys and Simpson which sought to answer the question: does the vertical component of ground motion constitute a signicant proportion of the loading that has to be resisted by a building and by its foundations. It concentrates on ground motion in the near-eld where the importance of vertical accelerations is greatest. Ambraseys & Simpson (1996) give results concerning the near-eld behaviour of the vertical to horizontal acceleration ratio. Since their work was completed new near-eld strong-motion data has been recorded and published, therefore the equations and graphs given need to be updated using these new records.
1.1
Background
Before proceeding, it should be explained why it became necessary to reappraise the question of the importance of vertical acceleration for design purposes. After almost every destructive earthquake some engineers postulate that structural damage, such as the elephants foot buckling of large liquid storage tanks, or the fracture of large diameter reinforced concrete columns supporting buildings and freeway structures was due to strong vertical ground-motions. However, there is no consensus as to the importance on damage due to vertical motions, and little that has been learned from the recent earthquake in Loma Prieta, Northridge, or Kobe indicates conclusively that damage to structures or to their foundations was caused predominantly by vertical motions.
1.2
Previous Work
There are numerous publications on the effect of vertical accelerations on structures. On buildings: Elnashai & Papazoglou (1997) On uid tanks: Veletsos & Tang (1986); Ramerstorfer et al. (1988), Chiba et al. (1987); On soil Structure interaction: Haroun & Abdelhaz (1986), Fischer & Seeber (1988), Veletsos & Tang (1988); On bridges: Saadeghvaziri & Foutch (1991); Foutch & Saadeghvaziri (1986); Field and laboratory evidence of the deterioration of the mechanical properties of building materials due to the high frequency content of the vertical acceleration time-history; Benedetti & Carydis (1999); On structural response: Orabi & Ahmadi (1988); On offshore platforms: Liou et al. (1988); On base system isolation: Liu & Tadjkbakhsh (1986). However, there are few works that examine design input parameters in terms of the zero-period and spectral response of the combined effect of vertical and horizontal, or the attenuation of these design characteristics as a function of magnitude and source distance.
1.3
Expert opinion
As a rst step for a qualitative answer to the question of how important is vertical acceleration in the design of ordinary buildings, we carried out a survey in which we asked engineers, with more that 25 years of eld experience in earthquake engineering, whose opinion commands respect in the profession, to sum up their views on whether they would rank vertical acceleration among the most important design parameters and whether they had clear evidence that its effects were responsible for observed behaviour of engineering structures. The consensus of opinion from this survey is that they could of course imagine possible structural situations in which the vertical motions could potentially be damaging, but these cases that they had found were
1. Introduction
always associated with or required some degree of deciency in design or construction. The dominant factors leading to damage in such cases were inadequate detailing of ductile behaviour, torsion of asymmetric structures, discontinuity in heightwise distribution of strength and stiffness, and the existence of clearly poor structures, the performance of which may well be exasperated by vertical acceleration. The general agreement was that although vertical acceleration has been held responsible for observed damage by some engineers, they have not seen enough properly documented evidence or a consensus of this.
1.4
The effect of vertical acceleration on stability
Structural stability depends on the amplitude, duration and frequency content of the driving forces and equally, on the structural resistance that can be mobilized during motion. In contrast with reinforced concrete and steel structures, foundations, natural slopes and ll dams can accommodate relatively large displacements without failure. However they may fail in shear along slip surfaces not so much because of excessive loading but rather because of the rapid deterioration of their strength through loss of cohesion and the development of large pore pressures and they are more sensitive than other type of structures to vertical accelerations. Even in the absence of horizontal excitation, a series of vertical acceleration pulses can cause a saturated slope of granular material to fail if their amplitude or duration is sufciently large to induce the development of large pore water pressures. Also steep natural slopes, whose stability under static conditions depends primarily on the peak strength of the material, would be vulnerable equally to horizontal, or vertical accelerations or to a combination of both. A series of vertical acceleration pulses, sufciently strong to destroy the cohesion intercept of the material and consequently bring about a drop of its strength from peak to residual, could easily lead to collapse, not so much on account of an increase in the driving forces but rather because of the adverse effect that the stresses had, regardless of their direction, on strength, reducing it from peak to residual values. Thus, for soils and soft rocks vertical acceleration alone can affect adversely both sides of the equilibrium equation: by increasing the driving forces and, at the same time, by reducing the shearing resistance of the material through an increase in the pore water pressures or by bringing about a sudden drop in strength to a residual value below that which is required to maintain static equilibrium. In the extreme, vertical acceleration can bring about liquefaction of loose saturated granular soil as effectively as the horizontal.
1.5
Masonary buildings
Because unreinforced masonry buildings are nonductile, brittle structures whose lateral systems are incapable of dissipating energy in an inelastic manner, the strength of these short-period structures must exceed the product of their reactive weight and the peak ground acceleration in order to survive seismic shaking. Again here, as in the case of soil structures, horizontal and vertical accelerations can easily cause severe deterioration of the cohesive part of the resistance of the structure and lead to collapse.
1.6
Engineered structures
All structures are designed to carry safely gravity loads and the vertical acceleration within a range of frequencies is perceived to cause changes in the level of vertical forces, the accommodation of which is the chief aim of the design. It is thus likely that signicant variations in vertical forces can be readily absorbed by elements specically designed to carry gravity loads. Therefore, one is inclined to be unconcerned about beams. Columns are more sensitive because a change of vertical load on them may signicantly affect their lateral force carrying capacity particularly at large drifts. However, the coincidence of peak lateral displacement with peak vertical inertia forces during the inelastic response of a structural system is sufciently rare to warrant great concern but worth examining and this is one of the purposes of this paper.
1. Introduction
Vertical motion will be important for special structures such as cable-stayed and arch bridges because of the coupling of the response in the two directions, while in discontinuous structures such as deck-and-column type bridges the modes are not coupled.
1.7
EUROCODE-8
The provisions in building codes, regarding the use of vertical acceleration in design are, rudimentary. The usual approach in codes is to treat the vertical design acceleration as some fraction of the horizontal, typically around 1 to 2 . 2 3 EUROCODE-8 at present recommends the use of vertical acceleration, in its spectral form, for the verication of structural integrity only in special cases of structural congurations that include long cantilevering elements, planted columns, large eccentricities or other congurations which enhance vulnerability to vertical motions. In such cases the Code recommends that for T < 0.15 s, Sa,v should be take at 0.7Sa,h , decreasing linearly with T to Sa,v = 0.5Sa,h and remaining constant for T > 0.5 s. Until now the Code does not consider the need to take into account vertical motions in the near-eld of medium to large earthquakes or in regions of predominantly normal or thrust faulting, where not only transient but also permanent vertical displacements occur. This simple approach, perhaps, reects the overall engineering experience from the design and performance of ordinary buildings, in which vertical acceleration is placed at a rather low ranking order among other considerations required for the safe earthquake resistant design of buildings. Alternative, it may reect the complexity of the problem which is not amenable to the degree of simplication required in codes. However, it should be kept in mind that near-eld ground motions from relatively large magnitude earthquakes can produce signicant ground motions in both the horizontal and vertical directions, to the extent that the vertical motion may signicantly exceed the horizontal. Furthermore, both horizontal and vertical design spectra in building codes are meant only for the condition of zero gravity and they are independent of P- effects. It would be expected therefore, that potential dual loading from both horizontal and vertical motions should be taken into account in design that would require also consideration of combined action of vertical and horizontal acceleration with gravity effects in the determination of the prediction of the attenuation of spectral ordinates.
1.8
Previous work on the attenuation of horizontal and vertical peak and spectral acceleration
1.8.1 Previous sets of data In the earlier studies by Ambraseys (1995), Ambraseys et al. (1996) and Ambraseys & Simpson (1996) use was made of four datasets for analysis: 1. A European dataset which consists of 422 triaxial acceleration records in the magnitude and distance ranges 4.0 Ms 7.9 and 0 d 260 km from 157 shallow (h 30 km) earthquakes in the European area (Ambraseys et al., 1996; Ambraseys & Simpson, 1996) 2. A World dataset which is restricted to near eld data with MS 6.0, d 15 km and peak vertical acceleration of 10 per cent g or more (plus one record of less than 0.1 g which escaped into their data set). In terms of peak acceleration, this dataset comprises 113 records from 34 earthquakes, however there are only 90 records for which full time histories are available to enable spectral ordinate determination (Ambraseys & Simpson, 1996). A point worth noting is that whilst we refer to the World dataset as near-eld data, we are not dening near-eld by these limits. See Table A.1 for a list of these records. 3. A Combined dataset of the European and World sets.
1. Introduction
4. A much larger Regional dataset of 830 records generated by 334 shallow (h 26 km) earthquakes in Europe and adjacent regions with 4.0 Ms 7.3 was used only for the prediction of the attenuation of peak acceleration (Ambraseys, 1995) 1.8.2 Previous results In a previous paper (Ambraseys, 1995), using the Regional dataset, attenuation laws for peak horizontal (ah ) and vertical (av ) ground accelerations in g have been derived which are given by: log ah = 1.43 + 0.245Ms 0.0010r 0.786 log r with: = 0.24 (1.1)
in which r = d2 + 2.72 , d is the source distance in km. For the vertical accelerations, using a smaller dataset of 620 records from 296 shallow earthquakes, it was found: log av = 1.72 + 0.243Ms 0.0017r 0.750 log r with: = 0.24 (1.2)
with r = d2 + 1.92 . A comparison of the attenuation of horizontal and vertical ground motions predicted by Equations 1.1 & 1.2 is shown in Figure 4.1. Attenuation laws of vertical peak acceleration for other regions have been investigated, among others, by Abrahamson & Litehiser (1989), Bozorgnia et al. (1995), and Ohno et al. (1996). The inuence of depth h on peak acceleration is important at close distances from small to medium magnitude crustal events and it has been investigated in a previous study (Ambraseys, 1995). For this case, with r = d2 + h2 , Equations 1.1 & 1.2 become: log ah = 1.06 + 0.245Ms 0.00045r 1.02 log r and: log av = 1.33 + 0.248Ms 0.0011r 1.00 log r with: = 0.25 (1.4) with: = 0.25 (1.3)
Figure 4.2 shows the signicant effect that source depth has in the near eld on the attenuation of acceleration for small earthquakes. Note however, that for the World dataset (Ms 6), h is not a signicant parameter. The inuence of local site effects have been tested by Ambraseys et al. (1996); Ambraseys & Simpson (1996) using a reduced European dataset of 416 triaxial records for which site conditions have been classied in terms of the local shear-wave velocity Vs,30 , that is the time-average shear-wave velocity to a reference depth of 30 m. The best t of the data gives: log ah = 1.48 + 0.266Ms 0.922 log r + 0.117SA + 0.124SS with: r = d2 + 3.5, and: with: = 0.25 (1.5)
log av = 1.74 + 0.273Ms 0.954 log r + 0.076SA + 0.058SS with: = 0.26 (1.6) with: r = d2 + 4.72 . For bedrock conditions (Vs,30 > 750 ms1 ) SA = SS = 0, for stiff soil (360 Vs,30 750 ms1 ) SS = 0 and SA = 1 and for soft soil (180 Vs,30 < 360 ms1 ) SA = 0 and SS = 1. The variation in predicted ground motion from Equations 1.5 & 1.6 is shown in Figure 4.3. Figure 4.4 and 4.5 show the predicted horizontal and vertical spectral accelerations, for different soil conditions, for Ms = 6.0, d = 10 km and for Ms = 7.0 and d = 10 km using Ambraseys et al. (1996); Ambraseys & Simpson (1996).
2. RESULTS OBTAINED USING THE EXTENDED SET OF RECORDS
2.1
Introduction
Since Ambraseys and Simpson complete their study many more near-eld strong-motion records have become available. Therefore it becomes possible to reexamine the question how important is vertical acceleration is in the near-eld? using this new data.
2.2
Denition of near-eld
The area around the fault which can be dened as near-eld is dependent on the size of the earthquake. Figure 4.6 shows a number of different denitions of the near-eld given by different authors1 . From Figure 4.6 it can be seen that there is little consensus about what the actual denition of near-eld should be. Martnez-Pereira & Bommer (1998) try to dene the near-eld in a more rigorous manner by examining different strong-motion parameters for predicting damage (damage potential parameters) against distance and magnitude. They dene the near-eld as that part of magnitude-distance space in which certain damage potential parameters are above certain thresholds. Unfortunately they do not give equations, in terms of magnitude and distance, for this space but note that it is similar to that given by Krinitzky & Chang (1987)2 . For this report it was decide to use the same denition of near-eld as Ambraseys & Simpson (1996), except the lower limit on the PGA was removed and the lower magnitude limit was reduced to Ms = 5.8. The PGA lower limit was removed for two reasons, rstly the PGA at a site cannot be known before hand and so it will not be known whether the attenuation equations given by Ambraseys & Simpson (1996) apply and secondly this limit introduces a bias which will cause the size of the accelerations to be over-estimated. Reducing the lower magnitude limit from Ms = 6.0 to Ms = 5.8 has increased the set of suitable records especially those from Europe3 . A magnitude dependent denition was not employed for two reasons. Firstly as was noted above no commonly accepted criteria exist to dene near-eld therefore choosing one over another may lead to results inconsistent with other near-eld studies where another denition is used. From Figure 4.6 it can be seen that Ambraseys & Simpson (1996) denition (which is being followed here) is one of strictest of all those published, thus using it although it may exclude some records which may actually have the characteristics of a near-eld record but should not include records which do not have these characteristics. This is thought to be better than a more relaxed denition in which some of the selected time-histories are not actual from the near-eld. The second reason is more technical and it concerns the problems of nding unbiased coefcients by regression analysis if the independent variables are highly correlated. If the upper bound on the highest allowable fault distance increased with increasing magnitude then the set of data which would full this criteria would suffer from a positive correlation between magnitude and distance. This can lead to problems with the one-stage regression method as mentioned in Section 2.6. Since attenuation relations are found in this study different severities of shaking can be accommodated within the suite of accelerograms used and this differences between them taken care of by the attenuation equations found. This is different than when a mean is taken for all records such as is the case in Nisar &
The denitions given by some authors are not explicit in their denition and so an approximate distance bound was found using Ambraseys & Jackson (1998) fault length expression 2 The denition given by Krinitzsky et al. (1993) is thought to be similar 3 Ambraseys & Simpson (1996) used a number of records from the Whittier Narrows (1/10/1987) earthquake for which they assigned Ms = 6.00. For this study the magnitude was recalculated and it was found to be equal to Ms = 5.94. Hence the difference in criteria is in fact minimal
1
2. Results obtained using the extended set of records
Golesorkhi (1995) and Elnashai & Papazoglou (1997). In these case it is important that all the records in the dataset have roughly the same level of shaking therefore the criteria used should be more strict.
2.3
Selected records
From the ESEE strong-motion database 186 free-eld4 , mainly triaxial, strong-motion records from 42 earthquakes were selected using the criteria: Ms 5.8, d 15 km and h 20 km. The chosen records are listed in Table B.1. The majority came from western North America (133 or 72 %), the rest were from Europe (40 or 22%) or from other parts of the world (Canada, Nicaragua, Japan and Taiwan) (13 or 7%). The distribution with earthquake mechanisms is: thrust (98 or 53%), strike-slip (72 or 39%) and normal (16 or 9%). Figures 4.7 to 4.10 show the distribution of the data with magnitude, distance and mechanism. As can be seen from Figure 4.7 the records are well distributed in magnitude and distance therefore the equations obtained based on this set of data are well constrained and representative of the entire dataspace (0 d 15 km and 5.8 Ms 7.8). Figures 4.8 and 4.9 also show reasonably uniform distribution of records. Figure 4.10 shows the lack of near-eld recordings of earthquakes with normal mechanisms and also the upper limit on the size of such type of earthquakes (about Ms = 6.9) which occurs due to theoretical reasons. Site conditions at the stations are also given in Table B.1 using the categorisation proposed by Ambraseys et al. (1996); Ambraseys & Simpson (1996), i.e. L: Vs,30 < 180 ms1 , S: 180 Vs,30 < 360 ms1 , A: 360 Vs,30 < 750 ms1 and R: Vs,30 750 ms1 . For many of the Californian stations and European stations soil proles were found, from which Vs,30 were found directly. The other stations site conditions were converted from other workers site categories. The site conditions for 178 of the 186 records have been classied and there are 4, 87, 68 and 23 records in the (L), (S), (A) and (R) categories respectively.
2.4
Correction procedure
Ideally all of the records used for this study would have been processed in a uniform way, for this to be possible the time-history need to be available in uncorrected format. Unfortunately some of the records (19) could only be obtained in corrected format. Since these records came from large earthquakes (Petrolia/Cape Mendocino, Landers, Big Bear, Northridge and Hyogo-Ken Nan-Bu) it was thought better to incorporate them in the study. The short period range of interest for this study (0.1 s2 s) means that any differences in the correction procedure should make little difference. The 19 records corrected in a different way are labelled with C in Table B.1. The uncorrected records were corrected using an elliptical lter (Menu, 1986) with pass band 0.2 25 Hz. For this study the values of these parameters used were: roll-off frequency: 1.001 Hz, sampling interval: 0.02 s, ripple in pass-band: 0.005 and ripple in stop-band: 0.015. An instrument correction was applied if the necessary characteristics were known for a particular record, most have the required characteristics. This pass band was chosen because some of the records which could not obtain in uncorrected form were corrected with a similar pass band. Also because of difference in record quality between the different accelerograms used means that a narrower pass band should be used than when all the records are of a high quality. The correction procedure though should not signicantly affect the results within the period range of interest.
2.5
Ground motion model used
The ground motion model assumed has the form: log y = b1 + b2 Ms + b3 d + bA SA + bS SS

The denition of free-eld used by Joyner & Boore (1981) was adopted, i.e. records from instruments in buildings of three or more storeys and from abutments were excluded unless the structures were thought not to have affected the records within the period range of interest.
4
where Ms is the surface-wave magnitude, d is the distance to the surface projection of the rupture plane, SA takes the value of 1 if the site is classied as stiff soil (A) and 0 otherwise, and SS takes the value of 1 if the site is classied as soft soil (S) and 0 otherwise. The distance dependence was not dened in terms of r = d2 + h2 because if a h term is included it is almost indistinguishable from zero and hence was dropped. Decay is assumed to be associated with anelastic effects due to large strains, which is reasonable in the near-eld. Also since both the geometrical decay terms of the form log d and the anelastic terms of the form d are highly correlated within the short distance range used in study they cannot be found simultaneously. Another point is that because of the lack of a h term log d cannot be used since some of the fault distances are equal to zero. The largest horizontal component was used for deriving the following attenuation relations for consistency with previous work.
2.6
Regression methods
A number of different regression methods exist for deriving attenuation relations, which one is used could affect the equations obtained and hence the predicted accelerations. There are two main types of regression technique used: one-stage and two-stage. In one-stage the magnitude and distance coefcients are found simultaneously whereas in the two-stage method the distance coefcients are found rst and then the magnitude coefcients are found in the second step. Within these categories there are also two further procedures, ordinary least squares estimation and random-effects (or maximumlikelihood) models (Brillinger & Preisler, 1984, 1985). The rst of these simply nds the coefcients which minimize the sum of squares of the residuals considering the error in each record to be independent from all other records. After the coefcients are found the standard deviation is found for the whole equation. In the random-effects technique the error is assumed to consist of two 5 parts: an earthquake-to-earthquake component, which is the same for all records from the same earthquake and a record-to-record component, which expresses the variability between each record not expressed by the earthquake-to-earthquake component. During the algorithm the standard deviation of these two errors is found along with the coefcients. This method is thought to better take account of the fact that each record from the same earthquake is not strictly independent. Most authors nd, if they consider the problem at all, that the regression technique used does not affect the results obtained within the range of distance and magnitude that are of engineering interest (Ambraseys & Bommer, 1991; Ambraseys et al., 1996). There have been some authors though who nd, that due to a high correlation between a records magnitude and distance, the one-stage method gives biased results (Fukushima et al., 1988; Fukushima & Tanaka, 1990) but that the two-stage technique eliminates this bias. Another study used so called reliability of the distance dependence coefcient associated with each earthquake in order to remove events which could bias the results and then nd the distance decay from the mean of the reliable earthquakes (Tong & Katayama, 1988). These three studies were all based on Japanese data where the depths and distances of the records are much larger than for this near-eld study. Also the magnitude and distances used for this study are not strongly correlated (in fact the correlation coefcient, rM,d = 0.10). Therefore these results do not apply directly to this study. The main advantage of the one-stage method is that it is simple. The two main disadvantages are that if magnitude and distance are correlated then it can yield incorrect coefcients and results are biased towards well recorded events. The main two advantages of the two-stage method is that each earthquakes amplitude factors can be examined to see whether linear or nonlinear scaling is justied and also well recorded events have less effect. The main disadvantage is that an assumption is required on how to weighted the second stage. Before deciding which regression technique to use for this study both one-stage and two-stage ordinary least squares were applied to the peak ground acceleration and spectral ordinates of the horizontal and vertiJoyner & Boore (1993) give an algorithm for assuming the error consist of three parts: an earthquake-to-earthquake component, a site component and a record component but this is not usually employed due to the limited number of records from each site
5
cal components. When the results are compared quite large differences in the magnitude dependence terms although the distance terms are almost identical. These two equations are for horizontal PGA: One-stage log ah = 0.634 + 0.202Ms 0.0238d with: = 0.21 Two-stage log ah = 0.311 + 0.151Ms 0.0228d with: = 0.25 Figure 4.11 shows the predicted horizontal PGA from both the one-stage and two-stage methods. As can be seen from Figure 4.11 the predicted PGA is similar for Ms = 6.8 but diverges for higher and lower magnitudes. The two-stage method equation predicts higher accelerations for small magnitudes and lower accelerations for large magnitude earthquakes. This is repeated for vertical PGA and horizontal and vertical spectral accelerations. Figures 4.12 and 4.13 show the cumulative frequency distribution for the magnitude scaling of both the one and two stage methods. The graphs show that the magnitude scaling of the two-stage method is based on a more evenly distributed set of magnitudes compared with the one-stage method (for example, about 25% of the points for the two-stage method are associated with earthquakes with Ms bigger than 7 whereas only about 15% of the one-stage points are). This seems to imply that the two-stage method gives a better representation over the entire magnitude range. To further investigate this question the amplitude factors, ai , which are used to nd the magnitude dependence in the second step were plotted against magnitude, see Figure 4.14. As can be seen there is much scatter in amplitude factors although a linear trend with magnitude can be seen. From this graph it is noticeable that ai from the Kocaeli earthquake (Ms = 7.8) falls well below the obtain lines and seems to be having a large inuence on the least squares line from the second step. To check this the two records associated with this earthquake were removed from the data and the analysis repeated. It was found that the distance dependence stays almost exactly the same but the magnitude dependence changes by a lot. From the one-stage method the magnitude coefcient b2 became 0.222 and from the two-stage method the coefcient became 0.195 (with corresponding changes in b1 ). This shows that the Kocaeli records are having a large effect on the magnitude dependence especially in the two-stage method. This large effect is probably not justied because there is not enough data from the large magnitude range to check whether this lowering in predicted acceleration for large magnitudes is normal or a special facet of the Kocaeli earthquake. Therefore it was decided to use the one-stage method for the rest of the analysis. 2.6.1 Inclusion of site geology in the attenuation model
The method used by Ambraseys et al. (1996) to include site geology in the attenuation was also used in this study. The residuals, i = log yi b1 b2 Mi b3 di , from the rst stage of the regression are found. Then the regression is performed on = b4 SR + b5 SA + b6 SS , where: SR takes the value 1 if the site is classied as rock and 0 otherwise, and SA and SS are similarly dened for stiff (A) and soft (S) soil sites. Then new coefcients are dened as follows: b1 = b1 + b4 , bA = b5 b4 and bS = b6 b4 and the error is recalculated with respect to the site-dependent prediction, using only those records with known site conditions.
2.7
Horizontal PGA (y = log ah )
This equation for the prediction of horizontal PGA (ah ) in the near-eld was obtained: log ah = 0.659 + 0.202Ms 0.0238d + 0.020SA + 0.029SS with = 0.214 (2.1)
Unlike Ambraseys and Simpsons study the magnitude dependence was found to be signicant at the 5% level. The soil terms were found not to be signicant at the 5% level, but were retained for comparison with other studies. Figure 4.15 shows the predicted horizontal peak ground acceleration for three different magnitudes and also the 84-percentile acceleration calculated in the way described in Draper & Smith (1981, pp. 210211).
10
These curves are for the model which does not include site geology, for which the equation is not given here. The difference between the 84-percentile acceleration calculated this way and in the normal way, i.e. adding on the standard error, is negligible. Figure 4.16 shows comparison between the peak ground acceleration predicted by Equation 1.5 and that predicted by Equation 2.1 for Ms = 6 and 7 and the three soil types (R,A and S). Note Equation 2.1 is converted to g when plotted. Comparing Equations 1.1 & 1.5 with Equation 2.1 a number of important features may be observed: Horizontal near-eld accelerations at all allowable magnitudes (Ms 5.8) predicted by the Region dataset exceed that of the new near-eld dataset, and when compared to the actual data can be considered over-conservative. This effect reects the relative distribution of the data in each dataset, in particular the Regional dataset which is dominated by small to intermediate magnitude events and in addition, the larger events which are associated predominantly with intermediate and far-eld data. This has the effect of swamping the large magnitude near-eld data and over emphasising trends determined by the smaller magnitude and distant data. The solutions from the new near-eld dataset suggest reduced magnitude and distance dependence. This reects the assumption of a point source in deriving the attenuation model which can not be considered true for large magnitude, near-eld data. Data of this near-eld type will actually reect the local fault conditions rather than the global fault features.
2.8 Vertical PGA (y = log av )

This equation for the prediction of vertical PGA (av ) in the near-eld was obtained: log av = 0.959 + 0.226Ms 0.0312d + 0.024SA + 0.075SS with = 0.270 (2.2)
As for the horizontal components the magnitude dependence was found to be signicant and the soil terms were found not to be signicant at the 5% level. Figure 4.17 shows the predicted horizontal peak ground acceleration for three different magnitudes and also the 84-percentile acceleration calculated in the way described in Draper & Smith (1981, pp. 210211). These curves are for the model which does not include site geology, for which the equation is not given here. The difference between the 84-percentile acceleration calculated this way and in the normal way, i.e. adding on the standard error, is negligible. Figure 4.18 shows comparison between the peak ground acceleration predicted by Equation 1.6 and that predicted by Equation 2.2 for Ms = 6 and 7 and the three soil types (R,A and S). Note Equation 2.2 is converted to g when plotted. Comparing Equations 1.2 & 1.6 with Equation 2.2 again a systematic difference in the predicted accelerations is noted although the new equations predict higher accelerations than those using the Regional and European sets of records. Also reduced magnitude and distance dependence is noticed in the new equations.
2.9
Vertical to horizontal absolute PGA ratio (y = log q = log av /ah )
The ratio, q, of the vertical, av , to the maximum horizontal, ah , ground acceleration can be derived either by combining the two equations which individually predict peak vertical and horizontal accelerations (Abrahamson & Litehiser, 1989), or by performing a regression directly on the ratios of maximum acceleration (Ambraseys & Bommer, 1991). Note that whereas Ambraseys and Simpson regressed directly on the ratio, av /ah , in this study the regression was done on the logarithm of the ratio. This was because it has a physical meaning and also because the error is multiplicative not additive so negative ratios cannot be predicted. No site coefcients were derived because they do not have a physical meaning, unlike those for horizontal and vertical components separately.
11
Since q represents the ratio of two functions whose maxima occur at different times, its value, which in building codes varies between 0.50 and 0.75, is a conservative estimate of the combined loading that could develop in real time during an earthquake. 2.9.1 All earthquakes
This equation for the prediction of vertical to horizontal PGA ratio, q, in the near-eld of all earthquakes was obtained: log q = log av /ah = 0.119 0.00799d with = 0.21 (2.3)
Magnitude dependence was found not to be signicant at the 5% level and so the regression was repeated with the magnitude coefcient constrained to zero. Distance dependence was signicant at the 5% level. Figure 4.19 shows the predicted q and the 84-percentile ratio for all earthquakes. 2.9.2 Normal earthquakes
Due to the limited number of records from earthquakes with normal faulting neither the magnitude or distance coefcients were found to be signicant and so the mean and the standard error of the logarithms was taken. The result is: log q = log av /ah = 0.216 with 2.9.3 Thrust earthquakes = 0.13 (2.4)
This equation for the prediction of vertical to horizontal PGA ratio, q, in the near-eld of thrust earthquakes was obtained: log q = log av /ah = 0.103 0.0133d with = 0.17 (2.5)
Magnitude dependence was found not to be signicant at the 5% level and so the regression was repeated with the magnitude coefcient constrained to zero. Distance dependence was signicant at the 5% level. Figure 4.20 shows the predicted q and the 84-percentile ratio for thrust earthquakes. 2.9.4 Strike-slip earthquakes
As for the records associated with normal faulting neither the magnitude or distance coefcients were found to be signicant and so the mean and the standard error of the logarithms was taken. The result is: log q = log av /ah = 0.138 with 2.9.5 European earthquakes = 0.25 (2.6)
This equation for the prediction of vertical to horizontal PGA ratio, q, in the near-eld of European earthquakes was obtained: log q = log av /ah = 1.11 + 0.132Ms with = 0.16 (2.7)
Distance dependence was found not to be signicant at the 5% level and so the regression was repeated with the distance coefcient constrained to zero. Magnitude dependence was signicant at the 5% level. Figure 4.21 shows the predicted q and the 84-percentile ratio for European earthquakes.
12
2.10
Vertical to horizontal simultaneous PGA ratio (y = log qsim = log av (tmax )/ah )
The vertical to horizontal simultaneous PGA ratio is dened as, qsim = av (tmax )/ah , where av (t) is the vertical ground acceleration, ah is the horizontal PGA and tmax is the time at which this peak occurs. It gives the vertical acceleration to be resisted at the time of the design horizontal acceleration. For all data subsets within this study neither the magnitude and/or the distance dependence were signicant, therefore the mean of the logarithm and the standard deviation were found. 2.10.1 All earthquakes
The result is: log qsim = log av (tmax )/ah = 0.996 2.10.2 Normal earthquakes with = 0.56 (2.8)
The result is: log qsim = log av (tmax )/ah = 0.830 2.10.3 Thrust earthquakes with = 0.44 (2.9)
The result is: log qsim = log av (tmax )/ah = 1.04 2.10.4 Strike-slip earthquakes with = 0.58 (2.10)
The result is: log qsim = log av (tmax )/ah = 0.978 2.10.5 European earthquakes with = 0.56 (2.11)
The result is: log qsim = log av (tmax )/ah = 0.939 with = 0.58 (2.12)
Note these predicted simultaneous PGA ratios are much smaller than the absolute ratios but the standard deviations are much higher.
2.11
Horizontal energy density (y = log Eh )

T
The energy density, E, of an strong motion record is dened as E = 0 v(t)2 dt, where T is the length of the record and v(t) is the ground velocity at time t (Sarma, 1971). This equation for the prediction of horizontal energy density (Eh ) in the near-eld was obtained: log Eh = 3.11 + 0.937Ms 0.0567d + 0.180SA + 0.417SS with = 0.419 (2.13)
The soil term for stiff soil (A) was found not to be signicant at the 5% level, but the term for soft soil (S) was signicant at the 5% level. Figure 4.22 shows the predicted horizontal energy density for M = 7 for the three different site categories.
13
2.12
Vertical energy density (y = log Ev )
This equation for the prediction of vertical energy density (Ev ) in the near-eld was obtained: log Ev = 3.88 + 0.96878Ms 0.0527d + 0.000SA + 0.111SS
with = 0.422
(2.14)
The soil terms were found not to be signicant at the 5% level. Figure 4.22 shows the predicted vertical energy density for M = 7 for the three different site categories.
2.13
Vertical to horizontal energy density ratio y = log qE = log Ev /Eh
The vertical to horizontal energy density ratio, qE = Ev /Eh . For all data subsets within this study neither the magnitude and/or the distance dependence were signicant, therefore the mean of the logarithm and the standard deviation were found. 2.13.1 All earthquakes
The result is: log qE = log Ev /Eh = 0.756 2.13.2 Normal earthquakes with = 0.34 (2.15)
The result is: log qE = log Ev /Eh = 0.540 2.13.3 Thrust earthquakes with = 0.25 (2.16)
The result is: log qE = log Ev /Eh = 0.762 2.13.4 Strike-slip earthquakes with = 0.34 (2.17)
The result is: log qE = log Ev /Eh = 0.795 2.13.5 European earthquakes with = 0.35 (2.18)
The result is: log qE = log Ev /Eh = 0.632 with = 0.32 (2.19)
2.14
Horizontal spectral acceleration (y = log SAh )
Table 5.1 contains the coefcients of the horizontal spectral acceleration attenuation relation for 5% damping and 46 periods between 0.1 and 2 s. Soil coefcients labelled with (*) are signicant at the 5% level. Figures 4.23 and 4.24 show the predicted horizontal response spectrum for Ms = 6.0, d = 10 km and Ms = 7.0, d = 10 km for the three soil types for comparison with Figures 4.4 and 4.5.
14
2.15
Vertical spectral acceleration (y = log SAv )
Table 5.2 contains the coefcients of the vertical spectral acceleration attenuation relation for 5% damping and 46 periods between 0.1 and 2 s. Soil coefcients labelled with (*) are signicant at the 5% level. Soil coefcients for both soft (S) and stiff (A) soils show de-amplication, with respect to rock, at long periods (T > 1 s). To check this was not simply due to the distribution of soil classes within the data space all the Northridge (17/1/1994) records were removed, because many of the soft soil records are from the Northridge earthquake, and the analysis repeated. All the coefcients, including the soil coefcients, were almost unchanged by the removal of those 34 records, hence the de-amplication seems to be genuine. Figures 4.23 and 4.24 show the predicted vertical response spectrum for Ms = 6.0, d = 10 km and Ms = 7.0, d = 10 km for the three soil types for comparison with Figures 4.4 and 4.5. A comparison of Figures 4.4 and 4.5 with Figures 4.23 and 4.24 reveals that the predictions given by the attenuation equations presented by Ambraseys et al. (1996); Ambraseys & Simpson (1996) which were derived using the European dataset, which includes many small magnitude earthquakes and records from large distances, roughly agree with the predictions made using the only the new near-eld dataset. For Ms = 6.0 at d = 10 km the new near-eld equations predict slightly higher (about 10 to 15%) spectral accelerations for the horizontal component and vertical accelerations about 50% larger in the short period range (T 0.25 s). For Ms = 7.0 at 10 km the new near-eld equations predict roughly the same horizontal accelerations but higher (about 20%) vertical accelerations for the short period range (T 0.25 s). Thus unlike for the case of peak ground acceleration the presence of many weak motion recordings, i.e. those from small magnitude events or large distances, does not seem to bias the equations towards given predictions based on the properties of these records. This could be because spectral accelerations are less affected, than peak ground acceleration, by properties other than magnitude and distance.
2.16
Horizontal spectral acceleration (bending model) (y = log SAh,b )
Although not always clearly stated, horizontal response spectra, in their usual form, are valid only for zero gravity conditions. This is obvious from the way they model structures, viz.:
2 u + 20 0 u + 0 u = ah (t)
(2.20)
in which, u, u and u are respectively the acceleration, velocity and displacement of the system, 0 is the fraction of critical damping and ah (t) is the absolute ground acceleration input. The undamped period, T0 = 2/0 , is independent of g. The horizontal response of a SDOF system under gravity, g, and vertical earthquake loads, av (t), can be modelled in two ways depending on whether the system bends or hinges at its base. It can be shown that the equations of both these models can be expressed as in the form of a Mathieu-Hill equation (Lubkin & Stoker, 1943):
2 u + 21 1 u + 1 [1 av (t)]u = ah (t)
(2.21)
2 2 For the bending case: = /g(1 ); 1 = 0 (1 ); and 1 = 0 0 /1 where = mg/Pcr (Pcr is the maximum Euler buckling load that the structure can withstand). 2 2 2 For the hinging case: = 1/(0 l g); 1 = 0 g/l; and 1 = 0 0 /1 where l is the length of column which approximates the structure. As can seen, from comparing Equations 2.20 and 2.21 with av (t) = 0, the effect of gravity is simply to change the natural period and damping of the system. In the literature, work in this area has been conducted on the bending model by Lin & Shih (1980); Shih & Lin (1982); Orabi & Ahmadi (1988); Orabi et al. (1989); Loh & Ma (1997) and on the hinging model for inelastic systems by Jennings & Husid (1968); Husid (1969); Sun et al. (1973); Tani & Soda (1977). Many of these works either do not use earthquake strong-motion and only apply stochastic simulations or do not consider the vertical earthquake motion acting at the same time as the horizontal. Thus there is a need for an
15
examination of Equation 2.21 under real structural and input conditions. In this study the bending model has been investigating using the near-eld set of records. The nonlinear nature of Equation 2.21 means that the piece-wise method of Nigam & Jennings (1969) to nd the response solution does not yield the correct results for large loads and high vertical accelerations. Thus an adaptive step length Runge-Kutta algorithm (Press & Teukolsky, 1992) was used to compute the response of SDOF systems in this study, Orabi & Ahmadi (1988) also used such a method. It was chosen to consider the response of systems governed by Equation 2.21 in terms of the loaded natural period, T1 = 2/1 , and damping, 1 , so that they can be compared easily with results from Equation 2.20. For the bending model there is a limit on the largest that the load ratio, , can be before the response of the system becomes unrealistically large. This limit is a function of the size of the maximum vertical accelerations present in the strong-motion record and is given by: g/{max[av (t)] + g}. Although this limit is not exact, experiments on the effect of varying show that it is a good bound. Loh & Ma (1997) nd that > 0.5 induces instability in the calculation of the response using Equation 2.21, this is for Taiwanese records normalised to 1g. This is exactly the maximum load ratio possible by using the limit dened above. In the suite of records used in this study the largest corrected vertical acceleration is 19.4 ms2 for the timehistory recorded at Nahanni-Station 1 (Iverson) during the Nahanni earthquake (23/12/1985 05:16). This means that the maximum possible load ratio which can be used is about 0.3. Thus this is the value chosen for this study. Table 5.1 contains the coefcients of the horizontal spectral acceleration (buckling model) attenuation relation for 5% damping, = 0.3 and 46 periods between 0.1 and 2 s. Soil coefcients labelled with (*) are signicant at the 5% level. The inclusion of the vertical ground motion has little effect. Figure 4.25 shows the ratio between the spectral acceleration including the effect of the vertical accelerations and not including the vertical accelerations (note that this ratio is between models not including soil terms) . For a site 0 km from a Ms = 7.8 the increase due to the vertical accelerations is about 8% and for smaller magnitudes and larger distances it is less. Loh & Ma (1997) contains a comparison of the mean of 5% acceleration response spectra from 30 Taiwanese records, normalised to 1g, with and with considering the vertical ground motion. Unfortunately they do not present the actual accelerations but only a graph. Also it is for a load ratio of = 0.5 and so a direct comparison is not possible. From their Figure 7(a) the percentage increases shown in Table 5.4 due to the vertical ground motion were found 6 . It is difcult to compare the results obtained in this study with Loh & Ma (1997) because of their different choice of load ratio and because they normalise their time-histories to 1g. The results obtained in this study though are more useful because they retain magnitude and distance scaling. The predicted increase in spectral acceleration given in this report seem to be in line with the results of Loh & Ma (1997), a lower load ratio, 0.3 compared with 0.5, leading to smaller increases. Note that the case examined is for innite vertical stiffness, i.e. the vertical input to the system is the ground motion and not the vertical response of the system which it would be if the vertical stiffness was nite. This introduces a further complexity into the problem because the vertical responses for each period, Rv (t, Tv ), need to be calculated and used as input to model 2.21 for all horizonal periods. The problem of stability noted above is worse for nite stiffness because 5% damped vertical spectral accelerations can reach 75 ms2 (at Tv = 0.11 s) (El Centro 6 accelerogram from the Imperial Valley 15/10/1979 earthquake) which corresponds to a maximum load ratio of = 0.115. Such a small does not occur in practice and hence structures (which can be modelled by Equation 2.21) subjected to such severe vertical loading can be expected to collapse. A full analysis of the nite vertical stiffness case is currently being undertaken and also being investigated is the hinging model mentioned above. Results from these studies will be presented at a later date.
6
Since they were read from a graph there may be some doubt over their accuracy
16
2.17
Horizontal maximum absolute input energy (y = log Ih )

t
The maximum absolute input energy, I, is dened as I = maxt 0 [utt (t) + a(t)]v(t) dt, where utt is the response acceleration of the SDOF system, a(t) is the ground acceleration and v(t) is the ground velocity (Chapman, 1999). Table 5.5 contains the coefcients of the horizontal maximum absolute input energy attenuation relation for 5% damping and 46 periods between 0.1 and 2 s. Soil coefcients labelled with (*) are signicant at the 5% level. Figure 4.26 show the predicted horizontal maximum absolute input energy for M = 7 and d = 10 km for the three soil types.
2.18
Vertical maximum absolute input energy (y = log Iv )
Table 5.6 contains the coefcients of the vertical maximum absolute input energy attenuation relation for 5% damping and 46 periods between 0.1 and 2 s. Soil coefcients labelled with (*) are signicant at the 5% level. Figure 4.26 show the predicted vertical maximum absolute input energy for M = 7 and d = 10 km for the three soil types.
2.19
Vertical to horizontal spectral ratio (Absolute) (y = log qs = log SAv /SAh )
There are two methods for nding the predicted vertical to horizontal spectral ratios a) divide the predicted vertical spectral accelerations (given by equations like Table 5.2) by the predicted horizontal accelerations (given by equations like Table 5.1), b) regress directly on the spectral ratio to nd new attenuation equations for the ratio. The rst technique was used in Niazi & Bozorgnia (1992); Bozorgnia & Niazi (1993); Bozorgnia et al. (1995, 2000). Its main advantage is simplicity. The second technique was used in Feng et al. (1988); Ambraseys & Simpson (1996) (although they changed the attenuation model by regressing on the ratio directly not on the logarithm). For the World dataset, results of a previous analysis are given in Ambraseys & Simpson (1996). This analysis however used the maximum ratio, that is qs = (SAv /SAh )max determined from the two horizontal components, rather than using the maximum horizontal component, that is qs = (SAv /(SAh )max ). Use of the maximum ratio for qs in Ambraseys & Simpson (1996), will if used in design in conjunction with peak horizontal spectral ordinate determinations, lead to an articially high value of the vertical spectral ordinate. Also as was noted above the the peak ground acceleration ratio their attenuation model was non-physical and the additive nature of the error meant negative ratios could be predicted. We have re-evaluated qs using the denition qs = (SAv /(SAh )max ) using the new near-eld set of records and the logarithmic ground motion model. Equations to predict the vertical to horizontal spectral ratio, qs , were derived assuming magnitude and distance dependence. For some periods, the magnitude coefcient was signicant, for some the distance coefcient was signicant but for most neither were signicant. Therefore it was decided to simply provide the mean of the logarithms and the standard error. Table 5.7 contains the coefcients and means for all earthquakes, normal, thrust and strike-slip and European earthquakes separately. No site coefcients were derived because they do not have a physical meaning, unlike those for horizontal and vertical components separately. Figure 4.27 shows the predicted qs for the all earthquakes and for each of the separate mechanism (normal, thrust and strike-slip). From this graph it can be seen that the predicted ratios are almost the same for each type of faulting except for normal faulting. The results for normal mechanism earthquakes are based on only 15 records; it is difcult to base conclusions on such a small number of records so more records are required from normal earthquakes in order to check this nding. Comparing Figure 4.27 with Figures 9, 10, 11 of Ambraseys & Simpson (1996) it can be seen that the difference in the vertical to horizontal spectral ratios for different fault mechanisms apparent in Ambraseys & Simpson (1996) simply seems to be a consequence of either the small dataset used or their non-physical attenuation model.
17
The results obtained are similar to those obtained by Feng et al. (1988); Niazi & Bozorgnia (1992); Bozorgnia & Niazi (1993); Bozorgnia et al. (1995); Ambraseys & Simpson (1996); Bozorgnia et al. (2000), i.e. ratios about one for short periods and ratios below about 0.5 for longer periods. Something to note is that whereas both the magnitude and distance dependence for horizontal and vertical spectral attenuation are signicant when considered separately, when considered as the ratio they are not usually signicant. Therefore workers who nd distance and magnitude dependence of the spectral ratios from regressing on the horizontal and vertical spectral ordinates may nd that this dependence is not signicant if the regression is done directly on the ratio. The absolute spectral ratios for the bending model were also calculated but they are almost identical to those for the normal model, due to the similarity in the horizontal response spectra from the two models. They will not be given here.
2.20 Vertical to horizontal spectral ratio (Simultaneous) (y = log qi = log utt,v (tmax )/SAh )
A major draw-back of the acceleration ratio q or qs for practical purposes is that in an earthquake the maximum ground or response accelerations in the vertical and horizontal direction occur at different times. Equations to predict the attenuation of vertical to horizontal simultaneous spectral ratio were found. This ratio is dened as: qi = utt,v (tmax )/SAh where: utt,v (t) is the vertical response acceleration and tmax is the time as which the maximum horizontal response acceleration occurs. Therefore this ratio gives the size of the vertical accelerations which need to be withstood at the time of the design maximum horizontal acceleration. The natural period of a structure in the vertical direction is usually different than that in the horizontal direction therefore these ratios, Qi , should be found for all combinations of vertical and horizontal natural period, i.e. 46x46 = 2116. Due to space limitations the actual coefcients are only be given for the case of equal vertical and horizontal period, see Table 5.8. As for the absolute ratio case, for some periods the magnitude coefcient was signicant, for some the distance coefcient was signicant but for most neither were signicant. Therefore it was decided to simply provide the mean of the logarithms and the standard error. Table 5.8 contains the coefcients and means for all earthquakes, normal, thrust and strike-slip and European earthquakes separately. No site coefcients were derived because they do not have a physical meaning, unlike those for horizontal and vertical components separately. Figure 4.27 shows the predicted qi for the all earthquakes and for each of the separate mechanism (normal, thrust and strike-slip). From this graph it can be seen that the predicted ratios are almost the same for each type of faulting except for normal faulting. The results for normal mechanism earthquakes are based on only 15 records; it is difcult to base conclusions on such a small number of records so more records are required from normal earthquakes in order to check this nding. As Figure 4.27 shows the simultaneous ratios, qi , are much less than the absolute ratios, qs , especially for short periods. Also it can be seen that the ratios are roughly independent of period. Figure 4.28 shows the predicted vertical to horizontal simultaneous spectral ratio, Qi , for all combinations of Th and Tv . Figure 4.29 shows the standard error of the regression 7 . Figure 4.28 show that for short vertical and long horizontal periods the simultaneous ratio, Qi , can reach about 2 but for most periods the ratio is less than about 0.3. Figure 4.29 shows that the standard error is much 3 higher than for the absolute ratio and it is roughly independent of period and equal to about 0.6. The cause of the much higher standard errors at certain, seemingly random, combinations of periods is not known.
2.21
Vertical to horizontal maximum absolute input energy ratio (y = log qe = log Iv /Ih )
Table 5.9 contains the coefcients and means for all earthquakes, normal, thrust and strike-slip and European earthquakes separately. No site coefcients were derived because they do not have a physical meaning, unlike those for horizontal and vertical components separately.
7
This the the standard error of the logarithms
18
Figure 4.30 shows the predicted ratio, qe , for all earthquakes and considering the three source mechanisms separately. As for the response spectral equations only predicted ratios for normal mechanism earthquakes are different than those for other types of faulting, although this may be due to a small number of records from normal earthquakes.
3. DISCUSSION
Two important features may be drawn from the results presented: treatment of near-eld data, both horizontal and vertical, should be made separately from intermediate and far-eld data in terms of predicted peak, spectral or acceleration ratios. the effect of vertical loading in the near-eld can be important for critical cases resulting in signicant deviations from near-eld averages and should therefore warrant consideration for separate treatment. In the near-eld ground motions from relatively large earthquakes (Ms > 7) can produce signicant ground motions in both the horizontal and vertical directions and these motions would have to be taken in account in an appropriate design. The ratio of the maximum peak ground acceleration to maximum horizontal in an earthquake, which individually occur at different instants, may exceed one, but their ratio q falls off with distance. Due to lack of data for strike-slip and normal events the magnitude and distance dependence of the ratio, q, cannot be found 2 but the mean ratios are 0.73 and 0.61 respectively, hence they are close to the commonly accepted ratio of 3 . The complete near-eld dataset and the thrust subset show signicant distance dependence, the ratios decrease 2 with distance, and the predicted ratios also are close to 3 . Note, that the maximum values of the vertical and horizontal peak acceleration in this ratio occur at different times in an earthquake. We may conclude with several results from this investigation. In what follows spectral comparisons are made for a damping factor of 5 per cent critical, and that conclusion are global, deduced from the statistical analysis of the European, World, a combination of these and the new near-eld datasets. The spectral response of the vertical acceleration and the attenuation of its spectral ordinates with magnitude and distance differ in amplitude and shape from those of the horizontal. The ratio qs of the maximum vertical spectral response to the horizontal may exceed one at short periods (T < 0.15 s), but falls off rapidly with period reaching a value of about 0.5 for long periods. Note that maximum spectral values in this ratio occur at different times but for the same response period in both horizontal and vertical direction (Th = Tv ). The ratio of the vertical spectral response that occurs at the time of the maximum horizontal response qi does not exceed 0.2 at all and is insensitive to magnitude and distance. Note that the response period in the horizontal and vertical directions are equal. The acceleration ratio Q(Th , Tv ) for response periods Th and Tv which are not equal, remains smaller than 0.3. The exception is for the rather unrealistic combination of very short vertical periods (0.1 s) with very long horizontal (T > 1.0 s) for which Q may reach 0.6. The effect of the combination of horizontal and vertical excitations on response is to increase slightly (510%) the horizontal response for very severe loading but the effect is smaller for less severe shaking. Note however that this is for innite vertical stiffness and an investigation is required into the more realistic nite stiffness cases. We may conclude that for ordinary buildings the effect of vertical acceleration for codied design purposes is not as important as one might have thought. However, for non-standard type of buildings and engineering structures, particularly in the near-eld of relatively large earthquakes (Ms > 7) that can produce signicant ground motions in both the horizontal and vertical directions, these motions would have to be taken in account in an appropriate design, particularly if they are likely to have a deterioration effect on building materials.
3. Discussion
20
3.1
Acknowledgments
We would like to thank Profs. A. Chopra, G. Housner, K. Kawashima, T. Paulay and A. Veletsos, among others, for their comments regarding the effect of vertical motion on design. Also we would like to thank Drs. K. Simpson, M. Srbulov, M. Free, S.K. Sarma and P. Smit for their help in this research which is supported by EPSRC grant no. GR/L87385.
4. FIGURES
4. Figures
22
1.0
H V
0.5
H V
Peak Ground Acceleration [g]
H 0.2 V
0.1
0.05 H
0.02
H V H V
0.01 0.1 0.3 0.6 1 3 6 10 30 V 60 100
Fault Distance [km]
Fig. 4.1: Attenuation of horizontal and vertical peak ground acceleration for Ms = 5,6 & 7 from Ambraseys (1995).
h=2km 1.0
0.5
h=5km
h=10km 0.2
h=20km 0.1
0.05
0.1
0.3
0.6
10
30
60 100
Fault Distance [km]
Fig. 4.2: Attenuation of peak horizontal acceleration predicted for an earthquake of magnitude Ms = 6 at focal depths of 2, 5, 10 and 20 km from Ambraseys (1995).
4. Figures
23
1.0
0.5
S A R A S R
0.2
0.1
0.05
0.02
S A R A S R 60 100
0.01 0.1
0.3
0.6
10
30
Fault Distance [km]
Fig. 4.3: Attenuation of horizontal (top set of curves) and vertical peak acceleration predicted from a Ms = 6 earthquake for the three soil catagories R,S & A from Ambraseys et al. (1996); Ambraseys & Simpson (1996).
0.6
0.5
Spectral Acceleration [g]
0.4 A 0.3 S R A S R S A 0.1 R
0.2
0.5
1.5
Period [s]
Fig. 4.4: Horizontal (top set of curves) and vertical absolute acceleration spectra for rock (R), stiff (A) and soft soil (S) for an earthquake of Ms = 6.0 at a fault distance of 10 km using Ambraseys et al. (1996); Ambraseys & Simpson (1996).
4. Figures
24
1.2
0.8
0.6
A S R A S R S A R
0.4
0.2
0.5
1.5
Period [s]
Fig. 4.5: Horizontal (top set of curves) and vertical absolute acceleration spectra for rock (R), stiff (A) and soft soil (S) for an earthquake of Ms = 7.0 at a fault distance of 10 km using Ambraseys et al. (1996); Ambraseys & Simpson (1996).
4. Figures
25
500
200
100
7 6
50
Distance [km]
20
8 11 3 10 9 5
10
5 2 12
2 1 1
0.5
5.5
6.5
7.5
Magnitude
Fig. 4.6: Graph showing denition of near-eld by different authors. 1) Berrill (1975) 2) Tocher et al. (1977) 3) Shteinburg et al. (1980) 4) Campbell (1981); Campbell & Bozorgnia (1994) (Use expression near-source) 5) Bolt & Abrahamson (1982) 6) Ambraseys & Menu (1988) 7) Hudson (1988) (Uses expression near-source) 8) Krinitzsky et al. (1993) 9) Nisar & Golesorkhi (1995) 10) Ambraseys & Simpson (1996), 11) Hu et al. (1996) & 12) Martnez-Pereira (1999)
4. Figures
26
15 Number of records
10
0 6 6.5 7 7.5 Magnitude (M )

s
10
12
14
Fault Distance (km)
Fig. 4.7: Distribution of all records in new near-eld dataset with respect to magnitude and distance.
10
0 6 6.5 7 7.5 Magnitude (M )

s
10
12
14
Fault Distance (km)
Fig. 4.8: Distribution of records associated with thrust earthquakes in new near-eld dataset with respect to magnitude and distance.
4. Figures
27
10
0 6 6.5 7 7.5 Magnitude (M )

s
10
12
14
Fault Distance (km)
Fig. 4.9: Distribution of records associated with strike-slip earthquakes in new near-eld dataset with respect to magnitude and distance.
10
0 6 6.5 7 7.5 Magnitude (M )

s
10
12
14
Fault Distance (km)
Fig. 4.10: Distribution of records associated with normal earthquakes in new near-eld dataset with respect to magnitude and distance.
4. Figures
28
10 9 8 7
PGA [ms ]
6 5 4 3 2 1 0 0
10
15
Distance [km]
Fig. 4.11: Comparison of predicted horizontal PGA using one-stage and two-stage regression methods for Ms = 5.8, 6.8 & 7.8. Solid line is from one-stage equation and dashed line is from two-stage equation.
0.9
0.8
0.7
Cumulative Frequency
0.6
0.5
0.4
0.3
0.2
0.1
0 5.8
6.2
6.4
6.6 6.8 7 7.2 Magnitude [Ms]
7.4
7.6
7.8
Fig. 4.12: Cumulative frequency distribution of records used for magnitude scaling of one-stage method
4. Figures
29
0.9
0.8
0.7
Cumulative frequency
0.6
0.5
0.4
0.3
0.2
0.1
0 5.8
6.2
6.4
6.6 6.8 7 7.2 Magnitude [Ms]
7.4
7.6
7.8
Fig. 4.13: Cumulative frequency distribution of records used for magnitude scaling of two-stage method
1.2 1.1 1 0.9 0.8 ai 0.7 0.6 0.5 0.4 0.3 0.2 5.5 6 6.5 7 Magnitude [M ]
s
Kocaeli
7.5
Fig. 4.14: Variation of amplitude factors, ai , with magnitude, crosses are those earthquakes which have more than one record associated with them and diamonds are for one record earthquakes. Solid line is the magnitude dependence obtained from the two-stage method ignoring the earthquakes with only one record, dashed line is from the two-stage method including all the earthquakes and dash-dotted line is from the one-stage method.
4. Figures
30
15
10
PGA [ms ]
5 0 0
10
15
Distance [km]
Fig. 4.15: Predicted horizontal PGA for Ms = 5.8, 6.8 and 7.8 (solid line) and the 84-percentile accelerations (dashed line).
1.0 2,S 0.5
2,A 2,R 2,A
2,S 2,R
0.2
0.1
0.05
1,S 1,A 1,R 1,S 1,A
1,R 0.02 0.1 0.3 0.6 1 3 6 10 30
60 100
Fault Distance [km]
Fig. 4.16: Comparision between the attenuation of horizontal peak ground acceleration for Ms = 6 and 7 for the three soil types (R,A and S) predicted by Ambraseys et al. (1996) equations (labelled 1) and the new near-eld equations presented in this study (labelled 2).
4. Figures
31
15
10
PGA [ms ]
5 0 0
10
15
Distance [km]
Fig. 4.17: Predicted vertical PGA for Ms = 5.8, 6.8 and 7.8 (solid line) and the 84-percentile accelerations (dashed line).
1.0
0.5
2,S 2,A 2,S 2,A 2,R 2,R
0.2
0.1
0.05
0.02 1,R 0.01 0.1 0.3 0.6 1 3 6 10 30
1,S 1,A 1,R 1,S 1,A
60 100
Fault Distance [km]
Fig. 4.18: Comparision between the attenuation of vertical peak ground acceleration for Ms = 6 and 7 for the three soil types (R,A and S) predicted by Ambraseys & Simpson (1996) equations (labelled 1) and the new near-eld equations presented in this study (labelled 2).
4. Figures
32
1.4
1.2
0.8
av/ah
0.6 0.4 0.2 0 0
10
15
Distance [km]
Fig. 4.19: Predicted vertical to horizontal PGA ratio, q = av /ah , for all earthquakes (solid line) and the 84-percentile accelerations (dashed line).
1.4
1.2
0.8
av/ah
0.6 0.4 0.2 0 0
10
15
Distance [km]
Fig. 4.20: Predicted vertical to horizontal PGA ratio, q = av /ah , for thrust earthquakes (solid line) and the 84-percentile accelerations (dashed line).
4. Figures
33
1.4
1.2
0.8
av/ah
0.6 0.4 0.2 0 0
10
15
Distance [km]
Fig. 4.21: Predicted vertical to horizontal PGA ratio, q = av /ah for Ms = 5.8, 6.8 and 7.8, for European earthquakes (solid line) and the 84-percentile accelerations (dashed line).
8000
6000
Energy density [cm2s1]
4000
A
2000
S A R
0 0
10
15
Distance [km]
Fig. 4.22: Predicted horizontal (top set of curves) and vertical energy density, for the three different soil categories, for M = 7.
4. Figures
34
0.6
0.5
0.4
S A A R R S
0.3
0.2
0.1
0.5
1.5
Period [s]
Fig. 4.23: Horizontal (top set of curves) and vertical absolute acceleration spectra for rock (R), stiff (A) and soft soil (S) for an earthquake of Ms = 6.0 at a fault distance of 10 km using the new equations.
1.2
0.8
S A
0.6
S A R
0.4
0.2
0.5
1.5
Period [s]
Fig. 4.24: Horizontal (top set of curves) and vertical absolute acceleration spectra for rock (R), stiff (A) and soft soil (S) for an earthquake of Ms = 7.0 at a fault distance of 10 km using the new equations.
4. Figures
35
1.1
Ratio of bending SA to nonbending SA
1.05
0.95 0
0.2
0.4
0.6
0.8
1.2
1.4
1.6
1.8
Period [sec]
Fig. 4.25: Ratio between the predicted spectral acceleration when the vertical ground motion is included and the predicted spectral acceleration when it is ignored for Ms = 5.8, 6.8 and 7.8 at a distance 0 km.
8000
Absolute unit input energy [cm2s2]
6000
S
4000
A
2000
A S R
0 0
0.5
1.5
Period [s]
Fig. 4.26: Predicted horizontal (top set of curves) and vertical maximum absolute input energy, for the three soil types, for M = 7 and d = 10 km.
4. Figures
36
1 0.9 0.8 0.7 0.6
qs
0.5 0.4 0.3 0.2 0.1 0 0
0.5
1.5
Period [s]
Fig. 4.27: Predicted vertical to horizontal spectral ratio, qs = SAv /SAh (top set of curves) and simultaneous ratio, qi = Rv (tmax )/SAh for different types of faulting. All earthquakes (solid line), normal (dashed line), thrust (dotted line) and strike-slip (dash-dotted line).
1 0.8 0.6
Qi
0.4 0.2 0 2 1.5 1 0.5 0.5 0 0 1.5 1 2
T [sec]
h
T [sec]
v
Fig. 4.28: Predicted vertical to horizontal simultaneous spectral ratio, Qi = Rv (tmax )/SAh .
qi
4. Figures
37
1.5
0.5
0 2 1.5 1 0.5 0.5 0 0 1.5 1 2
T [sec]
h
T [sec]
v
Fig. 4.29: Standard error of prediction, , of vertical to horizontal simultaneous spectral ratio, Qi = Rv (tmax )/SAh .
1 0.9 0.8 0.7 0.6
qe
0.5 0.4 0.3 0.2 0.1 0 0
0.5
1.5
Period [s]
Fig. 4.30: Predicted vertical to horizontal maximum absolute input energy ratio, qe = Iv /Ih for different types of faulting. All earthquakes (solid line), normal (dashed line), thrust (dotted line) and strike-slip (dash-dotted line).
5. TABLES
Tab. 5.1: Coefcients of horizontal spectral acceleration relations (T is the natural period) T 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 b1 0.028 0.008 0.132 0.114 0.100 0.110 0.008 0.036 0.083 0.101 0.182 0.289 0.570 0.652 0.587 0.554 0.584 0.559 0.543 0.610 0.714 0.812 0.903 0.971 0.955 0.992 1.113 1.100 1.054 1.088 1.182 1.236 1.318 1.406 1.540 1.726 1.980 2.265 b2 0.143 0.144 0.127 0.131 0.135 0.135 0.149 0.155 0.160 0.163 0.175 0.189 0.226 0.241 0.235 0.231 0.235 0.232 0.230 0.236 0.246 0.257 0.268 0.278 0.273 0.275 0.289 0.285 0.275 0.280 0.291 0.293 0.301 0.310 0.325 0.347 0.372 0.412 b3 bA 0.0238 0.042 0.0221 0.020 0.0215 0.016 0.0202 0.009 0.0191 0.011 0.0189 0.001 0.0175 0.004 0.0169 0.004 0.0157 0.004 0.0151 0.006 0.0164 0.006 0.0175 0.033 0.0176 0.081 0.0196 0.084 0.0226 0.064 0.0251 0.057 0.0259 0.048 0.0271 0.039 0.0272 0.043 0.0265 0.059 0.0263 0.086 0.0262 0.107 0.0262 0.114 0.0258 0.103 0.0257 0.102 0.0252 0.110 0.0275 0.126 0.0311 0.133 0.0331 0.129 0.0352 0.118 0.0352 0.113 0.0342 0.123 0.0350 0.142(*) 0.0335 0.145(*) 0.0321 0.142(*) 0.0307 0.153(*) 0.0293 0.186(*) 0.0307 0.173(*) continued on next page bS 0.014 0.010 0.016 0.010 0.000 0.001 0.005 0.025 0.035 0.034 0.049 0.066 0.114(*) 0.103 0.107 0.117(*) 0.120(*) 0.108 0.094 0.095 0.119(*) 0.134(*) 0.148(*) 0.150(*) 0.162(*) 0.178(*) 0.202(*) 0.226(*) 0.241(*) 0.234(*) 0.220(*) 0.222(*) 0.221(*) 0.216(*) 0.208(*) 0.220(*) 0.268(*) 0.252(*) 0.240 0.246 0.245 0.247 0.250 0.251 0.249 0.250 0.250 0.250 0.251 0.248 0.243 0.238 0.247 0.251 0.258 0.252 0.252 0.255 0.256 0.252 0.252 0.252 0.251 0.253 0.255 0.257 0.263 0.266 0.264 0.268 0.268 0.267 0.268 0.272 0.275 0.271
5. Tables
39
Tab. 5.1: continued T 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 2.505 2.725 2.904 3.052 3.127 3.206 3.300 3.380 b2 0.442 0.470 0.492 0.510 0.517 0.525 0.534 0.543 b3 0.0311 0.0297 0.0298 0.0304 0.0306 0.0316 0.0319 0.0326 bA 0.158(*) 0.139 0.128 0.123 0.114 0.107 0.102 0.098 bS 0.242(*) 0.229(*) 0.225(*) 0.221(*) 0.213(*) 0.213(*) 0.215(*) 0.215(*) 0.271 0.274 0.276 0.271 0.265 0.264 0.264 0.262
Tab. 5.2: Coefcients of vertical spectral acceleration relations (T is the natural period) T 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 b1 0.513 0.479 0.596 0.576 0.630 0.706 0.725 0.696 0.784 0.819 0.858 0.866 0.958 0.946 1.002 1.106 1.239 1.388 1.440 1.489 1.547 1.586 1.594 1.563 1.521 1.524 1.621 1.700 1.675 1.700 1.855 1.973 2.006 2.040 b2 0.209 0.202 0.218 0.213 0.218 0.226 0.231 0.227 0.236 0.240 0.241 0.238 0.245 0.239 0.244 0.261 0.277 0.298 0.303 0.303 0.309 0.312 0.312 0.305 0.299 0.302 0.316 0.324 0.317 0.318 0.337 0.348 0.349 0.352 b3 bA 0.0287 0.025 0.0297 0.033 0.0291 0.036 0.0270 0.032 0.0271 0.053 0.0268 0.070 0.0278 0.052 0.0297 0.047 0.0296 0.058 0.0288 0.053 0.0275 0.056 0.0269 0.051 0.0282 0.092 0.0273 0.096 0.0261 0.080 0.0265 0.050 0.0265 0.043 0.0274 0.040 0.0286 0.061 0.0291 0.105 0.0292 0.108(*) 0.0289 0.109(*) 0.0289 0.098 0.0296 0.099 0.0308 0.097 0.0325 0.081 0.0359 0.074 0.0364 0.061 0.0385 0.068 0.0380 0.065 0.0364 0.057 0.0358 0.064 0.0352 0.043 0.0346 0.031 continued on next page bS 0.113 0.132 0.136 0.109 0.116 0.118 0.087 0.082 0.090 0.068 0.066 0.041 0.070 0.041 0.033 0.012 0.019 0.025 0.034 0.053 0.043 0.047 0.038 0.037 0.030 0.015 0.001 0.005 0.023 0.002 0.006 0.012 0.009 0.013 0.308 0.308 0.303 0.296 0.295 0.287 0.291 0.290 0.288 0.286 0.282 0.275 0.264 0.265 0.264 0.256 0.256 0.258 0.255 0.255 0.255 0.253 0.255 0.252 0.248 0.243 0.244 0.253 0.253 0.263 0.258 0.252 0.256 0.257
5. Tables
40
Tab. 5.2: continued T 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 2.185 2.294 2.482 2.544 2.580 2.758 2.981 3.120 3.227 3.368 3.537 3.680 b2 0.370 0.384 0.406 0.411 0.412 0.435 0.466 0.481 0.492 0.509 0.529 0.543 b3 0.0338 0.0335 0.0333 0.0334 0.0324 0.0309 0.0292 0.0285 0.0301 0.0315 0.0310 0.0304 bA 0.032 0.028 0.029 0.011 0.010 0.044 0.072 0.076 0.062 0.060 0.057 0.042 bS 0.018 0.021 0.020 0.007 0.023 0.036 0.055 0.054 0.034 0.019 0.011 0.004 0.259 0.259 0.259 0.272 0.271 0.280 0.285 0.290 0.288 0.287 0.287 0.290
Tab. 5.3: Coefcients of horizontal spectral acceleration (buckling model) relations (T is the natural period) T 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 b1 0.078 0.118 0.015 0.004 0.008 0.030 0.058 0.084 0.149 0.179 0.247 0.355 0.627 0.712 0.641 0.632 0.662 0.645 0.639 0.709 0.815 0.913 1.013 1.078 1.053 1.086 1.219 1.189 1.149 1.175 b2 0.155 0.158 0.141 0.146 0.148 0.147 0.161 0.166 0.173 0.175 0.186 0.202 0.237 0.250 0.244 0.243 0.246 0.245 0.244 0.251 0.261 0.272 0.284 0.293 0.286 0.287 0.302 0.297 0.289 0.293 b3 bA 0.0240 0.013 0.0214 0.004 0.0212 0.006 0.0199 0.000 0.0190 0.006 0.0189 0.002 0.0183 0.004 0.0177 0.017 0.0161 0.007 0.0156 0.008 0.0169 0.010 0.0186 0.030 0.0188 0.079 0.0204 0.090 0.0233 0.061 0.0260 0.063 0.0264 0.058 0.0278 0.049 0.0282 0.054 0.0275 0.069 0.0266 0.098 0.0272 0.120 0.0271 0.128 0.0265 0.118 0.0267 0.125 0.0261 0.142(*) 0.0281 0.155(*) 0.0320 0.155(*) 0.0342 0.149(*) 0.0363 0.138 continued on next page bS 0.016 0.040 0.043 0.022 0.008 0.004 0.000 0.007 0.023 0.040 0.056 0.067 0.117(*) 0.111(*) 0.108 0.130(*) 0.137(*) 0.124(*) 0.111 0.109 0.132(*) 0.154(*) 0.171(*) 0.173(*) 0.189(*) 0.214(*) 0.234(*) 0.251(*) 0.265(*) 0.256(*) 0.235 0.241 0.242 0.247 0.250 0.255 0.252 0.254 0.253 0.252 0.254 0.253 0.248 0.242 0.251 0.254 0.259 0.256 0.259 0.258 0.257 0.255 0.255 0.254 0.253 0.254 0.256 0.260 0.267 0.270
5. Tables
41
Tab. 5.3: continued T 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 1.237 1.300 1.349 1.438 1.586 1.769 2.002 2.277 2.540 2.759 2.933 3.100 3.185 3.262 3.349 3.432 b2 0.299 0.301 0.305 0.314 0.333 0.354 0.377 0.416 0.451 0.478 0.499 0.519 0.527 0.534 0.544 0.552 b3 0.0361 0.0349 0.0357 0.0344 0.0330 0.0317 0.0305 0.0321 0.0323 0.0310 0.0308 0.0312 0.0317 0.0327 0.0332 0.0338 bA 0.134 0.143 0.160(*) 0.162(*) 0.151(*) 0.163(*) 0.191(*) 0.172(*) 0.158(*) 0.135 0.128 0.125 0.121 0.115 0.109 0.108 bS 0.244(*) 0.244(*) 0.241(*) 0.237(*) 0.222(*) 0.232(*) 0.271(*) 0.250(*) 0.243(*) 0.224(*) 0.224(*) 0.221(*) 0.221(*) 0.223(*) 0.225(*) 0.228(*) 0.266 0.271 0.272 0.270 0.271 0.276 0.279 0.275 0.274 0.277 0.279 0.274 0.268 0.268 0.268 0.266
Tab. 5.4: Percentage increase in response due to vertical accelerations found by Loh & Ma (1997) Period T 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.5 2.0 SAh (Ignoring vertical) 14.9 21.0 23.5 19.7 18.4 16.2 15.9 15.2 14.6 13.7 9.5 7.0 SAh (Including vertical) 27.9 31.7 29.2 28.6 24.1 22.9 20.3 18.4 17.8 16.5 11.7 8.9 % increase 87 51 24 45 31 41 28 21 22 20 23 27
Tab. 5.5: Coefcients of horizontal maximum absolute input energy relations (T is the natural period) T 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 b1 0.874 0.774 0.637 0.416 0.267 0.181 0.153 0.187 b2 0.613 0.598 0.582 0.556 0.538 0.529 0.531 0.538 b3 bA 0.0593 0.041 0.0569 0.070 0.0559 0.088 0.0528 0.066 0.0512 0.069 0.0492 0.082 0.0485 0.056 0.0458 0.046 continued on next page bS 0.169 0.186(*) 0.180(*) 0.138 0.132 0.133 0.116 0.129 0.397 0.392 0.402 0.408 0.405 0.413 0.420 0.420
5. Tables
42
Tab. 5.5: continued T 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 0.097 0.085 0.156 0.381 0.514 0.583 0.490 0.381 0.412 0.369 0.327 0.317 0.429 0.477 0.620 0.662 0.674 0.743 0.787 0.808 0.623 0.500 0.519 0.565 0.605 0.758 0.977 1.291 1.793 2.139 2.449 2.780 3.198 3.554 3.613 3.589 3.641 3.700 b2 0.526 0.527 0.539 0.579 0.598 0.614 0.605 0.600 0.609 0.608 0.604 0.604 0.617 0.622 0.645 0.652 0.652 0.662 0.667 0.675 0.651 0.640 0.639 0.640 0.648 0.672 0.699 0.740 0.805 0.855 0.900 0.945 1.005 1.055 1.062 1.059 1.066 1.074 b3 0.0442 0.0436 0.0426 0.0429 0.0430 0.0437 0.0459 0.0504 0.0511 0.0500 0.0500 0.0515 0.0512 0.0516 0.0502 0.0491 0.0476 0.0471 0.0497 0.0554 0.0616 0.0645 0.0640 0.0625 0.0629 0.0605 0.0582 0.0560 0.0550 0.0568 0.0559 0.0541 0.0509 0.0501 0.0530 0.0562 0.0575 0.0593 bA 0.048 0.055 0.052 0.060 0.122 0.142 0.143 0.120 0.106 0.102 0.128 0.155 0.195 0.221 0.208 0.208 0.213 0.215 0.265(*) 0.240 0.263(*) 0.241 0.243 0.274(*) 0.276(*) 0.266(*) 0.281(*) 0.296(*) 0.337(*) 0.349(*) 0.312(*) 0.293(*) 0.262 0.243 0.239 0.219 0.209 0.206 bS 0.149 0.159 0.171 0.166 0.233(*) 0.242(*) 0.268(*) 0.271(*) 0.272(*) 0.233(*) 0.216(*) 0.223(*) 0.257(*) 0.286(*) 0.286(*) 0.296(*) 0.320(*) 0.344(*) 0.414(*) 0.432(*) 0.476(*) 0.451(*) 0.460(*) 0.456(*) 0.435(*) 0.413(*) 0.424(*) 0.440(*) 0.521(*) 0.545(*) 0.504(*) 0.504(*) 0.487(*) 0.485(*) 0.479(*) 0.470(*) 0.460(*) 0.464(*) 0.416 0.412 0.412 0.423 0.424 0.412 0.425 0.435 0.434 0.441 0.444 0.453 0.458 0.461 0.465 0.470 0.473 0.478 0.473 0.480 0.489 0.499 0.499 0.487 0.499 0.501 0.507 0.523 0.527 0.530 0.517 0.510 0.502 0.497 0.496 0.491 0.492 0.487
Tab. 5.6: Coefcients of vertical maximum absolute input energy relations (T is the natural period) T 0.10 0.11 0.12 0.13 b1 0.986 1.045 1.007 0.899 b2 0.572 0.584 0.584 0.569 b3 bA 0.0585 0.009 0.0592 0.034 0.0578 0.038 0.0549 0.047 continued on next page bS 0.177 0.225 0.219 0.202 0.477 0.494 0.501 0.502
5. Tables
43
Tab. 5.6: continued T 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 0.905 0.928 0.939 1.019 1.046 1.060 1.096 1.067 0.987 1.074 1.192 1.327 1.505 1.587 1.586 1.613 1.673 1.711 1.676 1.598 1.539 1.532 1.526 1.615 1.526 1.497 1.589 1.719 1.679 1.739 1.935 2.044 2.161 2.321 2.553 2.847 3.177 3.437 3.515 3.644 3.816 3.997 b2 0.571 0.576 0.580 0.594 0.599 0.599 0.606 0.604 0.590 0.601 0.619 0.644 0.668 0.679 0.678 0.680 0.689 0.697 0.692 0.680 0.672 0.675 0.687 0.705 0.688 0.689 0.699 0.716 0.710 0.720 0.747 0.762 0.773 0.803 0.838 0.881 0.932 0.970 0.980 0.996 1.020 1.042 b3 0.0534 0.0522 0.0528 0.0542 0.0544 0.0536 0.0508 0.0512 0.0528 0.0510 0.0506 0.0497 0.0489 0.0500 0.0512 0.0520 0.0528 0.0530 0.0533 0.0550 0.0554 0.0571 0.0628 0.0643 0.0664 0.0676 0.0652 0.0634 0.0619 0.0606 0.0602 0.0600 0.0586 0.0598 0.0571 0.0536 0.0516 0.0543 0.0573 0.0583 0.0593 0.0582 bA 0.084 0.093 0.112 0.108 0.110 0.141 0.112 0.106 0.143 0.152 0.127 0.088 0.083 0.105 0.148 0.174 0.184 0.177 0.184 0.200(*) 0.198 0.177 0.109 0.082 0.151 0.151 0.144 0.115 0.076 0.065 0.085 0.090 0.098 0.045 0.004 0.054 0.079 0.070 0.079 0.081 0.085 0.070 bS 0.204 0.194 0.191 0.193 0.189 0.198 0.154 0.118 0.120 0.105 0.102 0.064 0.072 0.102 0.123 0.127 0.122 0.116 0.114 0.124 0.121 0.098 0.044 0.035 0.078 0.055 0.067 0.061 0.045 0.041 0.082 0.105 0.131 0.072 0.032 0.001 0.010 0.006 0.038 0.067 0.075 0.077 0.505 0.505 0.512 0.509 0.509 0.505 0.499 0.485 0.475 0.462 0.461 0.457 0.462 0.470 0.473 0.469 0.465 0.458 0.460 0.463 0.458 0.457 0.457 0.466 0.472 0.474 0.476 0.469 0.460 0.471 0.486 0.488 0.491 0.506 0.501 0.506 0.521 0.535 0.542 0.537 0.531 0.528
Coefcients of vertical to horizontal spectral ratio, qs , relations (T is the natural period)

Tab. 5.7: All T 0.10 b1 0.052 0.261 Normal Thrust b1 b1 0.152 0.157 0.063 0.245 continued on next page Strike-slip b1 0.017 0.292 European b1 0.150 0.264
5. Tables
Tab. 5.7: continued All T 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 0.080 0.110 0.140 0.160 0.193 0.215 0.246 0.264 0.285 0.300 0.337 0.364 0.385 0.412 0.421 0.430 0.439 0.446 0.459 0.474 0.478 0.485 0.492 0.498 0.502 0.514 0.527 0.525 0.520 0.511 0.513 0.518 0.511 0.499 0.496 0.494 0.499 0.485 0.476 0.465 0.461 0.452 0.447 0.443 0.444 0.262 0.266 0.258 0.258 0.258 0.267 0.263 0.261 0.256 0.250 0.242 0.239 0.233 0.229 0.229 0.228 0.233 0.239 0.243 0.250 0.248 0.242 0.239 0.240 0.240 0.250 0.253 0.249 0.271 0.267 0.260 0.264 0.266 0.267 0.270 0.274 0.285 0.285 0.290 0.295 0.292 0.280 0.272 0.259 0.254 Normal b1 0.121 0.162 0.194 0.229 0.260 0.274 0.268 0.267 0.285 0.303 0.336 0.326 0.311 0.363 0.363 0.372 0.378 0.387 0.397 0.414 0.405 0.371 0.354 0.327 0.334 0.364 0.426 0.430 0.385 0.359 0.379 0.397 0.391 0.347 0.335 0.347 0.338 0.310 0.255 0.235 0.229 0.227 0.229 0.247 0.257 0.141 0.178 0.182 0.189 0.187 0.198 0.182 0.165 0.157 0.156 0.151 0.191 0.212 0.217 0.224 0.218 0.215 0.214 0.216 0.205 0.174 0.167 0.174 0.154 0.141 0.167 0.194 0.176 0.153 0.150 0.125 0.138 0.153 0.153 0.161 0.188 0.212 0.209 0.221 0.199 0.196 0.197 0.204 0.207 0.193 Thrust b1 0.106 0.140 0.165 0.181 0.213 0.233 0.266 0.279 0.296 0.305 0.341 0.366 0.388 0.415 0.419 0.438 0.449 0.451 0.464 0.479 0.486 0.503 0.513 0.526 0.529 0.537 0.544 0.535 0.540 0.538 0.545 0.554 0.546 0.538 0.540 0.533 0.540 0.523 0.517 0.503 0.488 0.472 0.463 0.450 0.450 0.243 0.252 0.231 0.232 0.240 0.255 0.251 0.256 0.252 0.248 0.243 0.243 0.230 0.215 0.214 0.207 0.201 0.206 0.215 0.226 0.227 0.220 0.218 0.224 0.226 0.233 0.248 0.254 0.281 0.277 0.274 0.280 0.283 0.279 0.284 0.289 0.285 0.282 0.279 0.281 0.272 0.260 0.251 0.241 0.234 Strike-slip b1 0.038 0.300 0.061 0.293 0.096 0.299 0.118 0.298 0.152 0.289 0.178 0.292 0.215 0.290 0.243 0.285 0.270 0.279 0.293 0.269 0.334 0.258 0.368 0.246 0.396 0.242 0.419 0.250 0.435 0.249 0.431 0.257 0.438 0.274 0.451 0.283 0.465 0.281 0.480 0.288 0.482 0.285 0.484 0.277 0.494 0.268 0.496 0.263 0.500 0.262 0.516 0.277 0.525 0.267 0.533 0.253 0.523 0.271 0.507 0.265 0.499 0.253 0.495 0.254 0.489 0.255 0.479 0.257 0.470 0.255 0.473 0.258 0.479 0.287 0.472 0.292 0.468 0.298 0.462 0.312 0.473 0.315 0.473 0.302 0.471 0.293 0.474 0.276 0.475 0.277
44
European b1 0.168 0.251 0.210 0.272 0.231 0.261 0.254 0.254 0.285 0.243 0.304 0.243 0.324 0.234 0.335 0.229 0.350 0.220 0.351 0.211 0.343 0.184 0.357 0.190 0.368 0.191 0.387 0.180 0.365 0.180 0.365 0.180 0.374 0.185 0.387 0.193 0.400 0.193 0.408 0.201 0.399 0.207 0.395 0.205 0.389 0.208 0.383 0.203 0.394 0.203 0.423 0.223 0.443 0.242 0.456 0.244 0.428 0.237 0.405 0.236 0.407 0.235 0.423 0.247 0.427 0.269 0.404 0.266 0.394 0.267 0.383 0.250 0.383 0.252 0.370 0.252 0.351 0.252 0.343 0.261 0.348 0.277 0.355 0.278 0.363 0.265 0.373 0.258 0.378 0.259
Coefcients of vertical to horizontal simultaneous spectral ratio, qi , relations (T is the natural period)
Tab. 5.8: All T 0.10 0.11 0.12 0.13 0.14 0.15 b1 0.860 0.863 0.955 0.891 0.823 0.921 0.512 0.577 1.236 0.576 0.483 0.594 Normal b1 1.067 0.919 1.124 0.858 0.755 0.951 Thrust b1 0.656 0.836 0.518 0.684 0.893 0.581 0.904 0.843 0.521 0.486 0.848 0.497 0.422 0.861 0.482 0.416 0.924 0.606 continued on next page Strike-slip b1 0.849 0.467 0.810 0.553 1.069 1.832 0.955 0.684 0.786 0.497 0.910 0.616 European b1 0.966 0.618 0.810 0.530 0.962 0.699 0.880 0.489 0.949 0.548 0.913 0.506
5. Tables
Tab. 5.8: continued All T 0.16 0.17 0.18 0.19 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 0.912 0.997 0.921 0.928 0.960 0.944 1.029 1.008 0.983 1.079 1.104 1.118 1.051 1.067 1.117 1.047 1.041 1.125 1.089 1.137 1.123 1.154 1.189 1.129 1.123 1.095 1.140 1.164 1.067 1.099 1.063 1.078 1.053 1.036 1.076 1.049 0.962 0.966 0.990 0.973 0.561 0.533 0.491 0.502 0.555 0.517 0.482 0.506 0.473 0.542 0.540 0.568 0.529 0.564 0.585 0.554 0.569 0.550 0.552 0.527 0.526 0.542 0.511 0.562 0.499 0.528 0.574 1.190 0.539 0.590 0.508 0.560 0.549 0.563 0.569 0.573 0.509 0.539 0.541 0.497 Normal b1 0.864 1.029 1.072 1.100 0.950 0.957 0.863 1.108 1.111 1.152 1.092 1.151 1.025 0.866 1.131 1.099 0.938 1.249 0.932 1.067 1.095 1.060 1.091 0.958 1.015 1.038 0.953 0.962 0.946 0.893 0.805 0.851 0.903 0.775 0.919 0.721 0.767 0.614 0.635 0.777 0.455 0.670 0.776 0.793 0.521 0.355 0.336 0.435 0.681 0.494 0.500 0.615 0.433 0.409 0.642 0.630 0.314 0.598 0.424 0.482 0.575 0.493 0.431 0.403 0.382 0.617 0.443 0.460 0.524 0.600 0.367 0.548 0.680 0.499 0.599 0.365 0.444 0.345 0.456 0.429 Thrust b1 0.928 1.012 0.921 0.882 0.896 0.920 1.023 0.934 0.913 1.055 1.063 1.096 0.998 1.061 1.090 1.051 1.006 1.053 1.069 1.157 1.047 1.135 1.171 1.075 1.141 1.102 1.133 1.092 1.085 1.104 1.084 1.113 1.135 1.067 1.094 1.098 1.002 0.999 1.053 1.025 0.628 0.553 0.488 0.431 0.505 0.575 0.499 0.574 0.468 0.598 0.526 0.542 0.530 0.554 0.572 0.555 0.563 0.529 0.598 0.555 0.475 0.541 0.523 0.521 0.553 0.561 0.601 0.496 0.541 0.569 0.521 0.494 0.568 0.541 0.521 0.527 0.522 0.505 0.557 0.522 Strike-slip b1 0.900 0.485 0.972 0.479 0.888 0.417 0.953 0.514 1.047 0.618 0.974 0.465 1.070 0.482 1.084 0.404 1.048 0.417 1.096 0.473 1.162 0.569 1.140 0.600 1.126 0.542 1.117 0.602 1.149 0.596 1.030 0.544 1.109 0.614 1.195 0.561 1.149 0.508 1.125 0.503 1.231 0.567 1.199 0.556 1.234 0.511 1.237 0.625 1.123 0.445 1.096 0.467 1.188 0.558 1.302 1.796 1.069 0.544 1.134 0.616 1.088 0.506 1.077 0.636 0.975 0.480 1.049 0.596 1.085 0.624 1.052 0.646 0.948 0.501 0.994 0.592 0.979 0.512 0.945 0.469
45
European b1 0.950 0.541 1.019 0.561 0.937 0.370 1.047 0.626 0.962 0.553 1.011 0.641 0.943 0.420 0.989 0.495 0.986 0.530 1.123 0.537 1.031 0.513 0.940 0.407 1.000 0.438 0.929 0.467 0.979 0.568 0.948 0.520 0.866 0.389 0.960 0.414 0.911 0.464 0.917 0.401 0.936 0.483 0.973 0.527 1.154 0.528 1.113 0.658 0.975 0.478 0.952 0.562 0.964 0.491 0.916 0.431 0.832 0.407 0.933 0.544 0.867 0.447 0.906 0.512 0.856 0.539 0.812 0.483 0.911 0.611 0.800 0.391 0.811 0.644 0.764 0.458 0.809 0.471 0.896 0.480
Coefcients of vertical to horizontal maximum absolute input energy ratio, qe , relations (T is the natural period)
Tab. 5.9: All T 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 b1 0.385 0.372 0.369 0.392 0.395 0.421 0.439 0.476 0.505 0.525 0.550 0.365 0.375 0.378 0.369 0.363 0.371 0.377 0.373 0.375 0.378 0.374 Normal b1 0.401 0.400 0.435 0.461 0.469 0.492 0.475 0.480 0.509 0.513 0.517 Thrust b1 0.261 0.403 0.337 0.269 0.392 0.352 0.277 0.378 0.355 0.318 0.391 0.345 0.286 0.388 0.347 0.281 0.421 0.363 0.304 0.431 0.367 0.312 0.454 0.371 0.271 0.476 0.393 0.260 0.494 0.395 0.252 0.514 0.385 continued on next page Strike-slip b1 0.358 0.419 0.339 0.423 0.343 0.426 0.379 0.412 0.390 0.400 0.406 0.399 0.444 0.406 0.505 0.389 0.543 0.370 0.568 0.375 0.606 0.377 European b1 0.442 0.323 0.452 0.338 0.453 0.352 0.472 0.370 0.490 0.351 0.514 0.357 0.510 0.353 0.554 0.347 0.575 0.365 0.580 0.370 0.562 0.346
5. Tables
Tab. 5.9: continued All T 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 0.42 0.44 0.46 0.48 0.50 0.55 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.10 1.20 1.30 1.40 1.50 1.60 1.70 1.80 1.90 2.00 b1 0.598 0.652 0.691 0.733 0.761 0.793 0.810 0.818 0.835 0.852 0.853 0.861 0.879 0.891 0.908 0.950 0.936 0.935 0.928 0.919 0.916 0.923 0.922 0.899 0.884 0.892 0.897 0.881 0.871 0.847 0.839 0.825 0.800 0.795 0.797 0.353 0.364 0.343 0.348 0.357 0.357 0.374 0.391 0.401 0.405 0.408 0.403 0.403 0.408 0.419 0.434 0.429 0.442 0.460 0.459 0.450 0.450 0.462 0.462 0.458 0.463 0.496 0.481 0.502 0.503 0.499 0.485 0.471 0.463 0.453 Normal b1 0.563 0.597 0.580 0.654 0.686 0.678 0.714 0.761 0.787 0.755 0.710 0.673 0.665 0.637 0.615 0.666 0.714 0.744 0.704 0.685 0.725 0.761 0.768 0.724 0.696 0.627 0.621 0.584 0.509 0.448 0.415 0.422 0.428 0.451 0.463 0.228 0.250 0.251 0.367 0.356 0.301 0.318 0.306 0.277 0.311 0.349 0.347 0.337 0.310 0.314 0.345 0.349 0.336 0.302 0.305 0.292 0.316 0.332 0.313 0.346 0.307 0.315 0.326 0.372 0.408 0.423 0.442 0.385 0.357 0.374 Thrust b1 0.567 0.603 0.649 0.705 0.736 0.774 0.791 0.784 0.802 0.824 0.836 0.860 0.889 0.908 0.928 0.964 0.941 0.932 0.939 0.938 0.938 0.942 0.941 0.933 0.923 0.932 0.926 0.919 0.910 0.871 0.858 0.853 0.823 0.818 0.810 0.372 0.375 0.343 0.340 0.354 0.347 0.347 0.360 0.366 0.372 0.380 0.376 0.377 0.385 0.398 0.406 0.425 0.434 0.459 0.458 0.456 0.460 0.471 0.478 0.473 0.466 0.499 0.485 0.486 0.476 0.455 0.429 0.425 0.427 0.417 Strike-slip b1 0.646 0.348 0.729 0.360 0.769 0.348 0.785 0.351 0.808 0.360 0.843 0.376 0.854 0.415 0.875 0.442 0.888 0.462 0.909 0.457 0.905 0.450 0.901 0.440 0.910 0.439 0.922 0.441 0.941 0.445 0.990 0.469 0.975 0.440 0.979 0.466 0.959 0.480 0.943 0.479 0.926 0.463 0.931 0.458 0.928 0.471 0.891 0.463 0.871 0.453 0.895 0.472 0.915 0.511 0.892 0.486 0.894 0.522 0.898 0.524 0.902 0.532 0.871 0.530 0.848 0.515 0.837 0.503 0.848 0.491
46
European b1 0.544 0.282 0.612 0.268 0.631 0.253 0.637 0.280 0.634 0.302 0.638 0.276 0.685 0.302 0.744 0.305 0.743 0.292 0.719 0.312 0.702 0.345 0.699 0.348 0.705 0.368 0.707 0.373 0.711 0.395 0.773 0.409 0.780 0.443 0.803 0.475 0.793 0.443 0.792 0.443 0.799 0.429 0.809 0.457 0.821 0.505 0.788 0.509 0.766 0.510 0.725 0.455 0.736 0.438 0.718 0.424 0.676 0.428 0.662 0.465 0.663 0.504 0.667 0.485 0.653 0.464 0.654 0.470 0.655 0.475
BIBLIOGRAPHY
Abrahamson, N.A., & Litehiser, J.J. 1989. Attenuation of vertical peak acceleration. Bulletin of the Seismological Society of America, 79(3), 549580. Ambraseys, N.N. 1995. The prediction of earthquake peak ground acceleration in Europe. Earthquake Engineering and Structural Dynamics, 24(4), 467490. Ambraseys, N.N., & Bommer, J.J. 1991. The attenuation of ground accelerations in Europe. Earthquake Engineering and Structural Dynamics, 20(12), 11791202. Ambraseys, N.N., & Jackson, J.A. 1998. Faulting associated with historical and recent earthquakes in the Eastern Mediterranean region. Geophysical Journal International, 133, 390406. Ambraseys, N.N., & Menu, J. 1988. Earthquake-induced ground displacements. Earthquake Engineering and Structural Dynamics, 16, 9851006. Ambraseys, N.N., & Simpson, K.A. 1996. Prediction of vertical response spectra in Europe. Earthquake Engineering and Structural Dynamics, 25(4), 401412. Ambraseys, N.N., Simpson, K.A., & Bommer, J.J. 1996. Prediction of horizontal response spectra in Europe. Earthquake Engineering and Structural Dynamics, 25(4), 371400. Benedetti, D., & Carydis, P. 1999. Inuence of the vertical component on damage during shallow-neareld earthquakes. European Earthquake Engineering, 13, 312. Berrill, J.B. 1975. A study of high-frequency strong ground motion from the San Fernando earthquake. Ph.D. thesis, California Institute of Technology, Pasadena, California. Bolt, B.A., & Abrahamson, N.A. 1982. New attenuation relations for peak and expected accelerations of strong ground motion. Bulletin of the Seismological Society of America, 72(6), 23072321. Bozorgnia, Y., & Niazi, M. 1993. Distance scaling of vertical and horizontal response spectra and the Loma Prieta earthquake. Earthquake Engineering and Structural Dynamics, 22, 695707. Bozorgnia, Y., Niazi, M., & Campbell, K.W. 1995. Characteristics of free-eld vertical ground motion during the Northridge earthquake. Earthquake Spectra, 11(4), 515525. Bozorgnia, Y., Campbell, K.W., & Niazi, M. 2000. Observed spectral characteristics of vertical ground motion recorded during worldwide earthquakes from 1957 to 1995. In: Proceedings of Twelfth World Conference on Earthquake Engineering. Paper No. 2671. Brillinger, D.R., & Preisler, H.K. 1984. An exploratory analysis of the Joyner-Boore attenuation data. Bulletin of the Seismological Society of America, 74(4), 14411450. Brillinger, D.R., & Preisler, H.K. 1985. Further analysis of the Joyner-Boore attenuation data. Bulletin of the Seismological Society of America, 75(2), 611614. Brune, J.N., & Anooshehpoor, A. 1991. Foam rubber modeling of the El Centro Terminal Substation building. Earthquake Spectra, 7(1), 4579.
BIBLIOGRAPHY
48
Brune, J.N., Vernon III, F.L., & Simons, R.S. 1982. Strong-motion data recorded in Mexico during the main shock. Pages 319349 of: U.S. Geological Survey Professional Paper 1254, The Imperial Valley, California, Earthquake of October 15, 1979. Campbell, K.W. 1981. Near-source attenuation of peak horizontal acceleration. Bulletin of the Seismological Society of America, 71(6), 20392070. Campbell, K.W., & Bozorgnia, Y. 1994 (Jul). Near-source attenuation of peak horizontal acceleration from worldwide accelerograms recorded from 1957 to 1993. Pages 283292 of: Proceedings of the Fifth U.S. National Conference on Earthquake Engineering, vol. III. Chapman, M.C. 1999. On the use of elastic input energy for seismic hazard analysis. Earthquake Spectra, 15(4), 607635. Chiba, M., Tani, J., & Yamaki, N. 1987. Dynamic stability of liquid lled cylindrical shells under vertical excitation; experimental results. Earthquake Engineering and Structural Dynamics, 15, 2336. Draper, N.R., & Smith, H. 1981. Applied regression analysis. 2nd edn. John Wiley & Sons. Elnashai, A.S., & Papazoglou, A. J. 1997. Procedure and spectra for analysis of RC structures subjected to strong vertical earthquake loads. Journal of Earthquake Engineering, 1(1), 121155. Feng, D., Theofanopoulos, N., & Watabe, M. 1988. Consideration on the design velocity response spectra along the principal axes. Pages 855860 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. II. Fischer, F.D., & Seeber, R. 1988. Dynamic response of vertically excited liquid storage tanks considering liquid-soil interaction. Earthquake Engineering and Structural Dynamics, 16, 329342. Foutch, D.A., & Saadeghvaziri, M.A. 1986. The effect of vertical earthquake motions on the response of highway bridges. Pages 115122 of: Proceedings of the Third ASCE Conference on the Dynamic Response of Structures. Fukushima, Y., & Tanaka, T. 1990. A new attenuation relation for peak horizontal acceleration of strong earthquake ground motion in Japan. Bulletin of the Seismological Society of America, 80(4), 757783. Fukushima, Y., Tanaka, T., & Kataoka, S. 1988. A new attenuation relationship for peak ground acceleration derived from strong-motion accelerograms. Pages 343348 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. II. Haroun, M.A., & Abdelhaz, E.A. 1986. A simplied seismic analysis of rigid base liquid storage tanks under vertical excitation with soil structure interaction. Soil Dynamics and Earthquake Engineering, 5, 217225. Hu, Y.-X., Liu, S.-C., & Dong, W. 1996. Earthquake engineering. 1st edn. E & FN Spon, Imprint of Chapman & Hall, London. Hudson, D.E. 1988. Some recent near-source strong motion accelerograms. Pages 271276 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. II. Husid, R. 1969. The effect of gravity on the collapse of yielding structures with earthquake excitation. Pages A4 3143 of: Proceedings of Fourth World Conference of Earthquake Engineering, vol. II. Jennings, P.C., & Husid, R. 1968. Collapse of yielding structures during earthquakes. Journal of the Engineering Mechanics Division, ASCE, 94(EM5), 10451065. Joyner, W. B., & Boore, D.M. 1993. Methods for regression analysis of strong-motion data. Bulletin of the Seismological Society of America, 83(2), 469487.
BIBLIOGRAPHY
49
Joyner, W.B., & Boore, D.M. 1981. Peak horizontal acceleration and velocity from strong-motion records including records from the 1979 Imperial Valley, California, earthquake. Bulletin of the Seismological Society of America, 71(6), 20112038. Krinitzky, E.L., & Chang, F.K. 1987. State-of-the-art for assessing earthquake hazards in the United States: Parameters for specifying intensity-related earthquake ground motions. Tech. rept. 25. U.S. Army Corps of Engineers Waterways Experiment Station. Not seen. Krinitzsky, E.L., Gould, J.P., & Edinger, P.H. 1993. Fundamentals of Earthquake Resistant Construction. John Wiley & Sons. Lin, Y.K., & Shih, T.-Y. 1980. Column response to horizontal-vertical earthquakes. Journal of the Engineering Mechanics Division, ASCE, 106(EM6), 10991109. Liou, G.S., Penzien, J., & Yeung, R.W. 1988. Response of tension leg platforms to vertical seismic excitations. Earthquake Engineering and Structural Dynamics, 16, 157182. Liu, B.C., & Tadjkbakhsh, I. 1986. Effects of vertical motion on friction driven isolation systems. Earthquake Engineering and Structural Dynamics, 14, 609622. Loh, C.H., & Ma, M.J. 1997. Reliability assessment of structures subjected to horizontal-vertical random earthquake excitations. Structural Safety, 19(1), 153168. Lubkin, S., & Stoker, J.J. 1943. Stability of columns and strings under periodically varying forces. Quarterly of Applied Mathematics, 1, 215236. Martnez-Pereira, A. 1999. The characterisation of near-eld earthquake ground-motions for engineering design. Ph.D. thesis, University of London. Martnez-Pereira, A., & Bommer, J.J. 1998 (Mar). What is the near-eld? Pages 245252 of: Booth, E. (ed), Proceedings of the Sixth SECED Conference on Seismic Design Practice into the Next Century. Menu, J.M.H. 1986. Engineering study of near-eld earthquake strong-motions. Ph.D. thesis, University of London. Niazi, M., & Bozorgnia, Y. 1992. Behaviour of near-source vertical and horizontal response spectra at SMART-1 array, Taiwan. Earthquake Engineering and Structural Dynamics, 21, 3750. Nigam, N.C., & Jennings, P.C. 1969. Calculation of response spectra from strong-motion earthquake records. Bulletin of the Seismological Society of America, 59(2), 909922. Nisar, A., & Golesorkhi, C. 1995. Development of vertical design response spectrum for use in near-eld. Pages 10751082 of: Proceedings of the Fifth International Conference on Seismic Zonation, vol. II. Ohno, S., Konno, T., Abe, K., Masao, T., Iizuka, S., & Uebayashi, H. 1996. Method of evaluating horizontal and vertical earthquake ground motions for aseismic design. In: Proceedings of Eleventh World Conference on Earthquake Engineering. Paper No. 1791. Orabi, I. I., & Ahmadi, G. 1988. Response of structures subjected to horizontal-vertical random earthquake excitations. Soil Dynamics and Earthquake Engineering, 7(1), 914. Orabi, I. I., Ahmadi, G., & Su, L. 1989. Hysteretic column under earthquake excitations. Journal of the Engineering Mechanics Division, ASCE, 115(1), 3351. Press, W.H., & Teukolsky, S.A. 1992. Adaptive stepsize Runge-Kutta. Computers in Physics, 6(2), 188191.
BIBLIOGRAPHY
50
Ramerstorfer, F.G., Fisher, F.G., & Scharf, K. 1988. A proposal for the earthquake resistant design of tanks; results from the Austrian research project. Pages 715720 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. VI. Saadeghvaziri, M.A., & Foutch, D.A. 1991. Dynamic behaviour of R/C highway bridges under combined effects of vertical and horizontal earthquake motions. Earthquake Engineering and Structural Dynamics, 20, 535549. Sarma, S.K. 1971. Energy ux of strong earthquakes. Tectonophysics, 11, 159173. Shih, T.-Y., & Lin, Y.K. 1982. Vertical seismic load effect on hysteric columns. Journal of the Engineering Mechanics Division, ASCE, 108(EM2), 242254. Shteinburg, V.V., Chernov, Yu.K., & Ivanova, T.G. 1980. Near eld ground motion. Pages 373378 of: Proceedings of Seventh World Conference on Earthquake Engineering, vol. 2. Sun, C.K., Berg, G.V., & Hanson, R.D. 1973. Gravity effect on single-degree inelastic system. Journal of the Engineering Mechanics Division, ASCE, 99(EM1), 183200. Tani, S., & Soda, S. 1977. Vertical load effect on structural dynamics. Pages 10281033 of: Proceedings of Sixth World Conference on Earthquake Engineering, vol. II. Tocher, D., Patwardhan, A.S., & Cluff, L.S. 1977. Estimation of near eld characteristics of earthquake motion. Pages 470476 of: Proceedings of Sixth World Conference on Earthquake Engineering, vol. I. Tong, H., & Katayama, T. 1988. Peak acceleration attenuation by eliminating the ill-effect of the correlation between magnitude and epicentral distance. Pages 349354 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. II. Veletsos, A.S., & Tang, Y. 1986. Dynamics of vertically excited liquid storage tanks. Journal of Structural Division, ASCE, 112, 12281246. Veletsos, A.S., & Tang, Y. 1988. Soil structure interaction effects for vertically excited tanks. Pages 631636 of: Proceedings of Ninth World Conference on Earthquake Engineering, vol. VI. Wells, D.L., & Coppersmith, K.J. 1994. New empirical relationships among magnitude, rupture length, rupture width, rupture area, and surface displacement. Bulletin of the Seismological Society of America, 84(4), 9741002.
APPENDIX
A. STRONG-MOTION DATA USED BY AMBRASEYS AND SIMPSON
The information in this table is exactly the information used by Ambraseys and Simpson for deriving their results. Many of the distances, some of the magnitudes and moments and depths have been reassessed since completion of Ambraseys and Simpsons study but they are included here for comparison. Soil catagories given are those used by Ambraseys and Simpson. For those records which were only used for the derivation of the attenuation relations for PGA no PGV values are given. The two letter country abbreviations used are: AR for Armenia, CA for Canada, GR for Greece, IR for Iran, IT for Italy, JA for Japan, ME for Mexico, NI for Nicaragua, NZ for New Zealand, TA for Taiwan, TU for Turkey, US for United States of America, UZ for Uzbekistan and YU for Yugoslavia.
Tab. A.1: Near-eld records used by Ambraseys and Simpson (* means record could not be found in new database) Date 1933 1940 1966 3 5 6 11 19 28 Time d 0154 8.0 0436 6.0 0426 6.6 9.3 1401 3.0 7.0 10.0 14.0 15.0 0629 5.0 2000 6.0 0258 3.0 0315 14.0 0400 1.0 10.0 1336 4.0 1535 3.0 11.0 1522 Lat. ah 33.62 2.17 32.73 3.39 35.90 4.93 4.23 34.40 11.60 2.50 1.15 1.49 2.12 12.33 3.65 46.32 3.60 40.28 6.80 46.29 4.88 3.92 1.37 31.90 10.21 33.36 8.84 3.90 Long. av -117.97 2.07 -115.45 2.50 -120.90 2.72 1.27 -118.40 6.68 1.74 1.04 1.15 1.22 -86.13 3.14 13.32 2.37 63.39 12.92 13.20 1.81 1.47 1.18 50.76 5.01 57.42 7.30 1.84 Depth vh 10 25.7 7 36.5 7 77.6 25.0 8 111.2 29.9 32.1 13.8 22.5 6 33.9 6 30.9 13 66.6 15 27.7 6 Ms vv 6.41 19.4 7.15 9.8 6.13 14.5 7.6 6.61 54.0 31.1 18.2 6.0 8.1 6.16 18.1 6.50 9.6 7.10 66.4 6.06 9.5 6.38 log M0 S 25.79 S 26.59 S 25.36 S S 26.11 R S S A 25.34 S 25.53 R 26.26 L 24.80 A 25.67 F Station S Long Beach-City Hall S El Centro Array-Station 9 S Cholame Shandon Array 2W Cholame Shandon Array 5W T Pacoima Dam 8244 Orion Boulevard Vanowen Street 15910 Ventura Boulevard Jet Propulsion S Esso T Tolmezzo-Diga Ambiesta T Gazli S Breginj-Fab. IGL S C US US US
1971
US
1972 1976 1976 1976 1976
12 5 5 9 11
23 6 17 15 71
NI IT UZ IT
1977 1978
4 9
6 16
1978
1 2 3
11
43
10 6.02 25.11 61.7 A 5 7.33 27.11 106.4 34.7 S 25.9 11.5 A 20 6.16 25.58 continued on next page
T Naghan 12 T Tabas Dayhook T
IR IR
Earthquake not in ESEE database Only have one horizontal time history Earthquake not in ESEE database
A. Strong-motion data used by Ambraseys and Simpson Tab. A.1: continued Date Time d 4.0 0619 9.0 9.0 12.0 12.0 1723 7.0 9.0 12.0 2316 0.5 0.6 1.3 1.4 2.6 3.8 4.0 5.1 6.0 6.8 7.5 8.0 8.0 8.4 8.5 8.5 11.0 13.0 13.0 13.0 14.0 1633 1.0 3.0 1944 1.0 0328 2.0 10.0 1834 8.0 15.0 2053 4.0 13.0 1209 5.0 6.0 1722 15.0 2342 Lat. ah 3.53 41.98 2.26 2.73 3.67 4.55 42.23 2.91 2.66 2.62 32.64 3.01 4.56 4.27 3.04 7.53 6.03 4.93 4.77 3.92 4.83 2.49 3.43 2.33 4.26 2.19 2.24 2.67 3.75 2.62 2.97 2.01 37.61 4.41 3.03 37.56 2.26 32.19 8.45 5.90 40.78 1.83 1.69 38.10 2.93 2.98 33.13 4.48 1.85 1.96 36.24 Long. av 1.57 18.98 2.19 4.25 2.36 2.05 18.76 1.21 1.79 0.98 -115.31 2.48 6.16 15.35 1.42 3.59 4.57 5.47 5.97 3.73 2.44 2.31 1.86 2.27 3.82 1.50 1.00 1.95 1.31 1.27 1.23 1.59 -118.85 3.74 2.22 -118.84 1.70 -115.08 14.90 2.64 15.33 0.96 1.56 22.84 1.28 1.14 -115.65 6.12 2.02 Depth vh 12 28.3 52.0 52.2 39.8 5 16.7 28.1 16.6 8 91.6 109.5 110.0 42.9 45.0 53.9 88.4 70.0 77.4 50.7 66.5 31.7 37.5 46.9 19.6 41.9 47.8 36.7 16.8 8 23.3 15.9 1 19.2 5 76.3 31.6 10 33.1 29.9 10 23.7 25.8 6 43.7 13.2 11 Ms vv 7.04 13.5 16.9 14.3 15.4 6.34 9.7 7.3 7.9 6.86 29.2 24.5 61.4 5.5 14.1 21.4 37.1 20.3 15.3 9.9 17.1 9.4 8.5 8.5 6.4 11.3 8.7 3.0 6.7 6.16 22.1 11.1 6.01 8.2 6.60 89.3 11.7 6.87 14.2 24.3 6.69 7.2 8.6 6.05 16.7 6.0 7.11 log M0 S 26.49 R A A A 25.34 A A A 25.82 S S S S S S S S S S S S S S S S S S L S A 25.32 F Station T Ulcinj-Albatros Hotel Ulcinj-Olimpic Hotel Bar-Town Hall Petrovac-Oliva Hotel T Petrovac-Riviera Hotel Budva-PTT Bar-Town Hall S Meloland Overpass-Free Field El Centro Array-Station 7 El Centro Array-Station 6 Mexicali-Aeropuerto Bonds Corner-Ground El Centro Array-Station 8 El Centro Array-Station 5 El Centro Differential Array El Centro El Centro Array-Station 4 Holtville-Post Ofce ICSB Imperial County Centre (Lot) Mexicali-Casa Flores Brawley-Airport Hangar El Centro Array-Station 10 Calexico-Fire Station El Centro Array-Station 11 El Centro Array-Station 3 Cucapah El Centro-Parachute Test Site N Convict High School-Gym (Centre) N Convict S Victoria Cerro Prieto N Bagnoli-Irpino Calitri-Cabina Pittoli N Xilokastro-OTE Building Korinthos-OTE Building S Westmorland-Fire Station Salton Sea-Wildlife Refuge T Golbasi T C
53
1979
15
YU
1979
24
YU
1979
10
15
US
1980
25
US
1980 1980
5 6
25 9
24.92
US ME
1980
11
23
1981
24
1981
26
1981 1983
4
7 5
284 2
25.70 S A 26.40 R A 25.95 S S 25.20 S L 26.90 2.75 R -120.27 7 6.57 25.66 continued on next page
IT
GR
US
US
Earthquake not in ESEE database
A. Strong-motion data used by Ambraseys and Simpson Tab. A.1: continued Date Time d 7.0 7.0 2115 3.0 3.0 4.0 11.0 13.0 14.0 14.0 15.0 0920 7.0 8.0 9.0 9.0 1442 14.0 0142 9.0 1442 3.0 4.0 7.0 8.0 9.0 10.0 10.0 11.0 11.0 13.0 13.0 15.0 0154 15.0 0004 0.1 1.0 5.0 7.0 7.0 9.0 9.0 10.0 10.0 11.0 11.0 12.0 12.0 12.0 12.0 13.0 13.0 Lat. ah 5.86 2.94 37.31 12.70 3.05 4.16 2.87 3.53 1.90 1.97 2.03 34.00 6.06 3.22 6.51 2.14 37.54 4.51 -37.93 3.14 34.07 4.55 3.02 3.70 1.71 3.82 4.21 6.11 3.24 4.47 2.92 2.72 2.04 1.77 37.04 6.35 4.90 4.30 2.49 2.71 4.99 6.47 1.88 1.52 4.41 3.68 4.99 2.28 3.66 2.80 4.50 2.62 Long. av 3.48 2.05 -121.68 3.58 1.08 1.66 4.15 3.96 4.46 3.82 5.62 -116.61 4.30 4.36 6.84 2.96 -118.45 3.43 176.78 1.38 -118.08 3.20 3.81 1.76 1.19 1.37 1.40 2.24 1.27 4.38 1.71 1.50 0.63 1.37 -121.88 4.28 4.90 1.42 6.38 4.84 5.32 2.65 1.37 1.87 1.96 1.91 3.46 2.65 2.75 1.43 3.61 2.02 Depth vh 62.8 37.0 8 79.5 39.2 27.5 36.7 19.2 12.4 12.6 12 37.7 33.8 65.9 39.9 4 8 29.0 14 20.4 15.5 20.6 12.6 21.8 28.6 41.7 11.7 Ms vv 15.9 14.7 6.17 15.2 12.2 10.0 15.1 11.2 8.8 9.3 6.16 16.6 19.9 13.1 9.2 6.22 6.49 9.6 6.00 9.1 7.8 6.3 6.2 5.7 7.6 7.9 3.4 S 11 17 55.9 92.9 33.7 33.2 37.0 6.33 7.17 23.3 23.1 22.5 15.6 17.2 25.40 S 26.66 A R log M0 S F Station Pleasant Valley-Pumping Yard Pleasant Valley-Pumping (Base.) S Coyote Lake Dam-Abutment Halls Valley-Grant Park Anderson Dam-Downstream Gilroy Array-Station 6 Gilroy Array-Station 4 Gilroy Array-Station 7 Gilroy Array-Station 3 Gilroy Array-Station 2 T White Water Canyon-Trout Farm Desert Hot Springs North Palm Springs-Post Ofce Morongo Valley-Fire Station S Zack N Matahina Dam-Bottom Centre T Garvey Reservoir-Abutment Whittier Narrows Dam-Base. Up. Alhambra-Fremont School San Marino-S.W. Academy CSULA Obregon Park 7215 Bright Avenue (Base.) 4407 Jasp Bell-Bulk Mail Facility Altadena-Eaton Canyon Vernon CMD (*) Norwalk.B. (*) S Westmo Fir T Corralitos-Eureka Canyon Rd. UCSC Branc Lexington Dam-Left Abut. Watsonville-Station 4 Watsonville-Station 3 Capitola-Fire Station UCSC Walte SJ 3ST/1 SJ 3ST/2 Gilroy Array-Station 1 Gilroy-Gavilan College Saratoga-Aloha Avenue Sara E.Wal Gilroy Array-Station 2 Gilroy-2 Storey Hist. Com. Build. USCS Santa Cruz-Lick Obs. SJTeresa C
54
1984
24
1986
1986 1987 1987
7 3 10
21 2 1
25.41 A A A A S S S S 25.20 R A A A 25.45 25.81 24.90
US
US
US NZ US
1987 1989
11 10
24 18
US US
R A A A 39.0 15.3 S 42.8 11.6 A 20.4 11.7 A A continued on next page
33.8 29.4 41.3
14.8 13.0 25.7
A. Strong-motion data used by Ambraseys and Simpson Tab. A.1: continued Date Time d 14.0 2100 8.0 1718 4.0 1806 0.10 0.10 12.0 1257 2.0 11.0 Lat. ah 5.25 36.96 6.37 39.72 4.90 40.37 17.65 6.77 5.39 34.19 8.63 2.84 Long. av 2.84 49.41 5.10 39.63 2.39 -124.32 18.14 1.77 1.96 -116.53 3.92 1.86 Depth vh 44.0 19 10 109.5 13 Ms vv 13.7 7.32 6.85 20.9 7.15 log M0 S S 27.02 26.40 A 26.78 A S S 27.16 R S F Station Gilroy Array-Station 3 T S Erzincan-Meteorological Station T C.Mendocin Petrolia Rio Del.FF. S U. John Vly Joshua Tre C
55
1990 1992 1992
6 3 4
20 13 25
IR TU US
1992
28
11
7.64
US
B. STRONG-MOTION DATA USED FOR NEW EQUATIONS
Note that the PGA and PGV values in Table B.1 are different than those given in Table A.1 due to the different processing technique used. See Appendix A for two-letter country abbreviations used.
Tab. B.1: Near-eld records used for new equations Date 1940 1954 1966 5 12 6 19 21 28 Time d 0436 6 1956 1 0426 0 6 10 11 15 1401 15 0629 4 0258 3 0315 9 12 12 14 0921 6 7 8 9 9 12 1336 4 2333 11 1535 3 11 Lat. ah 32.73 3.39 40.82 2.67 35.90 4.85 4.19 2.63 3.64 0.64 34.40 3.50 12.33 3.60 40.28 6.78 46.29 1.06 1.19 2.59 4.87 46.32 1.32 4.15 0.91 3.45 2.33 0.87 31.90 10.28 38.27 1.59 33.36 9.34 3.92 Long. av -115.45 2.49 -124.08 0.43 -120.90 2.72 1.28 1.23 1.24 0.59 -118.40 1.62 -86.13 3.18 63.39 13.17 13.20 0.63 0.57 0.92 1.81 13.16 0.58 1.26 0.84 1.82 0.82 0.45 50.76 14.86 0.47 57.42 7.27 1.86 Depth Ms log M0 vh vv S 7 7.14 26.49 32.0 8.6 S 15 6.67 25.48 28.58 8.3 S 7 6.13 25.36 73.5 14.1 S 23.7 7.0 S 11.8 4.8 S 20.4 4.2 A 5.5 4.6 A 8 6.61 26.00 16.6 3.8 A 5 6.16 26.523 29.8 15.2 S 13 7.04 26.26 65.1 66.1 L 15 6.06 24.80 11.4 7.1 S 7.6 5.6 A 10.1 6.7 A 27.7 9.8 A 12 5.98 25.11 6.9 3.3 S 11.7 4.9 A 8.0 4.7 S 23.2 10.5 A 18.7 13.9 A 2.3 1.5 R 10 6.02 25.11 57.8 R 15 5.83 25.08 15.5 2.4 S 5 7.33 27.11 84.5 36.9 A 24.8 10.6 R continued on next page F Station S El Centro Array-Station 91 S Eureka-Myrtle & West S Cholame Shandon Array 2W2 Cholame Shandon Array 5W Cholame Shandon Array 8W Parkeld-Temblor Cholame Shandon Array 12W T Lake Hughes Array-Station 12 S Esso T Gazli S Buia San Rocco Forgaria-Cornio Breginj-Fabrika IGLI T Tarcento Breginj-Fabrika IGLI Buia Forgaria-Cornio San Rocco Robic T Naghan 14 T Patti-Cabina Prima T Tabas Dayhook C US US US
1971 1972 1976 1976
2 12 5 9
9 23 17 15
US NI UZ IT
1976
15
IT
1977 1978 1978
4 4 9
6 15 16
IR IT IR
The instrument was in a building with massive foundations which could have affected high frequencies (Brune & Anooshehpoor, 1991) 2 Only one horizontal and one vertical component 3 This seismic moment estimate comes from Wells & Coppersmith (1994) using an estimate of the rupture area and the slip which occurred along the fault. It is an overestimate 4 Only one horizontal component and no vertical component
B. Strong-motion data used for new equations Tab. B.1: continued Date 1979 4 15 Time d 0619 9 9 12 12 1723 7 9 12 15 2135 6 2316 0 0 1 1 2 2 3 3 5 6 6 7 9 9 10 11 11 11 12 13 15 15 0658 5 1633 1 12 12 1944 5 0328 2 10 1834 6 10 14 Lat. ah 41.98 2.75 2.22 4.52 3.66 42.23 2.91 2.67 2.62 1.69 42.76 1.87 32.64 3.03 4.84 4.19 2.90 3.33 7.45 4.86 4.41 2.32 2.19 5.92 4.71 2.61 4.26 2.37 2.65 4.09 2.18 2.94 1.06 3.88 1.97 33.02 1.53 37.61 4.42 3.02 2.44 37.56 2.25 32.19 8.48 5.83 40.78 1.80 0.58 2.99 Long. av 18.98 4.25 2.19 2.03 2.35 18.76 1.21 1.78 0.98 0.94 13.02 1.76 -115.31 1.42 5.39 15.46 2.45 8.84 3.58 2.46 6.11 2.32 1.52 4.47 6.06 1.26 3.81 2.31 1.94 1.15 0.95 1.24 0.84 1.34 1.59 -115.58 1.03 -118.85 3.73 2.22 -118.84 1.70 -115.08 14.66 2.64 15.33 0.97 0.35 1.67 Depth Ms log M0 vh vv S 12 7.04 26.49 48.1 16.2 A 26.1 11.9 R 39.0 13.3 A 51.7 15.1 A 5 6.34 25.34 16.8 9.0 A 27.3 7.1 A 16.4 7.5 A 8.5 5.3 A 4 5.84 24.84 14.0 5.6 R 10 6.87 25.70 42.0 5.1 S 71.8 30.0 S 88.1 49.7 S 84.4 28.4 S 38.9 13.6 A 40.8 13.9 S 66.7 11.7 S 90.6 22.6 S 38.4 7.9 S 31.3 7.0 S 48.0 19.6 S 54.7 17.8 S 42.5 8.1 L 28.9 8.2 S 52.8 13.0 S 14.7 4.7 S 28.6 6.3 S 39.5 8.8 S 34.7 3.1 S 17.9 6.0 S 33.1 10.8 S 15.8 4.7 A 5 5.86 24.51 10.8 2.3 S 9 6.15 25.26 23.8 19.8 S 15.7 10.8 A 15.8 A 16 5.99 24.92 18.0 7.9 S 5 6.49 25.79 65.2 62.2 S 31.7 11.1 A 16 6.87 26.40 30.8 14.4 R 5.9 3.5 A 57.2 20.1 R continued on next page F Station T Ulcinj-Hotel Olimpic Ulcinj-Hotel Albatros Petrovac-Hotel Oliva Bar-Skupstina Opstine T Petrovac-Hotel Rivijera Budva-PTT Bar-Skupstina Opstine Tivat-Aerodrom N Cascia-Cabina Petrucci S Mexicali-Aeropuerto El Centro Array-Station 5 El Centro Array-Station 6 Meloland Overpass-Free Field Agrarias (Mexicali Valley)5 Bonds Corner-Maintenance Shop El Centro Array-Station 4 El Centro Array-Station 7 Holtville-Post Ofce Brawley-Airport Hangar El Centro Array-Station 8 El Centro Differential Array El Centro Array-Station 3 Mexicali-Casa Flores Imperial County Centre (Lot) Calexico-Fire Station El Centro Array-Station 2 El Centro Array-Station 10 Cucapah (Mexicali Valley)6 Westmorland-Fire Station El Centro Array-Station 11 El Centro-Parachute Test Site S Westmorland-Fire Station N Convict High School-Gym (Centre) High School-Gym (West Wall)7 N Convict S Victoria (Mexicali Valley) Cerro Prieto (Mexicali Valley) N Bagnoli-Irpino Auletta Sturno C
57
YU
1979
24
YU
1979 1979
9 10
19 15
IT US
1979 1980
10 5
16 25
US US
1980 1980
5 6
25 9
US ME
1980
11
23
IT
5 6 7
Vertical components recording may be erroneous due to a loose mounting bolt (Brune et al., 1982) Only one horizontal and one vertical component Only one horizontal and no vertical component
B. Strong-motion data used for new equations Tab. B.1: continued Date Time d 14 2053 4 13 1209 6 7 15 2342 7 7 2115 0 10 13 14 15 15 15 15 15 0516 0 4 0920 0 1 5 6 7 11 15 1442 2 2 4 4 4 5 6 6 6 6 8 8 9 9 10 10 10 13 14 Lat. ah 1.70 38.10 2.88 2.97 33.13 1.85 4.51 1.73 36.24 5.77 2.92 37.31 3.05 1.68 2.87 3.53 0.89 2.03 1.97 1.15 1.50 62.19 10.41 3.48 34.00 6.05 8.91 6.48 3.22 2.15 2.04 1.96 34.07 2.43 2.63 2.86 3.70 4.21 4.63 2.17 1.30 1.71 4.48 3.39 2.32 1.65 2.91 1.34 1.81 2.02 2.23 0.78 Long. av 1.58 22.84 1.28 1.13 -115.65 2.02 6.12 1.12 -120.27 3.51 2.07 -121.68 1.08 0.77 4.15 3.96 0.91 5.61 3.81 1.08 0.59 -124.24 19.41 -116.61 4.31 4.09 6.82 4.37 2.95 2.99 1.80 -118.08 1.65 2.61 2.23 1.76 1.40 2.40 1.17 0.90 1.19 4.38 2.55 1.58 1.01 1.33 0.85 1.70 1.49 1.19 0.68 Depth Ms log M0 vh vv S 27.8 17.1 A 10 6.69 25.95 23.1 8.2 S 22.5 6.8 S 8 6.04 24.72 13.7 5.6 L 43.7 14.7 S 7.3 3.3 S 7 6.57 25.66 62.3 17.9 S 39.7 12.5 S 9 6.20 25.84 39.6 11.8 S 11.2 6.4 R 36.7 15.1 A 18.5 11.5 S 2.8 2.9 R 12.0 9.3 S 11.3 8.6 S 3.4 3.0 A 12.6 3.2 6 6.79 26.28 45.7 37.1 R 30.9 R 12 6.13 24.84 37.7 15.7 R 88.8 18.1 S 61.9 12.5 A 32.0 19.1 A 36.8 8.5 A 14.0 7.2 S 12.1 6.4 S 14 5.94 24.92 16.8 4.7 S 23.5 6.0 A 18.3 6.1 A 21.2 6.3 A 21.6 5.3 A 31.8 7.3 S 26.2 2.7 S 7.4 2.7 A 13.1 6.2 A 38.4 7.1 S 37.9 4.6 S 10.2 5.9 S 10.6 2.5 S 11.1 3.2 S 10.0 2.9 S 12.9 5.5 A 30.1 3.1 S 19.9 6.2 S 7.0 3.4 S continued on next page F Station Calitri-Cabina Pittoli N Xilokastro-OTE Building Korinthos-OTE Building S Salton Sea-Wildlife Refuge Westmorland-Fire Station Niland-Fire Station T Pleasant Valley-Pumping Yard Pleasant Valley-Pumping (Base.) S Halls Valley-Grant Park San Jose-IBM Building 12 Gilroy Array-Station 6 Gilroy Array-Station 4 Gilroy Array-Station 1 Gilroy Array-Station 2 Gilroy Array-Station 3 Gilroy-Gavilan College IBM Almaden T Nahanni-Station 1 Nahanni-Station 2 8 S White Water Canyon-Trout Farm Devers Substation North Palm Springs-Post Ofce Desert Hot Springs-Fire Station Morongo Valley-Fire Station Cabazon-Post Ofce Palm Springs-Airport T El Monte-11338 Fairview Ave. San Gabriel-600 E. Grand Ave. Arcadia-180 Campus Drive Alhambra-Fremont School Obregon Park 11500 E. Joslin St. Bell Gardens-7420 Jaboneria Baldwin-3699 N. Hollywood Ave. San Marino-S.W. Academy Bell-Bulk Mail Facility Downey-12500 Birchdale Pasadena-1488 Old House Rd 2369 E. Vernon Ave. Altadena-Eaton Canyon Covina-1271 W. Badillo 5921 N. Figueroa St. Downey-Co. Maintenance Build. 3036 Fletcher Drive Covina-656 S. Grand Ave. C
58
1981
24
GR
1981
26
US
1983
US
1984
24
US
1985
12
23
CA
1986
US
1987
10
US
No vertical component
B. Strong-motion data used for new equations Tab. B.1: continued Date Time d 14 14 14 14 15 15 0745 10 0004 1 7 9 12 12 14 15 15 15 1718 1 0450 15 1806 0 9 14 1157 3 11 1505 13 1230 0 0 0 0 0 0 0 0 0 0 0 1 5 5 6 7 8 9 9 10 Lat. ah 3.12 2.20 1.74 1.59 1.15 1.73 40.96 1.45 37.04 6.37 4.99 5.01 4.43 3.68 3.65 4.48 5.24 2.81 39.72 5.03 33.96 2.00 40.37 14.68 2.07 4.67 34.19 7.74 2.78 34.27 5.34 34.16 3.80 4.02 8.90 3.17 6.21 9.79 5.78 8.27 4.44 8.16 9.25 17.45 4.59 3.37 3.23 4.40 4.24 2.06 3.02 4.62 Long. av 1.69 1.56 0.69 1.11 0.83 0.79 44.27 0.39 -121.88 4.32 3.55 5.31 2.00 1.89 2.79 3.65 2.89 1.46 39.63 2.44 -116.32 1.83 -124.32 7.39 0.83 1.11 -116.53 8.07 1.78 -116.78 1.90 -118.57 4.98 2.72 3.84 2.52 3.85 7.30 5.37 5.25 7.85 8.19 4.52 10.28 2.89 5.41 1.77 2.92 1.77 1.58 2.70 1.65 Depth Ms log M0 vh vv S 29.1 3.7 S 30.9 3.0 S 12.9 3.3 S 4.5 2.5 R 5.4 2.8 A 17.7 3.0 S 11 5.8 11.0 3.0 S 18 7.17 26.66 55.2 16.8 A 39.0 19.8 A 38.9 17.8 S 35.4 10.1 R 26.0 8.1 A 40.0 10.9 S 18.8 10.8 A 42.7 12.4 S 43.2 7.7 A 10 6.75 26.40 102.4 15.7 A 12 6.20 25.26 12.9 6.2 S 15 7.10 26.92 126.1 60.3 R 43.5 10.9 R 52.1 16.5 A 1 7.55 27.04 146.5 41.1 R 42.7 14.8 A 15 6.52 25.83 34.4 11.3 A 14 6.81 26.08 60.7 13.5 S 96.3 30.4 A 52.3 14.0 A 12.8 11.4 R 92.1 30.9 A 66.9 27.2 A 94.7 30.7 S 128.9 18.6 S 61.2 35.4 S 147.5 35.6 S 75.8 22.0 S 110.2 72.3 S 46.2 18.3 S 40.4 17.5 S 15.7 5.1 A 41.9 12.3 S 50.9 14.2 A 30.4 7.0 A 28.6 11.4 A 32.6 13.5 S continued on next page F Station Compton-14637 Castlegate St. Long Beach-6979 Orange Ave. La Habra-950 Briarcliff Drive Mt. Wilson-Caltech Seism. Stat. Glendora-120 North Oakbank 924 W. 70th St. T Gukasian T Corralitos-Eureka Canyon Rd. Saratoga-Aloha Avenue Capitola-Fire Station Gilroy Array-Station 1 Gilroy-Gavilan College Gilroy Array-Station 2 USCS Santa Cruz-Lick Obs. Gilroy Array-Station 3 Gilroy-2 Storey Hist. Com. Build. S Erzincan-Meteorologij Mudurlugu S Thousand Palms-Post Ofce T Cape Mendocino (C) Eel River Valley-Bunker Hill Eel River Valley-Centreville Beach S Lucerne Valley (C) Joshua Tree-Fire Station(C) S Big Bear Lake-Civic Center(C) T 7769 Topanga Canyon Boulevard Newhall-26835 W. Pico Canyon Rd Simi Valley-6334 Katherine Rd Santa Susana-6633 Canoga Ave.9 Jensen Filtration-Admin. Build. Jensen Filtration-Filter Gen. Room Newhall-Fire Station (C) Sylmar-Co. Hosp. (Parking Lot) (C) Northridge-17645 Saticoy Street (C) Rinaldi Receiving Station Sepulveda VA Hospital Tarzana-Cedar Hill Nursey A (C) 16628 W. Lost Canyon Rd Arleta-Nordhoff Ave. Fire Stat. (C) Topanga Fire Station Sun Valley-13248 Roscoe Boulevard Pacoima Dam-Kagel Canyon(C) 12001 Chalon Road 6850 Coldwater Canyon Ave. 17291 Sunset Boulevard C
59
1988 1989
12 10
7 18
AR US
1992 1992 1992
3 4 4
13 23 25
TU US US
1992
28
US
1992 1994
6 1
28 17
US US
Only one horizontal and one vertical component
B. Strong-motion data used for new equations Tab. B.1: continued Date Time d 10 10 12 12 13 13 13 13 14 14 15 15 15 15 2046 0 1 1 3 6 11 12 12 0015 0 11 1557 1 0940 3 4 0001 3 14 1747 0 3 1657 9 11 Lat. ah 4.74 3.70 1.81 1.78 6.15 2.77 3.67 2.51 2.53 1.51 4.36 8.66 5.57 2.51 34.55 8.21 3.01 3.42 7.62 5.05 4.95 6.92 2.87 38.37 5.17 0.47 38.06 3.18 43.01 2.56 5.35 40.70 3.46 2.18 23.86 4.16 9.73 40.77 5.00 Long. av 3.09 1.40 1.21 1.37 2.91 1.89 1.30 1.60 1.15 1.90 1.50 2.28 2.13 1.13 135.04 3.34 4.23 5.58 3.62 4.49 2.93 3.40 1.23 22.15 1.92 0.29 30.15 1.33 12.84 1.81 3.62 29.99 2.41 1.32 120.81 2.98 5.91 31.15 3.33 Depth vh 65.6 37.1 8.3 23.6 30.2 26.1 31.9 33.0 11.7 16.6 63.9 41.8 52.6 25.1 19 90.4 55.4 91.1 54.3 91.7 60.9 174.1 24.2 3 49.7 3.5 5 41.1 8 17.6 32.5 17 34.6 27.5 11 36.0 62.2 14 66.5 Ms vv 19.5 5.9 3.6 9.3 10.2 6.8 10.8 8.8 4.5 12.4 12.3 14.0 12.2 6.2 6.95 40.1 19.7 62.0 35.5 32.6 24.6 32.6 5.9 6.35 14.2 2.3 6.07 14.9 5.9 7.7 28.5 7.8 32.7 11.2 7.6 28.6 36.7 7.3 13.0 log M0 S S A A A A A A A A A S A S 26.80 R R S R L F Station 14145 Mulholland Drive Stone Canyon Reservoir Monte Nido Fire Station Brentwood V.A. Hospital 12520 Mulholland Drive 700 North Faring Road Wadsworth V.A. Hosp. (GS South) Wadsworth V.A. Hosp. (GS North) Lake Hughes Array-Station 12 (C) Sunland-10965 Mt Gleason Ave. 1955 Purdue Ave. (Base.) Santa Monica-City Hall (C) Castaic-Ridge Route (C) Century City-LACC North (C) S Kobe (C) Kobe-University (C) Kobe-Port Island Array (Surface) (C) Kobe-Motoyama (C) Kobe-Port Amagasaki-G Kobe-8G Tadaoka (C) N Aigio-OTE Building Nafpaktos-OTE Building N Dinar-Meteoroloji Mudurlugu N Colorito Nocera Umbra S Sakarya10 Izmit-Meteoroloji Istasyonu T CWB station code: T078 CWB station code: T129 S Duzce-Meteoroloji Mudurlugu C
60
1995
16
JA
1995
15
1995 1997
10 9
1 26
1999
17
1999
20
25.78 A A 25.73 S 25.06 A R 27.15 A R 27.53
GR
TU IT
TU
TA
1999
11
12
26.65
TU
10 11
Only one horizontal and one vertical component This is the epicentral distance since fault location unknown at present
ESEE RESEARCH PUBLICATIONS The Engineering Seismology and Earthquake Engineering Section at Imperial College publishes reports on research work undertaken by members of the Section and the Department; a list of currently available reports is given below. Copies may be obtained from the ESEE Section Secretary, Department of Civil and Environmental Engineering, Imperial College, London SW7 2BU (Prices on application, Reports marked N are not for sale and some are out-of-print. The reports are sent free to selected libraries in Universities and Research Centres). Shake Table Facility Time History Matching and Exploratory Analysis, A.C. Hargreaves, A.H. Al- Sheikh and S.H. Perry, ESEE-86/1, April 1986. Regional Seismicity Studies, Eastern Mediterranean and West Africa, N.N. Ambraseys, C. Melville, R. Adams and C. Finkel, ESEE-86/2, May 1986. Regional Seismicity Studies, Northwest Europe and U.K., N.N. Ambraseys and J. Jackson, ESEE86/3, May 1986. UNESCO Earthquake Reconnaissance Missions, 1963-1981, N.N. Ambraseys, ESEE-86/4, June 1986. Shake Table Facility Strong Motion Database, A.C. Hargreaves, ESEE-86/5, July 1986. Composite Steel/Concrete Connections: Analytical Studies and a Design Equation, A.S. Elnashai and P.J. Dowling, ESEE-86/6, August 1986. Review of International Shake Table Facilities, A.C. Hargreaves and N.N. Ambraseys, ESEE86/7, August 1986. Corpus of Isoseismal Maps of the Eastern Mediterranean Region, N.N. Ambraseys and C. Finkel, ESEE-86/8, November 1986. The Kalamata (Greece) Earthquake of 13 September 1986, A.S. Elnashai and K. Pilakoutas, ESEE-86/9, December 1986. IC Shake Table Facility Quality Assurance Plan for Seismic Qualication, B. Hillary, A.C. Hargreaves and J.G. Wardrop, ESEE-87/1, June 1987. The State-of-the-Art in Inlled Frames, H.A. Moghaddam and P.J. Dowling, ESEE-87/2, August 1987. Karameh Dam Dynamic Behaviour of Foundation Materials, A.C. Hargreaves and N.N. Ambraseys, ESEE-87/3, August 1987. An Energy Absorption Device for Steel Braced Frames, G. Anagnostides, T.A. Wyatt and A.C. Hargreaves, ESEE-88/1, February 1988. ESEE-88/2 - missing. Dynamic Material Behaviour of Earth and Rockll dam, S.K. Sarma and M. Haeri, ESEE-88/3, February 1988.
Large Displacement Elasto-plastic Analysis of Steel Frames, A.S. Elnashai, B.A. Izzuddin and P.J. Dowling, ESEE-88/4, May 1988 (jointly with The Steel Construction Institute). (N) Equipment Qualication for Sizewell B and other Nuclear Facilities, A.C. Hargreaves, ESEE88/5, March 1988. (N) ESEE-88/6 - missing. Imperial College Shake Table Software Users Guide, A.C. Hargreaves, ESEE-88/7, November 1988. (N) Repair and Strengthening of Reinforced Concrete Structures Damaged by Earthquakes, P. Pitsillides, ESEE-88/8, November 1988. The Spitak (Armenia, USSR) Earthquake of 7 December 1988; A Preliminary Engineering Seismology Report, J.J. Bommer and N.N. Ambraseys, ESEE-89/1, March 1989. Transient Fluid-Structure Interaction Procedures, A.E.K. Chelghoum, A.S. Elnashai and P.J. Dowling, ESEE-89/2, March 1989. Earthquake Strong-motion Database Users Guide, J.J. Bommer, ESEE-89/3, May 1989. (N) Imperial College Shaketable Software Users Guide: Supplement No. 1, A.C. Hargreaves, ESEE-89/4, July 1989. Nonlinear Transient Dynamic Analysis of Steel Frames, B.A. Izzuddin, A.S. Elnashai and P.J. Dowling, ESEE-89/5, July 1989. (N) The Killini (Elias, Greece) Earthquake of October and November 1988, MSC/EFTU Group, ESEE-89/6, September 1989. ADAPTIC, A Program for Adaptive Large Displacement Nonlinear Dynamic Analysis of Space Frames, Users Manual, B.A. Izzuddin and A.S. Elnashai, ESEE-89/7, September 1989. Effect of Random Material Variability on the Structural Response of Steel Frames, P. Alexopoulou, A.S. Elnashai and M. Chryssanthopoulos, ESEE-89/8, September 1989. Transient Dynamic 3D Analysis of a Damaged Two Storey Reinforced Concrete Building, P. Makris and A.S. Elnashai, ESEE-89/9, September 1989. (N) A Theoretical Model for Composite Beam-Columns under Cyclic Loading, P. Madas and A.S. Elnashai, ESEE-89/10, September 1989. The Loma Prieta (Santa Cruz, California) Earthquake of 17 October 1989, A.S. Elnashai, J.J. Bommer and A.Y. Elghazouli, ESEE-89/11, December 1989. Materiaux Relatifs a la Sismicite de lAlgerie Occidentale au Cours de la Deuxieme Moitie du XIXe et du Premier Tiers du Xxe Siede, J. Vogt and N.N. Ambraseys, ESEE-90/1, January 1990. Time History Matching, A.C. Hargreaves, ESEE-90/2, May 1990. Friction Load Control Devices for Steel Braced Frames, G. Anagnostides, A.C. Hargreaves and
T.A. Wyatt, ESEE-90/3, June 1990. The Chenoua (Algeria) Earthquake of 29 October 1989, N.N. Ambraseys, A.S. Elnashai, J.J. Bommer, F. Haddar, P. Madas, A.Y. Elghazouli and J. Vogt, ESEE-90/4, July 1990. Mixed Eulerian-Lagrangian Finite Element Solution for Transient Fluid-Structure Interaction Problems, A.E.K. Chelghoum, A.S. Elnashai and P.J. Dowling, ESEE-91/1, February 1991. (N) A New Passive Connement Model for Concrete Subjected to Variable Amplitude Cyclic Loading, P. Madas and A.S. Elnashai, ESEE-91/2, March 1991. An Analytical Approach to Seismic Energy Absorption of Steel Frames with Random Material Characteristics, K. Koh, A.S. Elnashai and M. Chryssanthopoulos, ESEE-91/3, April 1991. Earthquake Resistant Design of RC Walls, K. Pilakoutas, supervised by A.S. Elnashai and N.N. Ambraseys, ESEE-91/4, April 1991. Database of European Strong Ground Motion, N.N. Ambraseys and J.J. Bommer, ESEE-91/5, April 1991. Attenuation of Peak Ground Acceleration for Europe, N.N. Ambraseys and J.J. Bommer, ESEE91/6, May 1991. Seismicity of Algeria; Cheliff Valley Earthquake, N.N. Ambraseys and D. Benouar, ESEE-91/7, June 1991. Long-term Seismicity of Istanbul, N.N. Ambraseys and C. Finkel, ESEE-91/8, July 1991. ESEE-92/1 - missing. Selective Repair and Retrotting Techniques for RC Structures in Seismic Regions, A.S. Elnashai and A.I. Salama, ESEE-92/2, April 1992. Performance of Steel/Concrete Composite Members under Earthquake Loading, A.Y. Elghazouli and A.S. Elnashai, ESEE-92/3, June 1992. Effect of Building Conguration on Seismic Response Parameters, M.M. Soliman, ESEE-92/4, July 1992. A Component-Based Model for Cyclic Analysis of Steel and Composite Beam-Column Connections, P.J. Madas and A.S. Elnashai, ESEE-92/5, July 1992. Testing and Analysis of Partially-Encased Beam-Columns under Combined Earthquake and Axial Loading, B.M. Broderick and A.S. Elnashai, ESEE-92/6, August 1992. Statistical Analysis of Steel Tensile Test Data and Implications on Seismic Design Criteria, G.M.E. Manzocchi, M. Chryssanthopoulos and A.S. Elnashai, ESEE-92/7, September 1992. Review of UK Seismic Design Parameters, N.N. Ambraseys, S.K. Sarma, J.J. Bommer and M. Free, ESEE- 92/8, November 1992. The Seismicity of Cyprus, N.N. Ambraseys and R. Adams, ESEE-92/9, November 1992.
Experimental Behaviour of Steel and Composite Frames under Cyclic and Dynamic Loading, K. Takanashi, A.S. Elnashai, A.Y. Elghazouli and K. Ohi, ESEE-92/10, November 1992. The Erzincan (Turkey) Earthquake of 13 March 1992, N.N. Ambraseys, A.S. Elnashai, B.M. Broderick, A.I. Salama and M.M. Soliman, ESEE-92/11, December 1992. Review of Repair Methods for Reinforced Concrete Structures, A.I. Salama and A.S. Elnashai, ESEE-93/1, July 1993. Re-Evaluation of Strong-Motion Data from Turkey, N.N. Ambraseys, E. Durukal and M. Free, ESEE-93/2, December 1993. Re-Evaluation of Strong-Motion Data from Greece, N.N. Ambraseys, A. Stavrakakis and K. Simpson, ESEE-93/3, December 1993. Reliability-Based Limits on Member Ductility, M. Manzocchi, M. Chryssanthopoulos and A.S. Elnashai, ESEE-94/1, January 1994. A Review of the Radiating Boundary Conditions for Earth Structures, S.G. Mahabadi and S.K. Sarma, ESEE-94/2, February 1994. On-Line Hybrid Testing of Steel Frames with High Strength Bracing Member, K. Takanashi, A.S. Elnashai, B.M. Broderick and K. Ohi, ESEE-94/3, March 1994. The Northridge (California) Earthquake of 17 January 1994: Observations and Correlative Response Analyses, B.M. Broderick, A.S. Elnashai, N.N. Ambraseys, J. Barr, R. Goodfellow and M. Higazy, ESEE-94/4, June 1994. Seismic Risk and Mitigation Planning in Egypt, A.S. Elnashai and M.M. Soliman, ESEE-94/5, July 1994. Attenuation of Spectral Ordinates in Europe, N.N. Ambraseys, K. Simpson and J.J. Bommer, ESEE-95/1, July 1995. Selected Engineering Seismology and Structural Engineering Studies of the Hyogo-Ken Nanbu (Kobe, Japan) Earthquake of 17 January 1995, A.S. Elnashai, J.J. Bommer, I. Baron, A.I. Salama and D. Lee, ESEE-95/2, September 1995. Effect of Modelling Assumptions and Input Motion Characteristics on Seismic Design Parameters of RC Bridge Piers, A.S. Elnashai and D. McClure, ESEE-95/3, September 1995. Surface Wave Magnitude Calibration for European Regions Earthquakes, N.N. Ambraseys and M. Free, ESEE-95/4, October 1995. Evaluation of Behaviour Factors for Frame-Wall RC Structures, L.M. Salvitti and A.S. Elnashai, ESEE-95/5, November 1995. Vertical Earthquake Ground Motion; Evidence, Effects and Simplied Analysis Procedures, A.S. Elnashai and A. Papazoglou, ESEE-95/6, December 1995. Engineering Prediction of Earthquake Strong-Motion Duration, J.J. Bommer and A. Martinez, ESEE-96/1, February 1996.
Re-appraisal of Large Central American Earthquakes, N.N. Ambraseys and R. Adams, ESEE96/2, February 1996. Seismicity of the Eastern Mediterranean and the Middle East, Part I, N.N. Ambraseys and D. White, ESEE-96/3, June 1996. An Earthquake Catalogue for Iran, N.N. Ambraseys and C. Melville, ESEE-96/4, June 1996. Effect of Model Conditions on the Response of Large RC Bridges, S. Dodd, A.S. Elnashai and G.M. Calvi, ESEE-96/5, August 1996. DRAIN-2D/90. Program for the inelastic analysis of plane structures subjected to seismic input - Users manual, A.J. Kappos, ESEE-96/6, November 1996. Experimental and Analytical Investigations into the Seismic Behaviour of Semi-Rigid Steel Frames, A.S. Elnashai, F.A. Danesh Ashtiani and A.Y. Elghazouli, ESEE-96/7, December 1996. Repair and Strengthening of RC Walls Using Selective Techniques, A.S. Elnashai and R. Pinho, ESEE-97/1, January 1997. Earthquake Ground-Motions Attenuation Relations for Stable Continental Intraplate Regions, M. Free, N.N. Ambraseys and S.K. Sarma, ESEE-98/1, February 1998. The Assessment of Total Seismic Moment, N.N. Ambraseys and S.K. Sarma, ESEE-98/2, March 1998. Review and Development of Response Spectra for Displacement-Based Seismic Design, J.J. Bommer, A.S. Elnashai, G.O. Chlimintzas and D. Lee, ESEE-98/3, March 1998. Inelastic Spectra and Ductility-Damping Relationships for Displacement-Based Seismic Design, B. Borzi, A.S. Elnashai, E. Faccioli, G.M. Calvi and J.J. Bommer, ESEE-98/4, May 1998. Observations on the Effects of the Adana-Ceyhan (Turkey) Earthquake of 27 June 1998, A.S. Elnashai, ESEE-98/5, August 1998. Inelastic Dynamic Response of RC Bridges to Non-Synchronous Earthquake Input-Motion, N. Tzanetos, A.S. Elnashai, F. Hamdan and S. Antoniou, ESEE-98/6, August 1998. Parameterised Displacement Spectra for Seismic Design, J.J. Bommer and A.S. Elnashai, ESEE98/7, August 1998. Seismic Performance and Cost-Benet Assessment of High Rise High Strength Concrete Buildings, B.T. Laogan and A.S. Elnashai, ESEE-98/8, November 1998. Surface-Wave Magnitude Reappraisal 10 - 44 N 18 - 70 E (Eastern Mediterranean and Middle East), N.N. Ambraseys and J. Douglas, ESEE-99/1, June 1999. Assessment of Inelastic Response of Buildings using Force and Displacement-Based Approaches, B.Borzi and A.S. Elnashai, ESEE-99/2, October 1999. The Kocaeli (Turkey) Earthquake of 17 August 1999: Assessment of Spectra and Structural Response Analysis, A.S. Elnashai, ESEE-99/3, October 1999.
The North Athens (Greece) Earthquake of 7 September 1999: Analytical Study of Structural Response and Spectra, A.S. Elnashai, ESEE-99/4, October 1999. Static Pushover versus Dynamic-to-Collapse Analysis of RC Buildings, A.M. Mwafy and A.S. Elnashai, ESEE-00/1, January 2000. INDYAS A Program for Inelastic Dynamic Analysis of Structures, A.S.Elnashai, R.Pinho and S.Antoniou, ESEE-00/2, June 2000. (N) Ductility of RC Members Constructed from High Strength Concrete and Reinforcing Steel, R.C.Goodfellow and A.S.Elnashai, ESEE-00/3, August 2000. Reappraisal of the Effect of Vertical Ground Motion on Response, N.N.Ambraseys and J.Douglas, ESEE-00/4, August 2000.

Esee00 4

Încărcat de

Informații document

Descriere originală:

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Esee00 4

Încărcat de

Drepturi de autor:

Formate disponibile

ESEE Report No.

00-4 August 2000

Engineering Seismology and Earthquake Engineering

Reappraisal of the effect of vertical ground motions on response

N. Ambraseys & J. Douglas

The effect of vertical acceleration on stability

2. RESULTS OBTAINED USING THE EXTENDED SET OF RECORDS

2. Results obtained using the extended set of records

Ground motion model used

The ground motion model assumed has the form: log y = b1 + b2 Ms + b3 d + bA SA + bS SS

2. Results obtained using the extended set of records

2. Results obtained using the extended set of records

Horizontal PGA (y = log ah )

2. Results obtained using the extended set of records

2.8 Vertical PGA (y = log av )

Vertical to horizontal absolute PGA ratio (y = log q = log av /ah )

2. Results obtained using the extended set of records

2. Results obtained using the extended set of records

Horizontal energy density (y = log Eh )

2. Results obtained using the extended set of records

Vertical energy density (y = log Ev )

Vertical to horizontal energy density ratio y = log qE = log Ev /Eh

Horizontal spectral acceleration (y = log SAh )

2. Results obtained using the extended set of records

Vertical spectral acceleration (y = log SAv )

Horizontal spectral acceleration (bending model) (y = log SAh,b )

2. Results obtained using the extended set of records

2. Results obtained using the extended set of records

Horizontal maximum absolute input energy (y = log Ih )

Vertical maximum absolute input energy (y = log Iv )

Vertical to horizontal spectral ratio (Absolute) (y = log qs = log SAv /SAh )

2. Results obtained using the extended set of records

This the the standard error of the logarithms

2. Results obtained using the extended set of records

Peak Ground Acceleration [g]

0.01 0.1 0.3 0.6 1 3 6 10 30 V 60 100

Fault Distance [km]

Peak Ground Acceleration [g]

Fault Distance [km]

Peak Ground Acceleration [g]

Fault Distance [km]

Spectral Acceleration [g]

0.4 A 0.3 S R A S R S A 0.1 R

Spectral Acceleration [g]

0 6 6.5 7 7.5 Magnitude (M )

Fault Distance (km)

0 6 6.5 7 7.5 Magnitude (M )

Fault Distance (km)

0 6 6.5 7 7.5 Magnitude (M )

Fault Distance (km)

0 6 6.5 7 7.5 Magnitude (M )

Fault Distance (km)

6.6 6.8 7 7.2 Magnitude [Ms]

6.6 6.8 7 7.2 Magnitude [Ms]

1.0 2,S 0.5

2,A 2,R 2,A

Peak Ground Acceleration [g]

1,S 1,A 1,R 1,S 1,A

1,R 0.02 0.1 0.3 0.6 1 3 6 10 30

Fault Distance [km]

2,S 2,A 2,S 2,A 2,R 2,R

Peak Ground Acceleration [g]