Keywords

1 Introduction

The development of seismic code for building design based on seismic hazard maps research was already conducted in 2012. The 2012 Indonesian Seismic Code [8] for building and non building design was developed based on seismic hazard map research carried on 2010. The research for 2010 Indonesian Seismic Hazard Maps (ISHM) was conducted by the Team for Revision of Seismic Hazard Maps 2010. The new ISHM-2017 was developed by The National Center for Earthquake Studies 2017 [5]. One of the most important information obtained from this research related with the new seismic sources that are detected and spread at the whole area of the country especially the additional of shallow crustal fault sources. A good example of additional shallow crustal fault sources from 5 (ISHM-2010) to 33 (ISHM-2017) are detected at the Java Island. There has been no significant improvement of 2017 subduction sources surrounding the Java Island.

SNI 1726:2012 [8] was developed based on the development of ISHM-2010. The improvement of SNI 1726 is still on-going, following the development of the new ISHM-2017. One of the important pieces of information needed for building resistance design is the design acceleration response spectrum (DARS) at the building location. Based on SNI 1726:2012 the DARS at the building location can be developed by conducting three basic steps: development of the maximum considered risk targeted acceleration spectrum (MCER) at the bedrock position; development of the surface acceleration spectrum (SMS, SM1 and PGAM); calculation of the DARS using SMS and SM1 surface spectral acceleration and following the same method described by SNI 1726:2012 and ASCE/SEI 7-16 [1].

The MCER spectral accelerations is divided into three: short period (Ss), long period (S1) and PGA (Peak Ground Acceleration) [4, 6]. All three acceleration spectra are calculated by combining Risk Targeted Ground Motion (RTGM) for a 1% probability of collapse in 50 years with 84th percentile deterministic seismic hazard analysis. Following the same method proposed by ASCE/SEI 7-16 [1], the new SS and S1 acceleration spectra are calculated by adjusting direction factors 1.1 and 1.3 for short and long period respectively. The logarithmic standard deviation (β) used for the SS and S1 acceleration spectrum calculation is 0.65.

The surface acceleration SMS, SM1 and PGAM are calculated by multiplying the three acceleration spectra Ss, S1 and PGA with site coefficients Fa, Fv and FPGA, respectively. For the new Indonesian seismic code, the three site coefficients are developed using the same values proposed by Stewart and Seyhan [9]. For a specific site or building location, the three site coefficients are calculated based on site class data and following the same method described by SNI 1726:2012.

The DARS for building resistance design is developed using two SDS and SD1 values which represent short and long period design spectra acceleration respectively. SDS and SD1 are equal to two third of SMS and SM1 values and are calculated using the same method described in SNI 1726:2012. The new Indonesian seismic code introduces a long transition period (TL) for developing the new DARS-2019. TL is a period on DARS-2019 which is separating the constant displacement and constant velocity and usually greater than 4 s [2, 3, 7]. DARS-2019 with TL values is developed using the same procedure proposed by ASCE/SEI 7-16 [1].

The MCER (SS, S1 and PGA) and TL data of Indonesia (national data) are developed by PUSGEN which are spread from (94° longitude, 8° latitude) through (142° longitude, −12° latitude) with a 0.1° interval (±11.05 km) in both directions. The SS, S1, PGA and TL at specific positions are usually obtained from the same values at the closest distance points (national data positions). Figure 1 shows the SS map and Fig. 2 shows the S1 map and the positions of national data surroundings the study area.

Fig. 1
figure 1

MCER–SS map of Central Java and Yogyakarta Provinces

Full size image
Fig. 2
figure 2

MCER–S1 map of Central Java and Yogyakarta Provinces

Full size image

This paper presents the development of MCER and TL values at specific location (cities) by conducting three different methods. The first method (method-1) for developing MCER and TL values is conducted by calculating the average of MCER and TL at the four closest points. The second method (method-2) is conducted by adjusting the MCER and TL values at the closest point. The final method (method-3) is conducted by using the weighting factor calculated from the four closest points. The weighting factor is calculated using four distance values at the four closest points. The MCER and TL values are calculated at 35 cities in Central Java and Yogyakarta regions. The positions of all cities can be seen in Figs. 1 and 2. The DARS-2019 for site class SC, SD and SE at specific city are then developed based on the MCER (SS and S1) and TL values at the specific point and calculated based on ASCE/SEI 7-16.

2 Methodology

The development of DARS-2019 for site class SC, SD and SE at specific locations or sites is carried out using the following algorithms:

  1. 1.

    Find the MCER acceleration spectrum (SS, S1 and PGA) and TL from national data points.

  2. 2.

    Find all points with a maximum 15 km distance from the site position and then find 4 closest points to the site position.

  3. 3.

    Find the MCER (SS, S1 and PGA) and TL values of the four closest points.

  4. 4.

    Sort the MCER and TL data of four points based on the distance of each point to the site position from minimum to maximum.

  5. 5.

    Find the average of MCER and TL values of all four points (method-1).

  6. 6.

    Find the MCER and TL values of the closest point (method-2).

  7. 7.

    Find the MCER and TL values of site based on the weighting factor of four points (method-3) following “Eqs. (1)–(3)”. The “w(i)” in Eq. 1 is the weighting factor of points number “i” where i = 1–4. The d(i) value is the distance of point no “i” to the site position. The MCER(i) and TL(i) are MCER and TL value at point no “i”.

  8. 8.

    Calculate the SMS, SM1 and PGAM following the same method proposed by SNI 1726:2012 and utilizing the site factor proposed by [9].

  9. 9.

    Calculate the SDS and SD1 using the same method proposed by SNI 1726:2012.

  10. 10.

    Develop the DARS-2019 based on ASCE/SEI 7-16 [1].

    $${\text{w(i)}} = \frac{{ 1/{\text{d(i)}}}}{{\sum\nolimits_{{{\text{i}} = 1}}^{{{\text{i}} = 4}} { 1/{\text{d(i)}}} }}$$
    (1)
    $${\text{MCER}}\,\, = \,\,\sum\limits_{{{\text{i}} = 1}}^{{{\text{i}} = 4}} {{\text{w}}({\text{i}})*{\text{MCER}}({\text{i}})}$$
    (2)
    $${\text{T}}_{\text{L}} = \sum\limits_{{{\text{i}} = 1}}^{{{\text{i}} = 4}} {{\text{w}}({\text{i}})*{\text{T}}_{\text{L}} ({\text{i}})}$$
    (3)

Table 1 shows an example of MCER and TL calculation for Yogyakarta, the capital city of Yogyakarta Province, with coordinates (110.35, −7.8). Based on the distance of the site to national data points there are six points with a distance less than 15 km to Yogyakarta. Based on method-1 the SS, S1, PGA and TL values of Yogyakarta are 1.2635 g, 0.54475 g, 0.53 g and 13 s respectively. By applying the second method (method-2) the SS, S1, PGA and TL values of Yogyakarta are 1.069 g (1 g = 9.81 m/s2), 0.493 g, 0.465 g and 20 s respectively. Finally by conducting method-3 the SS, S1, PGA and TL values of Yogyakarta are 1.211384 g, 0.533551 g, 0.516982 g and 13 s respectively. Table 2 shows another example of MCER and TL calculation for Semarang, the capital city of Central Java Province, with coordinates (104.0, −7.0). Following the same steps conducted for Yogyakarta, there are 5 (five) closest points to Semarang. Table 3 shows three different pair of SS and S1 results calculated at Semarang using three different methods.

Table 1 MCER-SS. S1, PGA and TL example calculation for Yogyakarta
Full size table
Table 2 MCER-SS. S1, PGA and TL example calculation for Semarang
Full size table
Table 3 SS and S1 acceleration spectrum calculated using three different methods
Full size table

By using SS and S1 values and applying site factor proposed by [9] and following ASCE/SEI 07-2016, the DARS-2019 for site class SC, SD and SE for Yogyakarta are presented in Fig. 3. Figure 4 shows an example of DARS-2019 for site class SC, SD and SE for Semarang. All DARS-2019 as can be seen in Figs. 3 and 4 are developed from SS and S1 results calculated using method-3.

Fig. 3
figure 3

DARS-2019 of Yogyakarta for site class SC, SD and SE calculated based on SS and S1 method-3

Full size image
Fig. 4
figure 4

DARS-2019 of Semarang for site class SC, SD and SE calculated based on SS and S1 method-3

Full size image

3 Results and Discussion

The calculation of SDS and SD1 for the development of DARS-2019 depends not only on the SS and S1 values but also depends on the site factor Fa and Fv values. The Fa and Fv values for specific site soil class (SC, SD and SE) are usually developed by linear interpolation and depends on the SS and S1 values. Table 4 shows the SS and S1 at 35 cities in the study area. As can be seen in Table 4 the difference of SS and S1 values are obtained due to different methods used for calculating SS and S1. The analysis for evaluating the probability best performance of three methods is conducted at 35 different cities. Figures 5 and 6 show the results of SS and S1 performance calculations using the three different methods.

Table 4 Improvement of SDS and SD1 for site class SC
Full size table
Fig. 5
figure 5

Difference of SS values in terms of closest site distance calculated using three methods

Full size image
Fig. 6
figure 6

Difference of S1 values in terms of closest site distance calculated using three methods

Full size image

Figure 5 shows the distribution of the SS difference values calculated using method-1 and method-2 (DifSS-12), the difference between method-1 and method-3 (DifSS-13) and the difference between method-2 and method-3 (DifSS-23). The SS difference distribution is developed in terms of closest site distance to national data positions. As can be seen in Fig. 5 when the distance of the site position to national data positions less than 2 km, there is no significant difference in SS values calculated using three different methods. However, when the distance of the site to national data greater than 2 km the SS values calculated using method-1 and method-3 are almost equal while the SS values calculated using method-2 is differ from the two other methods. It seems that the SS values calculated using method-3 have a better performance compared to another two methods.

Figure 6 shows the distribution of S1 difference values DifS1-12, DifS1-13 and DifS1-23. In terms of site distance, the distribution of DifS1-12, DifS1-13 and DifS1-23 are almost equal with DifSS-12, DifSS-13 and DifSS-23 distributions. The performance of the S1 values calculated using mehod-3 is better than method-1 or method-2.

A comparative study was conducted to evaluate the improvement of DARS calculated using [8] (DARS-2012) and DARS-2019. The analysis was conducted for 35 cities in Central Java and Yogyakarta Provinces for three different site classes (SC, SD and SE). DARS-2019 is developed using method-3. Figures 7 and 8 show the DARS-2019 and DARS-2012 for Semarang and Yogyakarta respectively. The maximum period of DARS-2019 display in Figs. 7 and 8 are equal to 4 s and is adjusted to the same period used by SNI 1726:2012.

Fig. 7
figure 7

DARS for Semarang

Full size image

As can be seen in Fig. 7, there is no significant improvement in the DARS for Semarang for site class SC, SD and SE. For site class SC and SE the DARS-2019 is slightly bigger than the DARS-2012. However, for site class SD, the DARS-2019 is slightly lower than the DARS-2012. The maximum improvement (increasing or decreasing) of DARS-2019 compared to DARS-2012 is less than 0.1 g. The improvement of DARS for Yogyakarta is slightly bigger than Semarang. As can be seen in Fig. 8, for site class SC and SD the DARS-2019 for Yogyakarta is slightly bigger than for DARS-2012. For site class SE the DARS-2019 of Yogyakarta is lower than the DARS-2012. The maximum improvement of the DARS-2019 compared to the DARS-2019 for site class SC is less than 0.2 g. However, the maximum improvement of DARS-2019 compared to Dars-2012 for site class SD and SE is less than 0.1 g.

Fig. 8
figure 8

DARS for Yogyakarta

Full size image

Tables 5, 6 and 7 illustrate the improvement of SDS and SD1 values for site classes SC, SD and SE respectively and calculated for developing DARS-2019 and DARS-2012 for Semarang and Yogyakarta. Positive or negative sign (±) inside this table represents increasing or decreasing of DARS-2019 compared to DARS-2012. Compared to SDS 2012, the SDS 2019 of Semarang city is increasing 8.32% and 17.33% for site classes SC and SE respectively. However, for site class SC the SDS value for Semarang is decreases 6.50%. For Yogyakarta SDS 2019 value for site classes SC, SD and SE are decreased 22.43%, 2.33% and 5.29% respectively. Compared to DARS-2012, the SD1 values for site class SC, SD and SE of DARS-2019 for Semarang are increased 19.21%, 28.68% and 6.72% respectively. For Yogyakarta, the SD1 value for DARS-2019 increased by 30.33%, 36.90% and 7.38% for site classes SC, SD and SE respectively.

Table 5 Improvement of SDS and SD1 for site class SC
Full size table
Table 6 Improvement of SDS and SD1 for site class SD
Full size table
Table 7 Improvement of SDS and SD1 for site class SE
Full size table

Base on the analysis conducted at 35 cities the largest improvement of SDS and SD1 are detected at five cities located close to Opak Fault Trace Yogyakarta. The SDS and SD1 of DARS-2019 at these five cities are larger than the SDS and SD1 of DARS-2012. The SDS values are increased in between 0.144 g and 0.297 g and the SD1 values are increased in between 0.104 and 0.148 g. The most significant improvement is detected at site class SC. However, the improvement of SDS and SD1 for site classes SD and SE at five cities are smaller than for site class SC and less than 0.1 g. Table 8 shows the improvement of SDS and SD1 values at five cities and Fig. 9 shows the position of all five cities. All five cities are located less than 25 km distance from the Opak fault trace. Klaten and Boyolali are located at Central Java Province. However, another three cities, Bantul, Wates and Yogyakarta are located at Yogyakarta Province. As can be seen in Table 8 the largest improvement of SDS value at Klaten is close to 0.3 g (43.5%) and the improvement of SD1 value at this city is 0.144 g (40.2%).

Table 8 Improvement of SDS and SD1 for site class SC at five cities
Full size table
Fig. 9
figure 9

The study area (a) and five cities with maximum improvement of SDS and SD1 values (b)

Full size image

4 Conclusions

The development of design acceleration response spectrum (DARS) for 35 cities in the Central Java and Yogyakarta Provinces, Indonesia, was performed in this study. The objective of this study was to evaluate the improvement of DARS-2019 compared to DARS-2012. The study was performed due to the improvement of the Seismic Hazard Maps of Indonesia 2017. The study was performed first by calculating the MCER acceleration spectra (SS and S1) and conducting three different methods. The DARS-2019 for site classes SC, SD and SE were then developed using the same method proposed by SNI 1726:2012 and ASCE/SEI 7-16.

Based on the analysis conducted at 35 different cities in Central Java and Yogyakarta Provinces, method-3 (weighting factor method) for developing of SS and S1 values at specific location has a better performance compared to two different methods (method-1 and method-2). Method-1 is developed using the averages of four different values obtained from the four closest points. Method-2 is developed based on the value at the closest point. All three methods are conducted in this study because the site coordinate position not always equal to the coordinates of national data points (developed by PUSGEN).

On average the DARS-2019 developed at 35 cities in Central Java and Yogyakarta Provinces are almost equal compared to the DARS-2012. Based on the analysis conducted at 35 cities, the improvement of the DARS-2019 compared to the DARS-2012 for site classes SC, SD and SE are less than 0.1 g except for site class SC at five cities located close to Opak fault trace. All five cities are located less than 25 km distance toward Opak fault trace. The maximum improvement value is detected at Klaten with SDS value close to 0.3 g (43.5%) and the improvement of SD1 value at this city is 0.144 g (40.2%).