LEARNING FROM LOCAL WISDOM : FRICTION DAMPER IN TRADITIONAL BUILDINGS IN INDONESIA

Indonesia is situated in the so called “Ring of Fire” where earthquake are very frequent. Despite of all the engineering effort, due to the March 28, 2005 strong earthquake (8.7 on Richter scale) a lot of modern buildings in Nias collapsed, while the traditional Northern Nias house (omohada) survived without any damage. Undoubtedly many other traditional buildings in other area in Indonesia have survived similar earthquake. Something in common of the traditional building are the columns which usually are not fixed on the ground, but rest on top of flat stones. In this paper some traditional building are subjected to non linear time history analysis to artificial earthquake equivalent to 500 years return period earthquake. This study shows that apparently the columns which rest on top of flat stone acts as friction damper or base isolation. The presence of sliding at the friction type support significantly reduces the internal forces in the structure.


INTRODUCTION
Indonesia is situated in the so called "Ring of Fire" where earthquakes are very frequent. Although the first Indonesian earthquake code was introduced in 1971, after more than forty years, despite all effort to disseminate the principle of good earthquake engineering design (Lumantarna, 2007), in the recent earthquake, such as Padang, October 2009, Bengkulu, September 2007, Yogya, Mei 2006, Nias, March 2005, a lot of modern buildings collapsed (Figure 1). On the other hand traditional building such as Northern Nias, omo hada ( Figure 2) survived without any damage (Lase, 2005).
Undoubtedly in every corner of Indonesia, there is traditional building that has survived the test of time through earthquakes. Just to mention a few, Figures 3 to 6 show some traditional building in different area. It can be seen from the seismic map in Figure 3, that these traditional buildings are located in high seismicity area. Studying the traditional buildings reveal that things in common in all traditional buildings are; the 2 elevated floor, wooden building, and columns that are not fixed on the ground but only placed on top of flat stones ( Fig. 8 and 9).
The authors suspect that beside the light weight wood structure, the column base which act as friction damper, reduces the effect of the seismic force to the upper structure. The behavior of oma hada with two base condition, i.e.: fixed base and base with Coulomb friction damper has been reported by Pudjisuryadi et al (2007), while the behavior of uma lengge was reported by Tiyanto and Shia (2012) in an undergraduate theses supervised by the authors.    Figures 8 and 9 show the column base of omo hada and uma lengge respectively. The schematic structural system of omo hada and uma lengge, is shown in Figures 10 and 11 respectively. Due to the difficulty in modelling the actual member connections it was decided to use rigid connection, only the diagonal bracing are pinned. These assumptions are considered reasonable, since the rigid connection will results in a higher column base shear. To study the effect of the column base, the two structures are modelled using fixed base and Coulomb friction damper and the structures are subjected to certain artificial ground acceleration and analysed using dynamic nonlinear time history analysis. SAP2000 Nonlinear was used for the time history analysis. The ground acceleration used in the analysis is spectrum consistent ground acceleration which is modified from El Centro 18 May 1940 NS to the acceleration response spectrum specific to the area where the buildings are. The modification of the earthquake record is performed using RESMAT, a software developed at Petra Christian University, Surabaya, Indonesia (Lumantarna and Lukito, 1997). The modified El Centro ground acceleration to be used in the analysis of uma lengge is shown in Figure 12, while the response spectra of the modified and the original El Centro 18 May 1940, NS component along with the target response spectrum are shown in Figure 13.

ANALYSIS RESULT
The member internal stresses due to load combination of 1Dead Load + 1Live Load + 1Quake of the two models are checked with respect to allowable stresses of the wood according to Indonesian standard (NI-5 PKKI 1961). The analysis result of oma hada has been reported elsewhere (Pudjisuryadi et al, 2007). The comparison of stress ratio of the fixed base and the Coulomb friction base of omo hada is presented here in Table 1, while the comparison of stress ratio of the fixed base and the Coulomb friction base of oma lengge is presented in Table 2 In Tables 1 and 2, stress ratio bigger than one suggest that the particular member exceeds its capacity. The highlighted numbers in Table 1 shows that the stress ratio in the Diwa (bracing) and Ehomo (column) reduce tremendously when the column bases are changed from fixed support to Coulomb fiction base support. Table 2 shows that the stress ratio of the column, diagonal bracing, and first floor beam (highlighted) which fail in fixed base, survive if Coulomb friction is used. It can be seen that compared to the fixed base, the Coulomb friction base reduces the stresses in the column and diagonal members markedly.

CONCLUDING REMARKS
Observing the results presented in Table 1 and 2, it can be concluded that the Coulomb friction base isolation of omo hada and oma lengge performs very well in reducing internal forces. If the columns are fixed on the ground, both traditional building would not have survived the 500 years return period earthquake As an aftermath, it may be worth to investigate if one departs from the traditional foundation design of modern low rise building (Fig. 15) by deleting the anchorage of the tie beam to the foundation (Fig. 16). It is interesting to see if the second option performs better during earthquake.