SEISMIC DESIGN CRITERIA FOR R.C. STRUCTURES IN SAUDI ARABIA: WHY DIFFERENT FROM THE UBC AND ACI REQUIREMENTS Abdulrahim M. Arafah, Mohammed S. Al-Haddad, and Rajeh Z. Al-Zaid Associate Professors, Civil Engineering Department, College of Engineering, King Saud University, Riyadh ABSTRACT: This paper presents the major differences between in the design criteria that recommended for the Kingdom and those required by the Uniform building Code and ACI 318M design code. These differences are classified to analytical and design aspects. The reasons of these differences are attributed to reliability, economical and quality aspects. The recommended deviations are intended to increase the structural reliability, strength, ductility and integrity.
1. INTRODUCTION Recently, there has been an increasing concern about the seismic activity along the western coast of the Kingdom. Several studies were conducted to estimate the level of the seismic risk in the Kingdom [1,2] and develop rational design criteria for reinforced concrete structures [3]. The seismic hazard analysis for the Kingdom was performed [1,2]. A zonation map, as shown in Fig. 1, was developed for the Kingdom based on the peak ground acceleration, PGA, values calculated for 50 years service lifetime with 10% probability of being exceeded.
Figure 1. Seismic Zonation Map for the Kingdom [1,2]
Following the Uniform Building Code (UBC 1991) model [4], the Kingdom was divided into four zones with seismic zone numbers (SZN) of 0, 1, 2A and 2B as shown in Table 1. The framework of ACI 318M-95 [5] code was adopted for the design of reinforced concrete structures in the Kingdom [3]. This paper highlights the reasons why the design criteria recommended for the Kingdom are different from those required by the UBC [4] and ACI318M [5] design codes. Table 1 : Seismic Zone Number (SZN) and Corresponding PGA According to UBC [4] SZN 0 1 2A 2B
PGA in g's < 0.05 0.05 to 0.10 0.10 to 0.15 0.15 - and above
2. DIFFERENCES IN ANALYSIS ASPECTS According to the UBC [4], the minimum design base shear , V, is calculated from,
V =
ZIC W Rw
(1)
where Z is the seismic zone factor, I is the importance factor, Rw is the system performance factor and W is the total seismic dead load. The factor C is a numerical factor which depends on the at site soil characteristics and the fundamental period of the structure. The proposed criteria [1,2] involved two modifications to this formula as explained in the following. 2.1 Earthquake Risk level According to the ACI 318M [5] table number R21.2.1, the zones of SZN = 0 and 1 are considered of no and low risk levels, respectively. The zones of SZN = 2A and 2B are considered as areas with moderate risk level whereas the zones of SZN = 3 and 4 are considered to be high seismic risk areas. Thus according to the seismic zonation map [1,2] most of the Kingdom regions fall in the zone of no and low risk level. Areas along the western coast, especially in the northwest and southwest are considered to be of moderate risk level. According to the ACI 318M [5] the design requirements depend on the risk and classified into three categories: special, intermediate and ordinary requirements for high, moderate, and low risk levels regardless the occupancy type. Al-Haddad et. al. [1,6] introduced the concept of seismic performance category, SPC, to identify the risk level and corresponding design requirements. The concept of SPC classifies structures according to importance (essential, special and standard) and risk level of its location as shown in Table 2 where A, B and C are corresponding to low, moderate and high risk levels as specified in ACI 318M [4]. As it is clear, the proposed modification is on the conservative side especially for essential structures.
Table 2 Seismic Performance Categories [1,6] OC
ES
SP
ST
2B
C
C
B
2A
B
B
A
1
B
A
A
0
A
A
A
SZN
Note: OC is for occupancy category. ES, SP, and ST are , respectively, for essential, special, and standard occupancy categories as specified in UBC. 2.2 System Performance Factor, Rw It is essential to design a reinforced concrete member with sufficient ductility to avoid brittle failure in flexure particularly for seismic resistant design. The current philosophy of seismic design of moment resisting reinforced concrete frames is based on formation of plastic hinges at the critical sections of a frame under the effect of substantial load reversals in the inelastic range. Therefore, the system performance factor, Rw , in the UBC equation for the design base shear accounts for the inelastic behavior and reduces the base shear force depending upon the type of the structural system and its level of ductility. Fig. 2 shows the relationship between ductility and performance factor according to equal displacement and equal energy principles.
(a) Equal displacement
(b) Equal energy
Figure 2 Relationship between ductility and force reduction factor
The proposed criteria [1,6] recommended reducing the factor Rw as required in UBC 1991 [5] as shown in Table 3. This reduction means an increase in the design base shear force. In latest edition of UBC (1997) Rw values are reduced to values consistent with the values proposed for the Kingdom. However the based shear equation was also modified.
Table 3 System Performance Factors for Reinforced Concrete Structures [1] Structural System
Lateral Load Resisting System
Rw (UBC) 12
Rw (KSA) 8
7
5
Ordinary MRSF (low seismic risk)
5
2
Reinforced Concrete
8
6
Reinforced Masonry
8
4
Concrete with Special MRSF
12
8
Concrete with Intermediate MRSF
9
5
Reinforced Concrete
6
4
Reinforced Masonry
6
3
Moment Resisting Space Special MRSF (high seismic risk) Frame, MRSF Intermediate MRSF (moderate seismic risk)
Shear Wall System
Dual System
Bearing Wall System
The modifications in the criteria of the seismic risk level and the system performance factor are attributed to (1) reliability, (2) economical, and (3) quality aspects. A brief discussion of these aspects is presented here after. 1.
High uncertainties associated with seismic hazard assessment involved in the development of the zonation map. This is mainly attributed to limited information on seismotictonics, past seismic activity, and ground attenuation in the Kingdom.
2.
In the Kingdom, it is usually recommended to be on the safe side and increase the safety margin in the design process even though this will slightly increase the initial cost of the structural system. This mainly attributed the fact that structural repair and rehabilitation is very costly process in the Kingdom.
3.
The quality control and quality assurance programs in the Kingdom are far behind those in the industrial countries. Majority of designers and contractors do not pay enough attention to the design and construction details. Therefore, such unacceptable low levels of practice adversely affect the structural strength, ductility, and integrity.
3. DIFFERENCES IN DESIGN ASPECTS 3.1 Flexural Design for Moderate Risk Levels To ensure that the failure of reinforced concrete beams is initiated and proceeded by yielding of tensile steel, the ACI 318M, Section 10.3.3 for non-seismic conditions limits the maximum tensile reinforcement ratio (ρ − ρ') to be not more than 0.75 ρb where ρ, ρ’, and ρb are the tension, compression and balanced reinforcement ratios, respectively. Reinforced concrete sections at the flexural limit state may fail by concrete crushing even when they are reinforced below the maximum reinforcement ratio specified by the ACI Code [5]. One of the factors contributing to this uncertainty is the variability of the strength of concrete and reinforcing steel. The margin provided by the ACI criterion for maximum reinforcement ratio does not ensure a ductile failure especially when the mean-to-nominal ratio of yield strength, λs, is high. In the Kingdom two types of concrete can be identified: the ready-mix (RM) concrete and the at-site mechanically-mixed (SM) concrete. Arafah [7], estimated the statistics of RM concrete and the SM concrete under the prevailing concreting practices in the Kingdom. The results from 636 strength tests on RM concrete indicated that mean-to-nominal ratio of concrete strength, λc , and the strength coefficient of variation, Vc , are about 1.0 and 20 percent respectively, and the strength is well represented by the normal distribution. The results of 45 strength tests on SM concrete indicated that λc and Vc are about 0.85 and 40 percent respectively, and concrete strength is well represented by the log-normal distribution. Al-Behairi [8] investigated the probabilistic characteristics of steel bars produced by the Saudi Steel and Iron Company through the bar quenching process. It was concluded that the mean-to-nominal yield strength of reinforcing steel, λs , and the strength coefficient of variation, Vs , are 1.34 and 4.3 percent respectively. The yield strength is found to be well represented by the normal distribution function. Since, the mean yield strength of Saudi steel is higher than its nominal value (420 MPa) and the mean compressive strength of saudi concrete is lower than its nominal value, it is recommended to replace the nominal strengths by their respective mean values [9]. The modified balanced reinforcement ratio, ρb, becomes, ρb = !ﺧﻄﺄ
(2)
where f c' and fy are the nominal compressive strength of concrete and nominal yield strength of reinforcing steel, and λc is the mean-to-nominal ratio for concrete which is about 1.0 and 0.85 for RM and SM concretes, respectively. The mean-to-nominal ratio for Saudi steel, λs , is about 1.34. This approach reduces the value of balanced reinforcement ratio and increases the ductility of reinforced concrete beams.
As an alternative approach, the nominal values for the concrete and reinforcement strengths can be employed in the ρb equation as specified in ACI 318M [4] and limit the maximum ratios of (ρ−ρ')/ρb to about 0.6 and 0.4 for RM and SM concretes, respectively [9]. 3.2 Seismic Design for High Seismic Risk The current philosophy of seismic design of moment resisting reinforced concrete frames, in high seismic risk regions, is based on formation of plastic hinges at the critical sections of a frame under the effect of substantial load reversals in the inelastic range. The approach is known as the Capacity Design Procedure. The following features characterize the capacity design procedure [10]: 1.
Potential plastic hinge regions within the structure are clearly defined. These are designed to have dependable flexural strengths as close as practicable to the required strength. Subsequently, these regions are carefully detailed to ensure that estimated ductility demands in these regions can be reliably accommodated. This is achieved primarily by close-spaced and well-anchored transverse reinforcement.
2.
Undesirable modes of inelastic deformation within members containing plastic hinges are inhibited by ensuring that the strengths of these modes exceeds the capacity of the plastic hinges at over-strength.
3.
Potentially brittle regions, or those components not suited for stable energy dissipation, are protected by ensuring that their strength exceeds the demands originating from the over-strength of the plastic hinges. Therefore, these regions are designed to remain elastic irrespective of the intensity of the ground shaking or the magnitudes of inelastic deformations that may occur.
The sequence of capacity design process includes: beam flexural design, beam shear design, column flexural strength, transverse reinforcement for columns, and beam-column joint design. It should be noted that only for the case of beam flexural design will design actions correspond to the code level of lateral seismic forces. For beam shear and all column design actions, the design forces are calculated on the assumption of beam plastic hinge sections developing maximum feasible flexural strength using simple equilibrium relationships. To ensure that the plastic hinges form at the ends of beams rather than in columns, ACI 318M design code requires that the sum of flexural strength of columns at any joint shall be 20 percent larger than that for beams connected to the same joint. Fig. 3a shows the energydissipating mechanism employing capacity design procedure and Fig. 3b shows a mechanism in which the plastic hinges formed in the columns causing the undesirable soft story mode of failure.
Figure 3 Comparison of energy-dissipation mechanisms 3.2.1 Shear Reinforcement in Beams and Columns: ACI 318M-95 [4], Section 21.2.4.1, requires that compressive strength fc' of the concrete shall be not less than 20 MPa. Therefore, SM concrete should not be permitted for structures with C performance category. ACI 318M-95, Section 21.2.5, requires that (a) the actual yield strength based on mill tests does not exceed the specified yield strength by more than 120 MPa , and (b) the ratio of the actual ultimate tensile strength to the actual tensile yield strength is not less than 1.25. These two conditions are not met by the steel produced by Saudi Steel and Iron Company [8]. The first requirement limits the magnitude of the actual shears that can develop in a flexural member in the inelastic range. Use of longitudinal reinforcement with strength substantially higher than that assumed in design will lead to higher shear and bond stresses at yield moments. These conditions may lead to brittle failures in shear or bond and should be avoided even if such failures may occur at higher loads than those anticipated in design. Therefore, a ceiling is placed on the actual yield strength of the steel. The second requirement is intended to ensure steel with a sufficiently long yield plateau. ACI 318M, Sections 21.3.4 and 21.4.5 requires using a factor of 1.25 for the reinforcement yield strength, fy , in calculating the design forces for shear strength of beams and columns. same factor for the joint design. Knowing that the mean to nominal value of U.S. steel is about 1.12, this factor from the statistical point of view means replacing the nominal value of yield strength, which is about the 5th percentile of the strength distribution, with the 95th percentile. This reduces the probability of exceeding the design strength by not more than 5 percent. When the same philosophy is employed to the Saudi reinforcing steel the factor 1.25 shall be increased to 1.5.
3.2.2 Design of Joints: ACI 318M, Section 21.5.1 specifies that forces in longitudinal beam reinforcement at the joint face shall be determined by assuming that the stress in the flexural tensile reinforcement is 1.25fy. consequently, joint shear forces generated by flexural reinforcement is calculated. To account for the high mean to nominal
ratio of Saudi steel, it is recommended to increase the design yield strength 1.25 fy specified by ACI code for seismic design of shear to 1.5 fy.. ACI 318M, Section 21.5.1.4, requires that where longitudinal beam reinforcement extends through a beam-column joint, the column dimension parallel to the beam reinforcement shall not be less than 20 times the diameter of the largest longitudinal bar, i.e.,
hcolumn ≥ 20 d beam bar
and
hbeam d column bar
≥ 20
(3)
In the Kingdom, it is recommended to increase the factor of 20 to 25 to account for the large mean to nominal ratio of the Saudi reinforcing steel. This condition increases the depth requirements for both columns and beams of the frame system. 3.2.3 Development Length of Bars in Tension: ACI 318M, Section 21.5.4.1 requires that the development length ldh for a bar with a standard 90-deg hook shall not be less than 8db, 150 mm, and the length required by
l dh =
f y db 5.4 f c'
(4)
for bar sizes No. 10 through No. 36. In the Kingdom the factor 5.4 should be reduced to 4.5. This will increase the development length to account for the high mean to nominal ratio of the Saudi reinforcing steel. 4. CONCLUSIONS This paper presents the major differences between the design criteria recommended for the Kingdom and those required by the Uniform building Code and ACI 318M design code. These differences include, 1. The design requirements are based on the concept of seismic performance category rather than the zone factor. 2. The system performance factor in the UBC base shear equation is reduced which means an increase in the design base shear force. 3. For flexural design under moderate seismic risk, the maximum tension reinforcement ratio is reduced to account for high yield strength in the Kingdom and improve flexural ductility. 4. In structures with C performance category, at-site mechanically mixed concrete is not permitted. 5. The design yield strength of 1.25 fy employed by the ACI Code for seismic design is increased to 1.5 fy to account for the high mean yield strength of the Saudi reinforcing steel. This factor is applied for the design of shear in beams and columns, design of joints and calculation of development length of bars in tension. 6. The column dimension parallel to the beam longitudinal reinforcement shall not be less than 25 times the diameter of the largest longitudinal bar instead of 20. 7. The factor 5.4 in the equation of development length for a bar with a 90-deg standard hook as specified in ACI code should be reduced to 4.5. This will increase the development length to account for the high mean yield strength of the Saudi reinforcing steel.
These differences were attributed to reliability, economical, and quality aspects. They account for the properties of concrete and reinforcing steel produced in the Kingdom. These modifications are intended to increase the structural strength, ductility and integrity. Acknowledgement This paper is part of a study sponsored by King Abdul-Aziz City for Science and Technology under grand number AR-11-57. The authors would like to express their thanks and appreciation for this support. References [1] AL-Haddad, M., Siddiqi, G.S., Al-Zaid, R., Arafah, A., Necioglu, A., and Turkelli, N., " A Study Leading to a Preliminary Seismic Design Criteria, for the Kindom," Final Report, KACST project No. AR-9-31, Riyadh, 1992. [2] Al-Haddad, M., Siddiqi, G.S., Al-Zaid, R., Arafah, A., Necioglu, A., and Turkelli, N., “A Basis for Evaluation of Seismic Hazard and Design Criteria for Saudi Arabia”, Journal of Earthquake Engineering Research Institute, EERI, Spectra, Vol. 10, No. 2, May 1994, Okland, California. [3] Al-Zaid, R., Arafah, A. M., AL-Haddad, M., Siddiqi, G. H., and Al-Sulimani, G., "Development of a National Design Code for RC Buildings -Phase II," Third Progress Report, KACST Project No. AR-12-58, Riyadh, 1994, 206 pp. [4] "Uniform Building Code", International Conference of Building Officials, California, USA, 1991. [5] ACI Committee 318, "Building Code Requirements for Reinforced Concrete (ACI 318M- 95) and Commentary ACI 318M", American Concrete Institute, Farmington Hill,,1989, 353 pp. [6] ] Al-Haddad, M., Siddiqi, G.S., “Seismic Design Recommendations for Building Structures in Saudi Arabia”, Journal of King Saud University – Engineering Science [1], Vol. 7, pp. 25-45, Riyadh, 1995. [7] Arafah, A. M., "Statistics for Concrete and Steel Quality in Saudi Arabia," Magazine of Concrete Research, London, Vol. 49, No. 180, September 1997, pp. 185-194. [8] Al-Behairi, S., "Mechanical Properties of Saudi Rebar and their Effect on Behavior of RC Members," Master Thesis, Civil Engineering Department, College of Engineering, KSU, Riyadh, November 1994. [9] Al-Nufaie, A., "Probabilistic Study of Brittle Flexural Failure in R.C. Beams Employing Saudi Materials," Master Thesis, Civil Engineering Department, College of Engineering, KSU, Riyadh, November 1996. [10] Pauley, T. and Priestly, M., Seismic Design of Reinforced Concrete and Masonry Buildings. John Wiley and Sons, Inc. , N.Y., 1995.
