COMPARISON OF RECENT U.S. SEIsMic CODES By R. D. McIntosh’ and S. Pezeshk,” Members, ASCE. ApernAct: This paper compares National Earthquake Hazards Reduction Program (NEHRP), Structural En- gineers Association of Califomia (SEAOC), ASCE 7, and Uniform Building Code (UBC) seismic design pro- Visions to address the differences in ther pi sophies and af bilities. These documents are compared by focusing on issues such a (1) purpose of earthquake codes: (2) type of document and target audience; (3) lateral forces; and (4) analysis provisions. NEHRP and ASCE 7 documents are based on strength design while UBC and SEAOC are based on allowable or working stress design. There are other fundamental differences among these documents such as the requited methods of analysis, building importance, detailing requirements, soil amplification factors, drift control and P-delta amplification, and the method of assigning an importance factor. Several tables and graphs are presented toi INTRODUCTION ‘This paper presents a comparison of the seismic design pro- visions in documents published by four organizations —Build- ing Seismic Safety Council, Structural Engineers Association of California (SEAOC), ASCE, and International Conference fof Building Officials. ‘The documents are (1) NEHRP Rec- ‘ommended Provisions for the Development of Seismic Regu lations for New Buildings (1992, 1995) 1991 and 1994 edi ‘Minimum Design Loads for Buildings (1995); (3) Uniform Building Code Volume 2, Structural Engineering Design Provisions (1994); and (4) SEAOC Recommended Lateral Force Requirements ‘and Commentary (1990). Throughout the remainder of this ppaper the foregoing documents shal be referenced as NEHRP, ‘ASCE, UBC, and SEAOC, respectively. The comparisons here ‘will focus on the differences in the philosophies and applica- bilities of these documents, ‘The organization of this paper will follow the format and periodically reference paper by Luft (1989). The reader should refer to the Luft paper fora brief history of the devel- ‘opment of the current earthquake codes and t0 sce how the evolution of seismic design provisions has accelerated since 1989 when the paper was published, PURPOSE OF EARTHQUAKE CODES. It is important to understand the expressed or implied pur- pose of a particular design document in order to fully under- stand its provisions. Of course the primary purpose of any earthquake code is to protect life. However, the way that this pporpose, as well as any additional purposes, is presented can provide ‘additional insight into the reasons for the presence of specific provisions in the body of the document and its in- tended audience. ‘The 1994 edition of NEHRP (1995) (hereafter referred to ‘as NEHRP-94) clearly states that it is intended as a reference document and not a model code. It states that the provisions present criteria for the design and construction of buildings and nonbuilding structures subject to earthquake ground mo- tions. The purpose is to minimize the hazard to life for all buildings and nonbuilding structures, to enhance the expected performance of higher occupancy structures as compared 10 {Se Pro), Engr, PI Process System lnc, Memphis, TW 38138, *Assoe, Prot, Dep. of Civ. Engrg. The Uni of Memphis, Memphis, ‘rN 38133 Note, Associate Editor Chit-Ming Uang. Discussion open until an- ‘uuy 1, 1998. To extend the closing date One mont, A writen Tequert ‘must Be fled withthe ASCE Manager of Journals The manuscript fr ‘his paper was submited for view and possible plication on June 2, 1996, This paper is put ofthe Journal of Structural Engineering. Vo. 123, No, 9, August, 1997. GASCE, ISSN 0733.94457/0008-0593— 100073400 + $50 per page. Paper No, 13520, strate the similarities and differences among these codes. normal occupancy structures, and to improve the expected pability of essential facilites to function during and following an earthquake. The purpose also states that because of the complexity and the great number of variables involved in seis- ‘mie design (e.g. asthe variability in ground motion, soil types, dynamic characteristics of the structure, material strength properties, quality assurance and control, and construction practices), the provisions provide minimum criteria that are considered prudent and economically practical forthe protec- tion of life in buildings subject to seismic ground motions anywhere in the United States. The provisions also clearly slate that the ground motions of the specified ""design earth- ‘quake"” may result in both structural and nonstructural dam- age. I states that for most structures designed and constructed according 10 these provisions, it is expected that structural 240 and Blige with oly type. 2 3 vertical iregulacties th are > S anes or>€5 8 Dynamic analysis require i zones 2,3, and ¢ |sec D and 3D response spectra analyse oroguivalet: Use| All buldingr > 20 Tor all ther baling with plan or vera regulars ‘Any other bung no qualifying fo ssc analy |sec'b and E—ste-specit reponse spesrum analy for bile. | Any tildng on Saal With 10.7 © ings in eis bard exponare group Io I with A, = 0 and ‘within 6 mi of faut wih capacity for > 7 magoitde earhyuake ‘or buldings with T = 09 tm tan of Ay > 02 and on type Se Scie (NEHRP-9) or sol pole ype E (NEHRP94) ‘erica face a [Force dsiibued ines proportion to son mata and tight for [Concentrated Force Fa top level where F = OUTIV = O28V and F tance (ELF) | T= 05 », praolicly for F> 25 ane witha near iterpolt-| =O when T= 07 thn of he exponent fom {02 for OS c= 28 Wik eemaining shear (V ~ F) dtbued tines i proportion to story mass and beg ed [rie bg established wing absurd methods. T= Gi whew T, [T= GIP o y= Rayegh mad o ny aber proven malas J Cin or 7, = ON, an in modal says compute V from |pCtTy = ONC SEAOC and by UBC Ty = 1 Sfeeaone 4 Ts = JEL ee scaling purpose using (Seon 24.8) T= 1 2G. (NEBRP- Jin r= 1 2car, (NEHRP-34) ‘nical scala | ovata preuessd componenta—check lad conBinaiony | ovina preston shal be deiged wing > O5DL for th Bava ton INEHRP-91"~(09-0SA)Ge * GRS)Ge where 29/5 = 1.0 and | Toad ating lone rf combat wth OSA, = 0 A= 008 INEHRP-54-—(03-08C)D + @RS)Qe where 28'S = 10 and orion srr cantilevers designed for net upward force of sc. = 0A. = 008 sam, Horizontal stra caneves— minimum ot upward force = 030) in aditon to athe lo combinations Provisions apply in 2086 3 and oly ‘Dynamic analyse [For enh of wo mutually perpeadiclar ares inlode all modes with [ Respsespecwum analysis recommended ‘tera "ra > 04 bu at eat thee lowest ode: For < te tres Ma ero mades = numberof ses Goa = 25418 except atepary D of buldings on. sl with “r. = 07's or baling oe 5, oS sol with To <0. (excep {6c fundaental then Cu'= ALLO + ST,VR of wheve Te > 40 Shen Co = SASS Nevins ana ASCE 7-98 ‘Same as above with C, and Cy substited for A, and Ay and site (ype Et Fsstiuted fr SS, sols Use sufcientaumber of modest include = 90% of the pariipa- ing ras oealcue the respons i each papal orzo [rime tistory analyse o ter aterative pocedues wing rion ‘sales especialy permiteg [Gncgoris Band C excep with ipo 5 plan kreplaiy: apply orcs aoncncurenly i each oto ong Sections [category € with type Span imeguaity ad ll aegis D&E sign fr dection tha causes ea effect ora leave. ap: Ply 100% in one decton and combine with 3D% aplied cone [Nenconcuen foreach pancipl mis excep for aoues 28 ea or ontogoal effec i ‘ype E pl egal exits ‘ype A plan egal ents about both major axes Cen fers pat of wo ele freeing sts And design fr 1005 in on dection in combiton with 30% in the ‘xbopoal Giecion spied caveat [Norton a sro levels ‘Gveruiing mo [Buicings = 10 wvies ro vodation a sory owl isings 2 20 stares —20% reduction at 20 soy from op and [Foe ane 3 an 4, can be taken as egal to eo fr cleution of “il levels Below: sighs ne interpolation from seo reduction at | ovetaning ome fou sl merase Tox sory ffm top 10205 a 200 wry fom tp Diatibuton to nooieal foreresiting vertical members specifcaly eat bigs rparsess of hig excpt invert pendulum 258 | "persed Teleco lowed a foundation interface Da [Diterence in clanic sry displacements, acuaed based on forces [Elasicdsplacomens decried without lover bound Wnt of 0078 ‘dtemine witout he upper bod init on T. multiplied BY. | for CIR and witout resuiction om Ty pation factor, Clit vce fom 0 10h, 10 098, e- |For P07 8 fending on seumic Maar exper poup rif = O04, = 0.005h, piferences exis ere Between NEHRE and ASCE) leer = 07 ‘bit < COR. < 0004, Praca alls pot vue wabliy cocci = 0.10 [Analysis aot rue if maby cocticione = O10 Stbltycoecent exam exceed (8) oF 025 atettion not eqied for znes 3 and 4 sory eit rio = (202 (998/ JOURNAL OF STRUCTURAL ENGINEERING / AUGUST 1997Some of the criteria for dynamic analyst was covered in the preceding discussion, The additional requirements are de- scribed in the row “Dynamic analysis criteria’ in Table 2. In the simplified modal analysis procedure of the NEHRP pro- visions specific guidance is given to the designer for imple- ‘menting the modal analysis. For each of two mutually perpen- dicular axes, the designer must include all modes with Ta, ‘greater than 0.4 but at least the three lowest modes. For buildings that have three stories or less the required number ‘of modes to be included equals the number of stories. The same spectral shape for the modal seismic design coefficient, ‘Com €xiss a5 for the ELF procedure. For SPC D or E buildings fon'S, soil (NEHRP-91) oF soil profile type E or F (NEHRP- 94) with Tx greater than or equal to 0.7 8 a site-specific spec- ‘rum must’be used, For buildings on S, or S, soil (NEHRP- 91) or soil profile type D, E, or F (NEHRP-94) all modes that have a period less than 0.3's (except the fundamental mode) have an alternative equation for calculating Cj. which will result in lower values, Also, for buildings where any modal period exceeds 4.0 5, Cin can be calculated using a different Alternative equation, which will result in lower values. Guid- ance for the more rigorous dynamic analysis is required when the modal analysis procedure isnot allowed. The commentary does provide a section on how to incorporate the applicable provisions; in general terms itis a sixstep procedure to per- form the recommended 3D response spectra analysis. Inclu- sion of sufficient modes to capture a minimum of 90% of the seismic reactive mass of the structure in each of two principal directions of response is specifically stated. “To meet the dynamic analysis requirements of UBC, a re sponse spectrum analysis is recommended as being sufficient except where the site-specific response spectrum analysis i required. General guidance as well as some specific require- iments are presented in the code. The designer is given several ‘options for determining the ground motion representation. A nommalized response spectrum is provided in the code as one of these options. A 3D analysis is required for structures with highly irregular plan configurations. Otherwise, no require- ment for 3D design is stated. For response spectrum analysis, the inclusion of sufficient modes such that atleast 90% of the participating mass of the structure is included inthe calculated response for each principal horizontal direction is required. ‘Time history analysis is permitted and must only conform 10 the general requirement that the rational analysis be based on established principles of mechanics. No other guidance for time history analysis is provided. ‘Both general approaches require scaling of the dynamic analysis results using the value obtained from the ELF pro- visions. If the base shear in any direction from the dynamic analysis is greater than the value calculated from the ELF pro- cedure the values from the dynamic analysis may be reduced to those values and the other results sealed accordingly. Ifthe base shear calculated from the dynamic analysis is less than the corresponding base shear from the ELF procedure, then the dynamic results must be scaled by the ratio of the ELF value to the corresponding dynamic base shear value. NEHRP ‘permits the base shear for this purpose to be calculated using 4 higher upper limit on T resulting in a possible lower value ‘of base shear for scaling purposes from the ELF method. UBC. ‘completely removes the upper limit on Ty for this purpose. ‘The comparison of T is shown in the row “Period” of Table 2 and was also presented during the discussion on lateral force "The results from each mode must be mathematically com- bined to obtain the total response. NEHRP requires that the ‘complete quadratic combination (CQC) method be used. UBC requires that resulis be combined by recognized methods, "A comparison of the vertical acceleration provisions is shown in the row ‘Vertical acceleration" of the Table 2. In the past vertical acceleration has not been generally considered ‘much of an issue since conventional structures tend to be very strong in the vertical direction due to gravity load design. However, the Northridge earthquake demonstrated that sign ‘ant damage can result from the vertical acceleration induced by thrust faults. Both documents address only special cases where vertical acceleration could be a significant problem. NEHRP requires that horizontal prestressed components be checked for a special load combination. Horizontal cantile- vered structural elements must be also designed for a mi ‘mum net upward force in addition to the other applicable load combinations. UBC requires horizontal prestressed compo- ‘ents to be checked for combinations using not more than SO% of the dead load for a gravity load acting alone or in com! nation with a seismic load. Horizontal cantilevered structural elements located in zones 3 or 4 must be designed for a spec- ified minimum net upward force. Past seismic codes have allowed some reduction in the over. turning moment. The justification of this practice was based fon the assumption that story level shears would not all reach ‘maximum value simultaneously and other factors. Both review ‘documents continue this practice to a limited degree. NEHRP. permits a 25% reduction atthe soil-foundation interface forall Buildings, except inverted, pendulum-type structures, regard- Jess of height. The overturning moment at the building levels ‘can be reduced if they have more than 10 stories. For buildings taller than 20 stories a 20% reduction in calculated overturning ‘moments at the 20th story from the top and all levels below ‘can be used. No reduction is permitted in the top 10 levels. A Tinear interpolation from 0% reduction at the 10th story to the full 20% reduction atthe 20th level is used for stories 11-19. NNo reduction in story level overturning moments is permitted for shears determined by the modal analysis method. However, 1 10% reduction is permitted at the soil-foundation interface. ‘The provisions do not appear to address this for other forms ‘of dynamic analysis. UBC does not permit any reduction in ‘overturning moment atthe structural levels. However, in zones 3 and 4 the force, if any, applied atthe top level of a regular building can be taken as zero for calculation ofthe overturning ‘moment at the soil-foundation interface. SEAOC permits re- distribution of overturning effects to other vertical members framing members of sufficient strength and stiffness are pro- vided to transfer the loads. UBC contains two additional load combinations, applicable only in zones 3 and 4, to be used for checking columns supporting a discontinuous, lateral, load- resisting element. SEAOC contains an identical provision ex- cept that its applicable for zone 2 in addition to zones 3 and 4 ‘The drift criteria for each document is given in the row “Drift” in Table 2. Drift provisions will control the member selection in many situations. The more flexible structural sys tems such as moment frames of steel and concrete will tend to have the most difficulty in conforming to stringent drift control requirements. NEHRP computes the drift as the dif- ference in elastic story displacements calculated using forces determined without the upper limit on T, multiplied by the deftection amplification factor C,. The amplification factor is determined by the type of structural system and varies from 1.25 to 6. Limits on daift in NEHRP Vary from 0.025 hy, to 0.010 hy depending on the seismic hazard exposure group. For some single story buildings, there is no limit on drift. NEHRP- 94 added more detailed information for masonry shear wall and masonry wal frame buildings and relaxed allowable story rift requirements for SPC group I. As previously stated ASCE increased the maximum allowable drift for some categories by ‘as much as 50% from the values shown in NEHRP. UBC uses clastic displacements determined without the lower bound ‘JOURNAL OF STRUCTURAL ENGINEERING / AUGUST 1997 / 990limit of 0.075 for C/R, and without the restriction on Ty, Drift limits in UBC are a function of the period, T, and R, as shown, Pedelta_ analysis requirements are shown in the row ""P- delta" of Table 2. Many of the most popular computer analysis programs used by engineers today have P-delta capability. This Capability eliminates the computational effort required and ‘makes simplified formulas unnecessary. Both code approaches use the stability coefficient as an indicator of the need to in- clude P-delta effects in the analysis. Ifthe stability coefficient exceeds 0:10 then the effects of P-delta must be included. NEHRP establishes an upper limit for the stability coeficient of 0.5/(C,) but less than or equal to 0.25, where Cy is the deflection amplification and beta is the ratio of shear demand to shear capacity for the story. Beta can be conservatively taken to be equal to 1.0, UBC does not require that P-delia effects be included ifthe stability coefficient for all stores is Tess than or equal to 0.10. In addition P-delta does not have to be considered in zones 3 and 4 if the story drift ratio does not exceed 0.02/R,. SEAOC does not have the lower limit for the stability coefficient included in the provisions but it is 1o- cated in the commentary ‘SUMMARY ‘The purpose of this paper was to make comparisons be- tween four of the major seismic design codes used in the United States. This comparison shows that the four codes can be easily categorized by the design approach adopted, NEHRP-91 and NEHRP.94 and ASCE 7-95 are based on the strength design approach. UBC and SEAOC use a working stress design approach, There are other fundamental differ- ‘ences such as the required methods of analysis, amplified dis- placements versus elastic displacement, and the method of as- Signing an importance factor. There appears to be a developing consensus thatthe material design and detail requirements can- not be easily separated from the load provisions ifthe assumed limit state occurs beyond elastic yield. As confirmation of this ‘one should note that significant detailing requirements are in- corporated into each of these documents. 1000 / JOURNAL OF STRUCTURAL ENGINEERING / AUGUST 1997 It is also guaranteed that these codes and others that model them will continue to undergo significant changes. Each new ‘major earthquake provides more data to be analyzed and new problems to be solved. It is important that the engineers, buld- ing officials, and others involved with using and administering the earthquake codes kesp up with the changing tide and be prepared for the inevitable changes ahead. APPENDIX. REFERENCES ‘Algerissn, 5. T, and Perkins, D. M. (1976). “A probabilistic estimate ‘f maximim acceleration in rock inthe contiguous United States” USGS Open Fie Rep. 76-416. US, Geoogial Survey (USGS), Reson using code rquiements for masonry sructues." ASCE 5.92/ACT $530.92 (19928, ASCE, New York, NY. “Bulling Code requirements for reinforced concrete” Concrete Inst (ACD, Detroit Mich Load and resistance facor design specication for structural sel bull ‘ngs (1993)Am. Inst of Sec! Cons. (AISC), Inc, Chicago, 1 Late. W. (0989). “Comparison among earthquake codes." Earthquake ‘Spectra, 5). “Minimum: design tude fo buildings and other structures.” (1998). "ASCE 795, ASCE, New York, NX LNEHIRP recommended provisions forthe development of seismic regu lato for new buildings, 1991 4. (192). Bul, Seismic Safety Counc, Washington, De: LNEHRP reconmended provisions forthe development of seismic rege ‘ations for new builings, 1986 E8.(1998) Bus. Seambe Salty Count Washington, DE SEAOC recommended lateral force requirements and commentary: (1990). Seamotogy Comminee, Swust Engr. Asse, of California {SEAOC), Sueno, Cal Slane provisons for sacar! ste buildings, (1952). Am. Ins. of Stel Const. (A130), ne, Chica. Spectcation for sacral buldng,sllowable stress design and plastic design (1989). Arm, Inst of Stel Cons. (AIS), Tne Chicag, I -Tenatve provisions forte develope of sts regulates for buildings" (1978) A7C-3.06 Rep, Appl. Techno. Coun, Redwood City, Cait Unitrin bulding code, Votue 2, structural engineering design prov- ‘ons, (1998), Ineratonal Coseence of Bung Oca, WHS, eat (419920). Am.