Published in SEAOSC Slender Wall Task Group, January 1, 2006, pages 1-47.
NOTE: At the time of publication, the author John Lawson was not yet affiliated with Cal Poly.
Recognizing there have been questions on the differences between the alternate slender wall design procedures in 1997 UBC and in ACI 318-02, the SEAOSC Board authorized a Task Group to provide a comprehensive review of the two design procedures. The ACI procedure was adopted by IBC 2000 and subsequent code editions. As quoted in ACI 318R-02 Commentary Section R14.8, Section 14.8 is based on the corresponding requirements in the UBC and experimental research of the Test Report by SCCACI-SEAOSC.
This summary report includes review of source documents, code comparison, and background of the design provisions under UBC and under ACI, respectively. A comprehensive review of the 1980 test data was made in addition to analytical comparison of sample wall panel design under each of the two procedures. Pursuant to the comparative design and validation of the original data, a list of findings is presented in the Report. Other design considerations though not part of the code comparison are discussed in order to encourage further studies by other groups. The report concludes with recommendations to SEAOSC Board and proposed changes to ACI.
Under 97 UBC Section 1914.8, the cracked moment is based on fr = 5 √ f ´c.; and in ACI 318-02 Section 14.8, the cracked moment is based on fr = 7.5 √ f ´c. This also means that the Mcr (UBC) = 2/3 Mcr (ACI) in the application of the two design procedures. In the 97 UBC, a linear interpolation between Δcr and Δn is permitted in obtaining Δs in order to simplify the slender wall panel design for Ms > 5 √ f ´c Ig/yt. The ACI procedure employs effective moment of inertia and a magnified moment for the combined moment due to lateral and eccentric vertical load, also know as the P-Δ effect. Table 1 gives section by section comparison between the alternate slender wall design procedures.
Review of 1980 Test Data
This Task Group was able to review and re-analyze the original test data. Verification of the 1980 data using adjusted lateral force and deflection data was performed. The analytical result follows closely with the bilinear load deflection characteristic. Lateral deflection increases rapidly when the moment exceeds two-third (2/3) of Mcr (as defined by ACI). The calculated moments for each of the twelve test panel correlate closely with the empirical test data. The load deflection curves and plots for the low axial loads versus moment interaction curve further validate the UBC design procedure. ACI needs to improve its methodology in computing Mu and Ie so that computed results would follow a bilinear load deflection characteristic.
Summary of Findings
Summary of comparative design examples is given on Table 5. Design based on ACI procedure is normally controlled by strength with service load deflection less than Δcr. ACI procedure significantly under-estimates service load deflection in comparison to the UBC procedure with increase lateral force and/ or axial load. Where wall panel design based on ACI procedures meets strength and deflection limit, the corresponding wall panel calculation based on UBC procedure may exceed the deflection limit.
- To calculate service load deflection, use E/1.4 for earthquake forces
- Recommend to appropriate enforcement agencies that adoption of the 2003 IBC provisions on alternate design of slender wall procedure should incorporate proposed changes to ACI 318-05 Section 14.8.4.
- Modification to ACI 318-05 Section 14.8.4 -delete equations (14-8) and (14-9) and the last paragraph in total, and replace with the following after the first paragraph:
“Δs = 0.67Δ cr + (Ms –0.67Mcr )(Δn –0.67Δcr)÷ (Mn-0.67Mcr); for Ms > 0.67Mcr (14-8)
Δs = 5 Ms lc2 ÷ (48Ec Ig) ; for Ms<.67Mcr (14-9)
- Send a letter to ACI-318 addressing the concerns in using the ACI alternate design of slender wall procedure and requesting ACI 318 to correct statements under Commentary R14.8.