Sheer Strength of Web-Tapered I-Shape Members

The scope of this research is the development of accurate methods for predicting the shear strength of web-tapered, I-shaped members such as those commonly used as rigid frame members in metal building systems. Historically, it has been common practice among metal building manufacturers to evaluate shear by comparing the allowable shear stress to the shear stress caused by a "modified shear," which accounts for the shear components of the inclined flange forces. This method has been recommended by Blodgett (1966), among others. The most modern guidance for the design of tapered I -shaped members is the upcoming MBMAIAISC design guide entitled Frame Design Using Web-Tapered Members (Kaehler et al. 2007). In Section 5.6, the authors state that the modified shear approach "has not been included below due to the lack ofresearch validation ofthe procedure." No research study has provided definitive evidence for or against the use of the modified shear approach. However, studies on tapered member strength and panel zone strength were performed at Virginia Tech by Sumner (1995) and Redmond (2007) under the supervision of Dr. Thomas M. Murray. Both studies included comparisons of measured and predicted tapered member shear. The predicted strengths were based on measured material properties and dimensions. Sumner's specimens all failed in flexure, so the actual shear strengths are unknown. However, his comparisons of maximum measured shear to predicted shear strength indicated that numerous members achieved much higher shears than the AISC Specification (AISC 1994) (same shear strength equations as in the current Specification (AISC 2005)) nominal strength equations predicted. Several of Redmond's specimens failed in combined shear and flexure. In each ofthese cases, the specimen reached a higher load than that predicted by the Chapter G (AISC 2005) nominal strength equations without consideration of modified shear. Half of the failure loads were under-predicted by 34% or more. Redmond also made comparisons of measured and predicted failure load using the modified shear procedure. If specimens with large initial web out-of-planeness were excluded, then strengths were underpredicted by an average of 4% (COV= 4%), indicating that the modified shear approach seems to be justifiable. (If two specimens with initial out-of-planeness near or exceeding the MBMA limits are included, the strength was over-predicted by 8% on the average (COV = 19%).) It is also noteworthy that his specimens were subject to combined shear and moment-predicted strengths were for shear only-and a complicated stress distribution in the knee area. The modified shear procedure would likely have been validated by a wider margin ifthe specimens had typical web initial out-of-planeness and ifthe experiments isolated the effects of shear and combined shear and moment.