Performance of Glass Fiber-Reinforced Polymer Reinforcing Bars in Tropical Environments-Part I: Structural Scale Tests. Paper by Abhijit Mukherjee and S. J. Arwikar

There are a number of reasons that make the experimental results presented in this paper invalid for any meaningful analysis. A list of the major ones follows:

1. All specimens are shear deficient and shear cracking is the originator of failure. According to ACI 440.1R-03, the shear capacity of this specimen is approximately 6.7 kN. Even though this value is intended for design and therefore is conservative, a safe shear arm (based on the nominal flexural strength) should have been no less that 1000 mm;

2. Some specimens were tested at a shear arm over reinforcement depth ratio of 1.37. The arching effect provides a very significant contribution to capacity in this instance;

3. The paper states that the ultimate capacity of the specimens tested as beams did increase as the strength of the concrete did due to curing. The authors fail to point out that this is because, for all specimens, shear strength controls failure. Flexural capacity cannot be affected by concrete strength gain when the failure, as for these specimens, is controlled by FRP rupture (in fact, ρ is approximately 1/3 of ρ^sub b^); and

4. The development length was less than or equal to 50 mm. For a No. 3 bar as used in this research, the minimum would be 300 mm (and this value may not be that conservative as the revision of ACI 440.1R-03 indicates.) The resulting bar slip, as documented in the paper for several specimens, makes computation of internal forces impossible because of the nonlinearity of strain at any given cross section. The assumption used in the reported analysis is incorrect.

A question for the authors: How can the authors say that specimens fail in compression if the reinforcement ratio is significantly below the balanced condition? (Is this because of arching? Is this because of bar slip? Is this because the failure mode was misinterpreted?

In conclusion, the test results cannot be used due to the poor design of the test specimens and the setup. The analysis is off mark because the most basic assumption of linearity of strain is invalid.

Discussion by A. Nanni

Jones Professor of Civil Engineering, University of Missouri-Rolla, Rolla, Mo.

Copyright American Concrete Institute Jul/Aug 2006
Provided by ProQuest Information and Learning Company. All rights Reserved