that was restrained by the grid. A total of 18 micro columns were restrained by the grid.

Therefore the total load used to calculate strain was determined based on the following:

 Pt,= P, +ncPR (8-16)

 Using the total load from Eq. 8-16 and substituting into Eqs 3-8 or 3-10 the new

strain was calculated. The lateral displacement of the micro columns controlled the

failure of the CFRP grid ring. It was discovered that based on the length and the strain of

the micro column adjacent to the CFRP grid, the CFRP grid was reaching failure strains

as soon as the micro column buckled. This was not observed during the tests and two

possibilities exist: (1) the CFRP grid layers were not bonded together and it is possible

that once the outside concrete was damaged there was some slippage developed since the

bond between the grid and the concrete weakened, and (b) as it is often the case with

actual structures it usually takes some initial movement before all components stiffen and

start working together. Therefore, an initial spacing between the micro column and the

CFRP grid of approximately 0.019 in. was assumed. This initial spacing was selected

based on the experimental data from grid confined cylinders presented in chapter 3.

Therefore, the selection of the initial spacing is tuned to that specific group of data.

 The initial spacing prevented the CFRP grid from engaging the buckled micro

columns immediately. The lateral displacement of the CFRP grid was then equal to the

total lateral displacement reduced by the assumed initial spacing. The lateral

displacement was compared to the failure lateral displacement (a6fl) of the CFRP grid

ring that was determined as follows:


 =- 1 (1 +f). d (8-17)
 2 IT