Well, I noticed a couple of things. First, you need to strip off the last part of your FFT array before displaying it to the graph, because you can only effectively see frequencies to 1/2 your scan rate, or 11025 Hz. Second, you used the wrong VI to get the real portion of your FFT--you chose the "complex to polar", when you really wanted "complex to Re/Im". Try that, and your results will be different. Also, for future reference, you may want to "window" your time array with a Hanning or Hamming window to smooth transition points. LabVIEW has a whole article devoted to these concepts in the help menu.
Mark