Laserfiche WebLink
<br />c.Direct Comuutation: <br /> <br />Area of rectangle ABDC <br /> <br />height (h) x base (b) <br />- 1000 x 8 - 8,000 <br />Area of triangle CDE - 1/2 (hb) <br /> <br />Total area - 8000 + 16000 <br /> <br />1/2 (4000 x 8) - 16,000 <br />24.000 <br /> <br />The three methods yield identical results for the straight-line <br /> <br />situation. However, for the typical non-linear situation, the <br /> <br />closeness of results will depend on the number of input points and, <br /> <br />therefore, on the number of rectangles defined by these points. <br /> <br />The second concern, regarding distortions from use of the <br /> <br />frequency interval calculation method, can be demonstrated from an <br /> <br />inspection of Figure V-ll. Curve 4a duplicates the straight-line <br /> <br />situation such that by inspection, the area excluded from rectangle <br /> <br />ABCD, under the curve, is equal to the area included, above the <br /> <br />curve. This, of course, is consistent with previous findings. Curve <br /> <br /> <br />4b is convex, and is more typical of the shape of damage-frequency <br /> <br /> <br />curves encountered. Rectangle ABCD is fitted to the last two points. <br /> <br /> <br />By inspection, it is observed that the area included in the <br /> <br />rectangle, above the curve, is significantly larger than the area <br /> <br />excluded, under the curve, such that the estimate for this part of <br /> <br />the curve appears to be overstated. It appears, then, that the <br /> <br />accuracy of the estimate is increasingly compromised the more convex <br /> <br />the curve becomes. It can, therefore, be concluded that, typically, <br /> <br />V-58 <br />