Figure 13-12.—Adjusting the zero point of the curve.indicates that the proving-ring dial used for this testnecessary to obtain the true or corrected load. In thiscontained a previously determined index error of 0.003inch. An error of this kind develops as a result ofrepeated compressions of the proving ring duringtesting.The figures under total load are the results obtainedby multiplying the corrected dial reading by theproving-ring constant. Each unit load was obtained bydividing the total load by 3.Figure 13-11B is the reverse side of the CBR testdata sheet. This side of the data sheet is used to plot theunit loads against the penetration depths to determine ifa correction to the unit load is necessary.In figure 13-11B you see the curve for sampleSF-PI-4 that we have been discussing. The convex shapeof this curve indicates that no corrections werenecessary. Sometimes, however, surface irregularities inthe soil sample or disturbances during the test will resultin curves having an initially concave shape, such asshown in figure 13-11B for samples SF-PI-5 andSF-PI-6. This shape indicates that a correction iscase, you must first adjust the zero point of the curve.Figure 13-12 illustrates the procedure you should use toadjust the zero point. First, draw a line that is tangent tothe steepest portion of the concave curve and extend thetangent to the zero base line. The point of intersectionof the tangent and the base line is the newzero-penetration point. The distance of the new zeropoint from the original zero point (distance X) is thedistance that the 0.1 and 0.2 points are moved to the rightto establish the corrected unit load.Finally, let’s look again at figure13-11A and discusshow the CBR values are determined. To compute theCBR values, you divide the unit load (or corrected unitload) at 0.1 and 0.2 inch by the standard unit loads of1,000 and 1,500 psi, respectively. Each result is thenmultiplied by 100 to obtain the CBR in percent. TheCBR is usually selected at 0.1 inch; however, when theCBR at 0.2 inch is greater, you should first rerun the testand then if the check tests give similar results at 0.2-inchpenetration, the CBR at 0.2 inch should be used.13-15