Effect of boundary condition	notation
Fig.1 Size of container and side wall ( 1) after Bolton and Gui, 1993, 4) after Renzi et. al, 1994)	Fig.2 Base of container (after Bolton and Gui, 1993)

Penetration tests carried out in relatively small soil container and near the side wall of container, tip resistance may be affected by the boundary restriction. To eliminate this boundary effect, ratios of D/B or L/B should be large enough, where L, D and B are distance from side wall of container to the cone, diameter of the soil container and base of the cone, respectively. Figure 1 shows results of centrifuge tests aiming at assessing the boundary effect. Tip resistance at a certain depth are plotted against the ratios. Open circles in the figure correspond to tests in which the cone was penetrated in the center of the model, while solid closed circles correspond to tests performed near the side wall of container. The ratios of L/B and D/B higher than 20 is recommended for sand, however, this threshold value may be influenced by compressibility - density of soil, stress levels, loading rate etc.

There may also be a threshold distance (x) between the cone tip and the base of the container, below which the tip resistance is affected by the base of the container. Ratio of x to the cone diameter (x/B) obtained from centrifuge tests on sand (Bolton and Gui, 1993) is plotted against relative density in Fig.2. The plot indicates that distance from the cone shoulder to the base of container should be 5B to 10B to eliminate the base boundary effect. Test results must be carefully interpreted when the x/B ratio is small.