Comments on Accuracy of Settlement Computations.

Accuracy of Predictions of Total Settlement One-dimensional settlement analyses are used in estimating total settlements, and the rates at which they occur, under the central parts of loads that are wide compared to the depth of the compressible layer. For soft, normally consolidated clays, the accuracy of the prediction of total settlement is mainly limited by the natural variability in soil properties coupled with a limited number of tests, as dictated by economic constraints. The accuracy of prediction is adequate for real design work.
For stiff, overconsolidated clays, the accuracy of the prediction of total settlement seems to be affected by sampling disturbance, which increases the slopes of reload curves in the laboratory. Although the predicted settlements are likely to be too large, the total settlements are small enough that they are often not a serious problem. The only serious problems are with lightly overconsolidated clays, where the engineer may be uncertain whether the loading will be entirely on a reloading curve or partially on a virgin curve.

Prediction of Time Rate of Settlement During Secondary Compression Predictions of time rates of settlement are relatively inaccurate in most cases. In the special case of saturated, homogeneous, normally consolidated clays, the prediction of the rate of primary consolidation is based on Terzaghi’s theory and appears to be reasonably accurate. Further, field studies suggest that the slope of the secondary settlement curve (S-log t) in the field is about the same as in the laboratory, so the secondary settlement in the field
is estimated just by plotting the primary field S-log t curve and then drawing in the secondary curve tangent to the laboratory curve, as suggested in Figure 3.29.
In most field problems, the simplifying assumptions of Terzaghi’s theory are not satisfied. Estimates of the times required for consolidation in such cases are usually based on previous experience in the area, although more advanced analytical techniques than those considered here have shown much promise.

0 comments:

Post a Comment