18.11.2012 Views

Cranfield University

Cranfield University

Cranfield University

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Soil Compaction Models<br />

VCL parameters. The predicted soil displacement was highly accurate. The models of<br />

Gupta and Larson (1982) and Bailey and Johnson (1989) which require more input pa-<br />

rameters can hardly be more accurate, however, the larger amount of input parameters<br />

these need make them unsuitable for such an in-situ approach. Chi et al. (1993) point out<br />

that the derivation of critical state soil mechanics parameters is the largest source of error<br />

when comparing soil compaction model predictions to real data. This source was reduced<br />

here due to the small number of necessary parameters and the large sample size.<br />

It was possible to use a soil mechanics model based on critical state soil mechanics theory<br />

(Schofield and Wroth, 1968) and gain its soil specific soil physical parameters from simple<br />

empirical measurements. This allowed qualitative but more importantly quantitative cor-<br />

rect predictions.<br />

The output from COMPSOIL is more distinct than for models such as TASC (Diserens<br />

and Spiess, 2004) or SOCOMO (van den Akker, 2004) which only predict soil stresses.<br />

Although the screw driver test for TASC is easier to perform than gaining an in-situ VCL,<br />

the additional information gained from COMPSOIL due to the accurate soil displacement<br />

and soil density increase prediction with depth make the additional effort worthwhile. Thus<br />

the in-situ VCL approach improves considerably the user-friendliness of COMPSOIL due<br />

to the ability to adapt easily to different soil types.<br />

Considering the different approaches taken to derive VCL parameters, the approach de-<br />

scribed here to gain an in-situ VCL has the advantage of easily accessible information and<br />

very large sample size compared to any laboratory based approach. The large sample size<br />

and in-field test avoids the major issue of variations with small sample testing for critical<br />

state soil mechanics models as pointed out by Chi et al. (1993) when investigating the ac-<br />

curacy of soil compaction models. The sample size is of concern for the approach taken by<br />

Gurtug and Sridgaran (2002) who obtained compaction characteristics from plastic limit<br />

tests for fine grained soils as these are usually carried out with only a few grams of soil.<br />

Compared to the improvement of Kirby et al. (1998) who derive all critical state soil me-<br />

chanics parameters from one constant cell volume triaxial cell test, the field test only pro-<br />

vides the information for soil compaction, but without any laboratory equipement. No pedo<br />

- transfer functions as for example included by Keller et al. (2007) in SOILFLEX are nec-<br />

essary as the approach can be repeated on any soil type and the resulting in-situ VCL will<br />

Ph.D. Thesis Dirk Ansorge (2007)<br />

119

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!