Correlation between Cone Penetration Test parameters, soil type, and soil liquidity index using long short-term memory neural network

Jocz, Mateusz; Lefik, Marek

doi:10.2478/sgem-2023-0023

Figures & Tables

An example of a research point consisting of: a) a cross section of the borehole, where Sa – sand, Si – silt, Cl – clay, Mg – made ground, Gr – gravel; b–d) graphs of the basic parameters of CPTU probing, such as qc, fs, and u2; e) a graph of the calculated laboratory values of the liquidity index IL for the selected layers.

Liquidity index IL before (green line) and after (orange line) transformation for the example research point.

Results of identification of liquidity index with the developed LSTM network in comparison with known values of the liquidity index. Explanation in the text.

Figure 5:

Comparison of the obtained results with the correlation proposed by PN-B-04452:2002 and with laboratory results transformed by Equation 3 for profile number 7.

Figure 6:

Results of identification of soil type with the developed LSTM network (predicted data) in comparison with known borehole profiles (actual data). The profile includes the following data numbers: profile 1 (0–540), profile 2 (540–2115), profile 3 (2115–2837), profile 4 (2837–3547), profile 5 (3547–4600), profile 6 (4600–5318), profile 7 (5318–6304), profile 8 (6304–8060).

Division of soil types into categories_

Soil types	Symbol	Category
Made ground	Mg	1
Fine sand	FSa	2
Medium sand	MSa	3
Glacial till	siSa, clSa, sasiCl	4
Settled deposits	clSi, saSi, Si,	5
Settled deposits with gravel	grsaSi, grclSi	6
Gravel	Gr	7

Correlation between Cone Penetration Test parameters, soil type, and soil liquidity index using long short-term memory neural network

Figures & Tables

Figure 1:

Figure 2:

Figure 3:

Figure 4:

Figure 5:

Figure 5:

Figure 6:

Division of soil types into categories_

Paradigm

My account