**5. Comparison of experimental and numerical simulation tests**

All DEM numerical simulation calculations ran on cluster due to faster calculation process. In this case, it is possible to separate modelled oedometer into four quarters where each quarter is calculated by separate computer (**Figure 11**). In DEM simulations, parallel calculations of clusters are widely applied. Using these clusters, it is possible to calculate much bigger simulations in the same period of time, as with a single computer. Nevertheless, applications of clusters in DEM simulations do not provide enough calculation capacity. More detailed

Used modelled particles shapes in numerical simulations have three shapes: ideal sphere, particle recreated from two spheres and particle recreated from three spheres (**Figure 12**). Recreation of particles shapes is based on experimental testing results obtained from mor‐

Providing DEM numerical simulations with different recreated particles shapes, it is possible to obtain particle shape influence for compression results. The analysis of influence of

**Figure 13.** Comparison of compression results with different particle shape: 1—particle recreated from sphere; 2—par‐

ticle recreated from two spheres; 3—particle recreated from three different sizes spheres.

simulated particles shape on compression results is presented in **Figure 13**.

explanation of calculations with DEM is presented in Refs. [1, 30].

phology parameters determination part.

256 Modeling and Simulation in Engineering Sciences

**Figure 12.** Recreation of particles shapes.

Reliability and accuracy of numerical DEM modelling of sand soil behaviour depend on the modelled soil and discretization level of particles. In this research work, the same numerical granulometric curve as obtained experimentally was used (**Figure 14**).

**Figure 14.** Numerical and experimental granulometric curves.

The analysis of obtained experimental and numerical morphology parameters is given in **Figure 15**. Comparison of results revealed that to recreate morphology parameters, involve‐ ment of different quantity of spheres for single‐particle subscription is necessary: 8–11 spheres for form coefficient, 7–11 spheres for sphericity, 8–12 spheres for circularity, 3–4 spheres for angularity, respectively. Analysing angularity recreation with different quantity of spheres, the angularity coefficient decreases when more than seven spheres are used for single‐particle subscription.

**Figure 15.** Comparison of morphology parameters: 1—experimental data; 2–12—spheres quantity of recreated numeri‐ cal particle shapes.

Validation of experimental and numerical (*Ep* = 78 GPa and *ρs* = 2,650,000 kg/m<sup>3</sup> ) compression results is presented in **Figure 16**. One can note that used high modelled particles density increases calculation speed and does not have any effect for contact mechanics due to the first Newton law of motion.

**Figure 16.** Comparison of compression results: blue—standard experiment; red—numerical experiment (one sphere); green—numerical experiment (two spheres); purple—numerical experiment (three spheres); black—improved experi‐ ment.

The analysis of compression results revealed (**Figure 16**) that the increased spheres quantity for single‐particle discretization lead the numerical compression curve to the experimental one. For a more qualitative match of experimental and numerical curves, it is better to use the improved experimental compression curve which is closer to the numerical results. In this case, the curves of numerical and improved experimental results get closer to each other. Usually, only numerical tests are corrected with non‐normal coefficients to get the same result curve as in the experiment. In this case, it was improved experimental test to get experimental compression curve more closely to numerical one. This effect was obtained realising the same compression procedure in experiments and in numerical simulations.
