Tesi etd-01252024-115627
Link copiato negli appunti
Tipo di tesi
Dottorato
Autore
PIRRONE, SERENA ROSA MARIA
URN
etd-01252024-115627
Titolo
Investigating the Optimal Design of Bioinspired Digging Robots for Earth and Space Soil Exploration
Settore scientifico disciplinare
ING-IND/15
Corso di studi
Istituto di Biorobotica - PHD IN BIOROBOTICA
Commissione
relatore Prof. CIANCHETTI, MATTEO
Membro Prof. SIBILLE, LUC
Presidente Dott.ssa MAZZOLAI, BARBARA
Membro Prof. SIBILLE, LUC
Presidente Dott.ssa MAZZOLAI, BARBARA
Parole chiave
- Bioinspired artificial exploratory systems
- Soil exploration
- Penetration performances
- Discrete Element Model
- Robot-soil interaction
- Robot design requirements
Data inizio appello
28/06/2024;
Disponibilità
completa
Riassunto analitico
This Ph.D. dissertation presents a methodological approach to investigate design requirements for autonomous soil diggers, taking inspiration from plant root features. Plant roots are efficient soil explorers that move by growing at their apical region and maintaining their mature, already grown part, static. This movement strategy has been proven to reduce lateral friction, thus facilitating soil intrusion, confining interaction with the environment at the apex, and enabling additional apical movements and morphing in response to mechanical constraints. Examples of such apical morphing include apex tapering and radial expansion.
A three-dimensional (3-D) Discrete Element Model (DEM) was developed to mimic selected features of plant roots (specifically apical growth, body radial expansion, and apex geometry), verify their usefulness in soil penetration operations, and identify conditions for their advantageous implementation in underground exploratory robots.
To this purpose, different combinations of intruder and soil characteristics were analyzed numerically and experimentally, employing a terrestrial soil model, the Hostun sand, and three extraterrestrial soil simulants. Performances were evaluated in terms of soil resistance pressure experienced by a probe implementing the selected root-inspired feature and compared to a non-bioinspired one.
Specifically, the numerical model has been employed to estimate the penetration pressure requirements of two probes differing in tip shape, one with a conic and another with a root-like tip, and penetrating by being pushed from the top. Modeling results were compared to outcomes obtained experimentally by performing penetration tests in Hostun sand samples. Both numerically and experimentally, the tip design effect was investigated in soil samples differing in initial density (i.e., loose, medium-dense, and dense media). At higher soil densities, modeling pressures resulted underestimated respect to the experimental measurements due to the bottom border effect of the soil chamber present in the experiments. At low soil density, bottom border effect was not observed experimentally and modeling and experimental outcomes resulted closer. In agreement with the experimental findings, the developed numerical model correctly approximated the behavior of probes during penetration in Hostun sand with changing initial porosity and correctly estimated the root-like tip design as advantageous with respect to the conic tip design in terms of penetration pressure requirements. The tip shape was also experimentally investigated in Lunar and Martian regolith simulants. As for the Hostun sand, the root-like tip produced lower penetration pressures, demonstrating its employment as advantageous in extra-terrestrial soils.
Furthermore, the developed 3-D DEM numerical model was adopted to simulate the penetration process of a digger using apical growth in both cohesionless granular and cemented deep soils. In the former case, dense and loose granular media were considered. In particular, the model was adopted to compare penetration performances (in terms of penetration pressure and energy requirements) with purely axial growth and a combination of radial and axial growths. Radial growth is hypothesized to facilitate root soil penetration. Results from our model suggest that implementing a radial growth preliminary to an axial growth is advantageous in cohesionless dense granular soil, reducing the soil resistance experienced by the digger for deeper penetration after radial expansion. For penetration in cemented soil, the radial expansion results advantageous over a lower penetration depth, and its beneficial effect drops with increasing inter-particle contact adhesion values.
Finally, the digger penetration process was analyzed for cohesionless granular soil as a function of the digger diameter (Droot) to soil median particle size (D50) ratio. In particular, the soil resistance force that penetration systems must overcome to move into a cohesionless granular soil was estimated based on the Droot/D50 ratio, and important differences were found for small and big ratios (Droot/D50<<1 and Droot/D50>>1). In addition, the penetration pressure requirements of a root-inspired digger penetrating through apical growth were compared to those of a traditional penetrometer that is pushed from the top, considering both small and big Droot/D50. The analysis confirmed the advantage of employing apical growth in all cases, showcasing a lower pressure while penetrating soil, with higher relevance for low Droot/D50.
This thesis's findings provide important guidelines for the design of artificial soil penetrometers in terms of size, geometry, penetration strategy, pressure, and energy requirements. The developed numerical model can thus be employed to inform autonomous explorative robots and suggest conditions to benefit the performance of autonomous exploration on Earth or in Space applicative scenarios.
A three-dimensional (3-D) Discrete Element Model (DEM) was developed to mimic selected features of plant roots (specifically apical growth, body radial expansion, and apex geometry), verify their usefulness in soil penetration operations, and identify conditions for their advantageous implementation in underground exploratory robots.
To this purpose, different combinations of intruder and soil characteristics were analyzed numerically and experimentally, employing a terrestrial soil model, the Hostun sand, and three extraterrestrial soil simulants. Performances were evaluated in terms of soil resistance pressure experienced by a probe implementing the selected root-inspired feature and compared to a non-bioinspired one.
Specifically, the numerical model has been employed to estimate the penetration pressure requirements of two probes differing in tip shape, one with a conic and another with a root-like tip, and penetrating by being pushed from the top. Modeling results were compared to outcomes obtained experimentally by performing penetration tests in Hostun sand samples. Both numerically and experimentally, the tip design effect was investigated in soil samples differing in initial density (i.e., loose, medium-dense, and dense media). At higher soil densities, modeling pressures resulted underestimated respect to the experimental measurements due to the bottom border effect of the soil chamber present in the experiments. At low soil density, bottom border effect was not observed experimentally and modeling and experimental outcomes resulted closer. In agreement with the experimental findings, the developed numerical model correctly approximated the behavior of probes during penetration in Hostun sand with changing initial porosity and correctly estimated the root-like tip design as advantageous with respect to the conic tip design in terms of penetration pressure requirements. The tip shape was also experimentally investigated in Lunar and Martian regolith simulants. As for the Hostun sand, the root-like tip produced lower penetration pressures, demonstrating its employment as advantageous in extra-terrestrial soils.
Furthermore, the developed 3-D DEM numerical model was adopted to simulate the penetration process of a digger using apical growth in both cohesionless granular and cemented deep soils. In the former case, dense and loose granular media were considered. In particular, the model was adopted to compare penetration performances (in terms of penetration pressure and energy requirements) with purely axial growth and a combination of radial and axial growths. Radial growth is hypothesized to facilitate root soil penetration. Results from our model suggest that implementing a radial growth preliminary to an axial growth is advantageous in cohesionless dense granular soil, reducing the soil resistance experienced by the digger for deeper penetration after radial expansion. For penetration in cemented soil, the radial expansion results advantageous over a lower penetration depth, and its beneficial effect drops with increasing inter-particle contact adhesion values.
Finally, the digger penetration process was analyzed for cohesionless granular soil as a function of the digger diameter (Droot) to soil median particle size (D50) ratio. In particular, the soil resistance force that penetration systems must overcome to move into a cohesionless granular soil was estimated based on the Droot/D50 ratio, and important differences were found for small and big ratios (Droot/D50<<1 and Droot/D50>>1). In addition, the penetration pressure requirements of a root-inspired digger penetrating through apical growth were compared to those of a traditional penetrometer that is pushed from the top, considering both small and big Droot/D50. The analysis confirmed the advantage of employing apical growth in all cases, showcasing a lower pressure while penetrating soil, with higher relevance for low Droot/D50.
This thesis's findings provide important guidelines for the design of artificial soil penetrometers in terms of size, geometry, penetration strategy, pressure, and energy requirements. The developed numerical model can thus be employed to inform autonomous explorative robots and suggest conditions to benefit the performance of autonomous exploration on Earth or in Space applicative scenarios.
File
| Nome file | Dimensione |
|---|---|
| PhD_Thes...rrone.pdf | 6.76 Mb |
Contatta l'autore |
|