Penetration Tests in a Mold on Regolith Quasi-Analogues at Different Relative Densities C. Vrettos 1, A. Becker 2, K. Merz 3, and L. Witte 4 1 Division of Soil Mechanics and Foundation Engineering, Technical University Kaiserslautern, 67663 Kaiserslautern, Germany; email: vrettos@rhrk.uni-kl.de 2 Division of Soil Mechanics and Foundation Engineering, Technical University Kaiserslautern, 67663 Kaiserslautern, Germany; email: becker@rhrk.uni-kl.de 3 Division of Soil Mechanics and Foundation Engineering, Technical University Kaiserslautern, 67663 Kaiserslautern, Germany; email: kaimerz@rhrk.uni-kl.de 4 German Aerospace Center (DLR), Institute of Space Systems, 28359 Bremen, Germany; email: lars.witte@dlr.de ABSTRACT Know-how from terrestrial geotechnical engineering may be applied for a variety of space problems such as the prediction of the soil mechanical response of extraterrestrial soils, the exploration of near-surface planetary soil, and the movement of vehicles. Penetration testing constitutes a well-established method for the backcalculation of the soil properties. In an initial study, indentation of a rod into granular material placed within a mold has been investigated for variable density states. Regolith simulants, natural soils, and glass have been considered. The test results are presented and key features are elucidated showing that the sole use of the grain size distribution curve as a prediction tool within each subgroup is not appropriate. The effects of the rigid base and the rigid walls are clearly revealed in the response. INTRODUCTION Experience and available know-how from geotechnical engineering can be transferred into space science for the exploration and characterization of planetary soil surfaces. In addition to the low gravity environment, the testing and characterization methods must take into consideration the reduced size of such equipment. Penetration testing constitutes a well-established method for the back-calculation of the soil properties variation with depth from the surface, cf. Zacny, Wilson, Craft, Asnani, Oravec, Creager, Johnson, and Fong (2010). Furthermore, penetration may be employed to install sensors within the soil for directly measuring physical properties. Finally, penetration resistance provides an estimate for the expected bearing strength at landing sites. Sophisticated equipment has been developed for assessing percussive excavation forces, Green and Zacny (2014), and a variety of numerical methods are available for the numerical simulation. Still, the greatest uncertainty pertains to the mechanical behavior of the regolith material and the transfer of the properties
observed under terrestrial conditions to the space environment. In an initial study on penetration resistance of small probes, several tests were carried out using a fixed geometry and varying the relative density of the dry granular material. Regolith simulants, natural soils and glass were considered. In the sequel, these tests are described and the essential findings are summarized. MATERIALS Within the context of the present study, the following nine materials were tested: a) filter dust; b) MMS-dust; c) MSS-D; d) quarry fines; e) sand WF34; f) MMS-sand; g) grey sand; h) glass beads; i) glass splinter. The following abbreviations are used: MMS for Mojave Mars Simulant (unaltered basaltic sand); MSS-D for Mars soil simulant as a mixture of quarz and olivine sand that has been used at DLR for several years; WF34 for silica sand from the Weferlingen plant in Germany. In the soil mechanics terminology, these materials can be roughly divided into two groups: silts (a to d) and sands (e to i). In the sequel, materials a to d are referred to as fine-grained. The grain size distribution curves of these materials are shown in Figure 1. Further properties determined are the specific gravity ρ s, and the void ratio in loosest and densest condition, e max and e min, respectively. The results are summarized in Table 1. It can be seen that the void ratio difference (e max e min ) varies between 0.14 and 0.58. The latter is a measure of the compactibility of the material. Figure 1. Grain size distribution curves of the materials tested
Table 1. Properties of the material tested Material ρ s [Mg/m 3 ] e max e min Filter dust 2.50 2.08 1.15 MMS dust 2.83 1.72 0.90 MSS-D 2.90 1.15 0.43 Quarry fines 2.74 1.53 0.80 Sand WF34 2.64 0.88 0.56 MMS sand 2.89 1.21 0.75 Grey sand 2.65 0.93 0.52 Glass beads 2.48 0.75 0.55 Glass splinter 2.76 0.93 0.53 TESTING PROCEDURE The majority of the tests were conducted on a mold of the diameter 71 mm and height of 112 mm. For some additional tests, a larger mold with a diameter of 150 mm and height of 125 mm was used. Material was placed in the mold at densities corresponding to loose, medium-dense and dense conditions, as defined below. A stainless-steel rod of 10 mm in diameter with a flat base was selected as penetrator, Figure 2. Tests were performed under a constant displacement rate of 1.2 mm/min. The force needed to penetrate the bar was continuously recorded by an appropriate load cell. The test was terminated when the rod reached a depth of 30 mm. Figure 2. Rod during penetration For each soil type, three values of density were considered, corresponding to loose, medium-dense, and dense conditions. The target values were defined in terms of the relative density D r with D r e = e max max e e min
The target values for the loose and dense conditions were D r = 0 and D r = 1, respectively, while for the medium-dense condition D r = 0.50 was chosen. Each test was repeated once. The loosest state was achieved by pouring the material from a funnel with a prescribed spout opening at a fixed distance from the base. The densest state was achieved by hitting the mold with a two-pronged fork for a specified count number. The respective procedures are defined in the German standard DIN 18126. The target values were achieved with sufficient accuracy. All material tested had been oven dried before placement into the mold. RESULTS Figure 3 to Figure 5 show the variation of the force with the penetration depth of the rod for the smaller size mold. The curves are made as the sum of two test series with comparable relative densities of the material. Figure 5 on the right shows the respective curves for the fine-grained materials tested (filter dust, MMS dust, MSS-D, quarry fines) for the larger size mold (150 mm) in a dense state. It should be mentioned that the two test series exhibited small scatter in the results. The following trends are noteworthy: - Comparing the fine-grained material with the sands and glass material at low densities, it can be seen that the force vs. displacement relationship of the latter is almost linear, while fine-grained material shows a sub-linear variation (Figure 3). - In loose condition, quarry fines and MSS dust exhibit similar behavior despite the different grain size distribution (Figure 3, left). - The medium-dense condition increases the force required for penetration about six times (Figures 3 and 4). - In the medium-dense state, the fine-grained materials showing similar behavior are the quarry fines and the filter dust (Figure 4). - In the medium-dense state, the force vs. displacement curves of the fine-grained material become more linear (Figure 4, left), while the curves for the sandy soil start exhibiting a super-linear behavior (Figure 4, right). - In the dense state, the fine-grained material clearly exhibits a super-linear increase of the force with advancing penetration (Figure 5, left). Furthermore, it is evident that the boundary conditions, i.e. the distance to the rigid base and the rigid wall, clearly affect the response: While in the loose state, the forcedisplacement relationship up to the penetration depth considered has depending on the material a sub-linear or linear form, it reverses to a super-linear form with a dramatic increase of resistance when dense conditions prevail. For example, the resistance at final penetration depth in MMS dust increases from 0.9 N in the loose state to 4 N in the medium-dense state and finally reaches 170 N in the dense state (Figure 3 to 5, left).
Figure 3. Penetration in loose material; mold diameter 71mm Figure 4. Penetration in medium-dense material; mold diameter 71mm Even in the larger mold, the effects of the rigid boundaries become apparent: The resistance value for MMS dust drops down to 35 N (instead of 170 N), but the forcedisplacement curves still follow a super-linear trend (Figure 5). Further, it may be observed that for the fine-grained material in Figure 5, the curves start to exhibit an accelerated increase after reaching a threshold penetration depth of approx. 15 to 20 mm. A classification and prediction of the penetration behavior solely from the grain size distribution curves does not seems possible. However, a distinction between finegrained material (dust/silt) and sandy material is appropriate. As expected, at the same relative density, sandy material exhibits greater penetration resistance compared to fine-grained material. The ratio is on the order of 2 to 3.
Figure 5. Penetration in dense material for two different mold geometries CONCLUSIONS The penetration tests carried out in two classes of soil, fine-grained dust and dust-like material on the one hand and sandy material on the other, showed the main trends expected to prevail when the relative density of the material is varied. The use of the grain size distribution curve as a prediction tool within each subgroup is not appropriate. Testing soil materials with nearly similar grain size distribution curves can yield different penetration resistances. One possible reason for this could be the influence of the particle shape. The effects of the rigid base and rigid walls have also been identified from the response. Future work will include investigations with respect to particle shape and boundary conditions (diameter and height of the mold). ACKNOWLEDGMENT This work was supported by the Helmholtz Alliance ROBEX. REFERENCES Zacny K., Wilson J., Craft J., Asnani V., Oravec H., Creager C., Johnson J., Fong T. (2010), Robotic lunar geotechnical tool, 2010 ASCE Earth and Space, Honolulu. Green A., Zacny K. (2014), Effect of Mars atmospheric pressure on percussive excavation forces, Journal of Terramechanics 51:43 52. DIN 18126:1996-11. Soil, investigation and testing - Determination of density of non-cohesive soils for maximum and minimum compactness.