Abstract-only submissions

Abstract-only submissions to the 11th ISTVS Asia-Pacific Regional Conference

The repository of conference recordings is: https://vimeo.com/showcase/istvs2022. The password for the repository was sent by email to registered attendees.

0196 / Analysis on Climbing Mode of Zhurong Rover

Zhen Chen, Meng Zou, Dong Pan, Baofeng Yuan, and Lining Chen

Video: https://vimeo.com/showcase/9710424/video/747934001

Abstract: Due to the complex situation on the surface of Mars and the long delay of signal transmission between Mars and the Earth, many rovers of different countries have experienced the situation of reduced mobility or even failed exploration missions due to climbing process on Mars. In order to avoid such a situation, this paper takes ‘Zhurong’ Rover as an example. Four kinds of climbing modes that it can use are analyzed namely normal climbing, Z-type climbing, oblique climbing, bionic wriggle climbing. The advantages and disadvantages of four kinds of climbing modes and their applicable scope are compared. When the slope is small, normal climbing should be used. When the slope is moderate, Z-type and oblique climbing should be used. When the slope is large, bionic wriggle climbing should be used. Through the experiment, the critical value of the rover climbing slope is 23°, when the rover climbing slope is greater than 23°, the normal can not climbing normally which should be backward downhill. In addition, the current criterion other than the slope criterion is given to improve the climbing efficiency and safety of the rover under different climbing modes. The critical value of the current is 1.92 A. the research results can provide experimental data and reference for the in-orbit use of ‘Zhurong’ Rover.

1704 / Comparative Analysis of Hydrodynamical Efficiency of Full-Submerged Archimedes Screws of Rotary-Screw Propulsion Units of Snow and Swamp-Going Amphibious Vehicles with Single and Tandem Propulsor Design

Svetlana Karaseva, Aleksey Papunin, Vladimir Belyakov, Vladimir Makarov, and Dmitry Malahov

Video: https://vimeo.com/showcase/9710424/video/748218893

Abstract. The paper considers the problems of increasing of efficiency of amphibious vehicles with rotary-screw propulsion units (RSP). The results of computer simulation of hydrodynamical interaction of full-submerged Archimedes screws of rotary-screw propulsion units with water area in travelling and mooring modes are presented. The computer simulation was done for geometrically similar Archimedes screws with single and tandem design. During simulation the hydrodynamical characteristics of Archimedes screws of RSP with the most typical for snow and swamp-going vehicles geometrical characteristics were studied. The three-run Archimedes screws with helix angles in the range from 24 to 39 degrees were considered. The non-dimensional dependences of thrust coefficient, torque coefficient and propulsive coefficient on advance ratio for travelling mode (performance curves) and of relative quality coefficient on rotation speed of Archimedes screw for mooring mode got by the results of computer simulation are given and analyzed. The advantages of tandem design of RSP in comparison with single design with relation to realization of overwater characteristics of snow and swamp-going amphibious vehicles are highlighted and quantitative estimated.

2398 / Modeling Tire-Soil Compression Resistance on Artificial Soil Using A Scaling Law of Pressure—Soil Sinkage Relationship

Pius Jjagwe, Mehari Tekeste, and Nisreen Alkalifa

Video: https://vimeo.com/showcase/9710424/video/750588095

Abstract: Semi-empirical traction models utilize soil parameters estimated from ASABE cone and flat plate soil sinkage data; however, limited studies exist that apply the scaling law of the soil measurement tool to a full-scale tire to soil system at various initial soil conditions. This study investigated the effects of three scaled plates (size and shape) and two density conditions on pressure-sinkage relationships on an artificial soil. Elliptical shape tire-soil and contact area of 484 cm2 were estimated from a vertical loading a tire (235/75R15, tire inflation pressure of 179 kPa, and 8 kN vertical load) soil bin test on an artificial soil. Pressure-sinkage data was collected on an artificial soil column at 1.21 Mg/m3 (loose) and 1.41 Mg/m3 (dense, 75% of Proctor Density) initial conditions using circular, rectangular, and square, each at three scaled areas (n = 0.5, n = 0.25, and n = 0.125, where n =1 is the tire-soil estimated area from the single tire soil bin test). A scaling law with a strong correlation was established between the geometry scale and the energy in compressing the soil. The plate pressure data from n = 0.125 exhibited an asymptotically decreased pressure with depth, similar to the soil cone penetrometer data. The pressure-sinkage data from n = 0.5 exhibited similar trend as Reece-Bekker-Wong models but varied by the initial soil bulk density. Using ground-contact pressure of the tested tire and pressure-sinkage prediction model (n = 0.5), the loose soil resulted in a 38% increase in simulated sinkage compared to the dense soil. The study demonstrated applying a scaling law to simulate the tire-soil system on soft and dense soils. Semi-empirical prediction of motion resistance ratio using data from single-tire soil bin on soft and dense soils will be presented.

2681 / Research on High Traction Bionic Wheel Based on Traction Characteristics of Jerboa Foot

Hao Pang, Liangliang Zhao, Huailiang Wang, and Rui Zhang

Abstract: The lunar surface is covered with loose lunar soil, and the bearing capacity and shear resistance of the lunar soil are relatively low. Therefore, when the wheels roll on the lunar soil, the lunar soil particles flow easily, which reduces the passing performance of the mobile system. Jerboa has superior traction in sandy terrain. In this paper, the 3D reconstruction of the jerboas foot is firstly carried out, and three models of jerboas feet with different postures are established according to the posture changes of the jerboas foot. The discrete element numerical simulation of the sand jumping process was carried out to explore the interaction mechanism of the jerboas feet and sand when jumping in the sand.The simulation shows that jerboas toenails play a role in improving traction during the sand jumping process, and the trapezoidal toe gap structure with narrow upper and lower lower feet plays a role in sand-fixing and current-limiting in the process of foot-sand interaction. The key structure to improve the traction performance of jerboas foot was extracted through simulation analysis, and a high-traction bionic lunar rover wheel was designed by using the jerboa's foot as a bionic prototype. The wheel structure was optimized by discrete element numerical simulation. Taking Yutu wheel as the control wheel, the traction performance of the bionic wheel was verified by the wheel soil test on the bench.

2928 / Bearing Capacity Analysis of Bionic Walking Wheel for Manned Lunar Rover.

Rui Zhang, Liangliang Zhao, Yupei Du, Bin Zhao, Hao Pang, Lige Wen, Weijun Wang, Hua Zhang, Zhenyu Hu, Zhenyu Hu, and Meng Zou

Abstract: The special curved shape of African ostrich sole can limit current and fix sand and improve the bearing capacity of wheel. In the case of limited size and mass of lunar rover wheels, the use of flexible wheels can reduce the occurrence of slippage and subsidence on the soft surface of the lunar rover, and the bearing capacity of the existing flexible wheels cannot meet the requirements of heavy load. Therefore, an innovative rigid-flexible coupling bionic walking wheel was proposed, which was composed of the bionic main monomer and the integrated rim hub formed by the array of rigid-flexible coupling sub-monomers of the flexible imitation ostrich sole morphology and rigid limit structure. In order to fully analyze the mechanical properties of the rigid-flexible coupling bionic walking wheel under different loads, the Apollo elastic wheel of the same size was proposed as the control group. Based on the lunar soil thrust received by the wheel when driving on the lunar surface, the lateral force received when turning and the huge impact force received when driving, The static load and impact load of vertical load, tangential force and lateral force are simulated by finite element method. The thickness of the sheet (1.5mm) was determined by vertical load simulation, taking into account elasticity and stiffness. Through the simulation of lateral force, tangential force and impact load, it is found that the control wheel outside the bionic wheel has plastic deformation. The results show that the rigid-flexible coupling bionic walking wheel can bear larger load and impact. This research has certain reference value for the follow-up development of heavy flexible metal wheel.

2960 / Terramechanics Model Augmentation Using Machine Learning

Eric Karpman, Jozsef Kovecses, and Marek Teichmann

Abstract: Thus far, terramechanics simulations for real time applications have focused largely on the well-established semi-empirical models as these are the only ones that are fast enough to be run in real time. For wheel-soil interaction these are the models hallmarked by the names of Bekker and Wong, and for other earthworking maneuvers the Fundamental Earthmoving Equation (FEE) proposed by Reece is commonly used. One major drawback to these models is that they are formulated with a steady state assumption which makes their accuracy in dynamic simulations questionable. More flexible simulation alternatives include the Finite Element Method (FEM) and the Discrete Element Method (DEM) - both of which are extremely computationally demanding which can be prohibitive in many practical applications. In this work, we propose a method for utilizing advances in machine learning to augment the forces predicted by existing models in an effort to improve their accuracy in dynamic simulations. This is done by running virtual experiments using DEM simulations and comparing the soil reaction forces at each time step to those that would be predicted by the semi-empirical models. The difference between the actual recorded forces and those predicted by the models are used to train a neural network. The objective is for the neural network to learn how much to add or subtract from the forces predicted by the steady state model under any given dynamic conditions. Once trained, the neural network could be used in combination with the semi-empirical models to run more accurate terramechanics simulations without the need to run computationally expensive DEM or FEM simulations. As a proof of concept for this approach, we will illustrate how it can be applied to the augmentation of the Fundamental Earthmoving Equation in the simulation of a blade cutting soil, and also for the augmentation of the Bekker-Wong wheel-soil model.

3004 / Obstacle Avoidance of Mobile Robots Using Modified Artificial Potential Field Algorithm Based on Vortex and PID Adjustment

Zhuoran Sheng and Song Wang

Abstract: In recent years, autonomous mobile robot has become a key research direction in academia due to their good application prospects. The ability to autonomously avoid obstacles and reach a preset target point is the basic requirement of an intelligent mobile robot. As a common obstacle avoidance algorithm for mobile robots, artificial potential field algorithm (APF) has defects such as local minima problem and GNRON problem. Aiming at the shortcomings of the traditional APF algorithm, this paper proposes an improved artificial potential field algorithm based on point vortex and PID adjustment. The implementation of the improved algorithm is based on the two following main points. Firstly, an irrotational point vortex flow field is added to the original repulsion field of the obstacle to form a composite potential field. Properties of point vortex are used to provide additional virtual deflection force to the robot to avoid getting stuck. Secondly, aiming at GNRON problem, the strength of vortex and the range of action of the vortex repulsive composite potential field are adjusted in real time by means of PID adjustment. In order to verify the feasibility of the improved algorithm proposed in this paper, it is compared with the traditional APF algorithm in 7 cases of environment with different obstacle layouts. The results show that the improved algorithm can be effectively applied to a variety of obstacle environments, and successfully solve the local minima problem and GNRON problem of the traditional APF algorithm. The mobile robot can flexibly and efficiently avoid obstacles and reach the end point.

3290 / Why we need alternative ground robots to traverse sandy and rocky extraterrestrial terrain, and how we can progress towards them

Chen Li and Kevin Lewis

Video: https://vimeo.com/showcase/9710424/video/748992412

Abstract: Robotic spacecrafts and vehicles have helped expand human’s reach in many planetary exploration missions. Most mobile ground robotic vehicles for planetary exploration use wheeled or modified wheeled platforms, and they have been extraordinarily successful at completing intended mission goals. However, due to the limitations of wheeled locomotion, these robotic vehicles have been largely limited to relatively benign, solid terrain with sparse obstacles, and they have avoided extreme terrain with loose soil/sand and cluttered large rocks. Unfortunately, such challenging terrain is often among the most scientifically interesting for planetary geology. Many animals traverse such challenging terrain seemingly at ease, but robots have not matched their locomotor performance and robustness. This lack is largely due to a lack of understanding of the fundamental principles of how effective locomotion (or lack thereof) is generated from controlled interaction with complex terrain, on the same level of flight aerodynamics and underwater vehicle hydrodynamics. Fundamental understanding of legged and limbless locomotor-ground interaction gained over the past few decades has already enabled stable and efficient bio-inspired robot locomotion on relatively flat ground with small, scattered obstacles. Recent progress in the new field of terradynamics of locomotor-terrain interaction has also elucidated the principles of legged and limbless locomotion on loose soil/sand via a diversity of surface and subsurface locomotor modes, as well as over cluttered large obstacles via destabilizing locomotor transitions across different modes overcoming obstacles comparable to or larger than robots themselves. Such bio-inspired multi-legged and limbless robots using terradynamic principles will provide versatile, robust alternative platforms for traversing scientifically important extreme extraterrestrial terrain and expand the reach in planetary exploration.

4136 / DEM Analysis of the Dynamics of Granular Media Considering with Interparticle Forces under Low Gravity Condition

Takuru Nishino, Kenta Takase, Shingo Ozaki, Takao Maeda, Mitsuhisa Baba, and Masatsugu Otsuki

Video: https://vimeo.com/showcase/9710424/video/750291615

Abstract: For future satellite and planetary explorations, it is important to understand the dynamic interaction between the soft ground and the mechanical system under low gravity. One of the key points in understanding the interactions in low gravity environments is the competition between gravity and the forces acting between particles. The typical forces between particles or between particle and object are considered to be electrostatic and van der Waals forces. These forces are very small compared to gravity under 1G environment on Earth and have little effect on the dynamics. However, it has been suggested that in low gravity environments on small bodies, these forces can compete with gravity and affect the dynamics of granular media. In this study, the analysis using discrete element method (DEM) of rod penetration into granular media and pad impact was performed. First, various parameters were determined for the interparticle forces. The parameters were determined by considering the relationship between the force and the particle size. Next, DEM analyses were performed under four different interparticle force conditions (none, constant, electro-static, and van der Waals) and several levels of gravity environments, to compare reaction forces, sinkage, and particle dispersal. Comparison of the reaction forces, sinkage, and particle dispersal obtained from the analysis confirms the influence of interparticle forces under low gravity condition.

4170 / Development and Traveling Performance Analysis of Driving Test Rover in Sandy Terrain

Masataku Sutoh, Yuji Katsumata, and Sachiko Wakabayashi

Video: https://vimeo.com/showcase/9710424/video/747931452

Abstract: The interests on the lunar exploration and utilization are recently increasing in worldwide. The Japan Aerospace Exploration Agency, JAXA, also plans to conduct a lunar polar exploration mission, in which lunar volatiles such as water ice are investigated using a rover, and various activities toward human lunar exploration in the future. Towards these missions, it is essential to analyze the traveling performances of vehicles operating in an environment similar to the surface of the Moon which is covered with loose regolith. In this study, the development and performance analysis of a driving test rover are described. We first developed a driving test rover for motion analysis of vehicles traveling on sandy terrain. This four-wheeled rover has a driving system compatible with ground vehicles on Earth. In addition to the lunar exploration, this rover is also used to analyze the motion behaviors of the vehicles on Earth. Subsequently, we conducted traveling tests using the rover in an indoor sand field having an area of approximately 400 m2. The motion of the rover was three-dimensionally measured on flat and inclined terrains using a motion capture system equipped in the field. Furthermore, numerical simulations were conducted using the dynamic model of the rover based on the wheel-terrain interaction. From the experimental and simulation results, a similar tendency was observed regarding the motion behaviors of the rover.

4323 / Development and Demonstration of Robotic Investigation Tool for Terrain Mechanical Property

Yuzuki Morita and Genya Ishigami

Video: https://vimeo.com/showcase/9710424/video/746407755

Abstract: The Artemis Program led by NASA and the Global Exploration Roadmap summarized by the International Space Exploration Coordination Group has been focusing on the construction of lunar infrastructure in the coming decades. Terrain parameter estimation is essential for the design phase of infrastructure construction in rough terrain. A robotic platform should perform the estimation because of the harsh environment on the moon. While the classical terramechanics model holds several terrain parameters such as cohesion, internal friction, or shear modulus, the resistive force theory needs one lumped parameter called scaling factor, which is basically identified with a flat-plate sinkage test. We developed a handheld investigation tool for the terrain scaling factor that can also be mounted on a mobile robot. The tool consists of a flat plate with a precise linear actuator for intruding the plate into the soil. A force sensor attached between the plate and the actuator measures the resistive force generated during the plate intrusion. The scaling factor is then determined by the relationship between the force and intrusion depth. In our work, we first conducted a laboratory control test of the tool and verified that the tool consistently identifies the scaling factor under different density and grain size distribution conditions. Subsequently, we mounted the tool on a four-wheeled robotic mobile platform and remotely operated the robot deployed on a sandy field, where the robot experimentally demonstrated a terrain parameter investigation. The demonstration confirmed that all of the scaling factors determined by the robotic tool were within the range of the data obtained in the laboratory control experiment.

5636 / Numerical Analysis of Multi-Pass Effect Based on the Extended Terramechanics Theory

Shingo Nakano and Shingo Ozaki

Video: https://vimeo.com/showcase/9710424/video/750254517

Abstract: In recent years, various off-road vehicles have been used for disaster response and space exploration. In the future, it is expected that off-road vehicles will become even more important. Along with this, there is also a need to improve the traveling performance of off-road vehicles. Currently, multibody dynamics analysis is promising as a preliminary examination method that is indispensable for research, however, to improve the analysis accuracy, the interaction between the ground and wheels, such as deformation of the ground surface shape and multipath effect, have to be appropriately considered. Therefore, in this study, we focus on the terramechanics theory and propose an extended model that combines cellular automata and conventional theory. In the proposed model, the stress distribution between the road surface and the wheel is evaluated by the conventional Bekker-Wong-Reece model, and change in the bulk density and soil property is evaluated based on the cellular automaton. The purpose of this study is to establish a methodology for MBD analysis that can take into account the deformation of the terrain surface and the change in soil state variables due to wheel traveling. Then, we performed the numerical analysis of traveling on flat ground with the single-wheel model and small rover model. Here, we used the general-purpose simulation software Simscape, and implemented the proposed model into it. As a result, we successfully showed the effectiveness of the proposed methodology by the systematic analysis.

5943 / Research on Bionic Anti-Skid Tires Suitable for Icy and Snow-Covered Roads

Rui Zhang, Yu Han, and Hao Pang

Abstract: In the icy and snowy environment, the tire surface of the vehicle, as a key component of the vehicle's interaction with the icy and snow-covered ground, plays a vital role in improving the vehicle's passability. Reindeer run on ice and snow all year round, and their excellent foot attachment characteristics are obvious. A bionic tire tread structure was designed according to the structural characteristics of the reindeer's sole and the geometry of the reindeer's sole, and the tire structure was optimized by the finite element method (CAE). The simulation results show that the hard ice-breaking part of the tire will cut into the ice layer and generate greater friction with the ice surface, and the soft material part of the tire adheres to the snow, reducing the effect of the snow layer on the tire's traction. Finally, the CAE models of the bionic wheel and the conventional snow wheel are established. Wheel simulations show that the driving traction of a wheel with a bionic tread is 85.4% greater than that of a conventional snow tire. The above studies show that the bionic structure greatly improves the adhesion performance of the wheel surface. In addition, the bionic hair can play a better role in combination with the bionic wheel contour based on the curve of the reindeer's foot. This paper can provide new inspiration and methods for the design of tires on icy and snowy roads.

7349 / Research on Energy-Saving Underactuated Bionic Biped Robot

Youhao Diao, Zhiqiang Zhuang, and Xuebo Wang

Abstract: There is still a certain gap between the current biped walking machines and humans in terms of walking efficiency and walking naturalness. important scientific value. Taking the human lower limb musculoskeletal system as the learning model, the bionic design of the biped robot based on the passive walking principle is carried out, and the physical prototypes of the fully passive walking machine and the underactuated walking machine are developed. Completed the development of a knee-free completely passive bionic walker with bionic knee joint, bionic foot, bionic subtalar joint and metatarsophalangeal joint, and carried out Adams-based simulation analysis for the virtual prototype to verify the feasibility of its walking. The experimental results show that the walking efficiency of the new kneeless fully passive bionic walking machine is increased by 50%, while achieving continuous and stable walking. In this paper, through the in-depth analysis of the human body's delicate and complex muscle-skeletal system and high-efficiency and energy-saving movement patterns, the inspiration is obtained, and then the mechanical body of the biped walking machine based on the passive walking principle is finely and subtly designed. Design brings bionic design inspiration.

7739 / Modeling of the Shear Displacement for Tracked Vehicles in Transient Maneuvers

Yang Jiao and Jozsef Kovecses

Abstract: Many design specifications of tracked vehicles are given in terms of the steady-state characteristics. They are indeed quite important to tracked vehicles because of the nature of their operations. However, in some situations such as starting or steering, the transient state performance of these tracked vehicles could also be of great significance. The main trend solution to model the traction force of off-road vehicles is to use shear stress and shear displacement curve by assuming the track soil interaction behaves like the one in the direct shear box test. Therefore, the approximation of shear displacement is quite important in order to accurately predict the shear stress which is directly related to the traction performance of a tracked vehicle. In this work, we propose a general representation of the shear displacement along a discretized track model. This is done by analyzing the shear displacement accumulation at each special point along the track in every time step. Then a new shear displacement model is proposed based on the general representation. The performance of this model is compared with Wong’s shear displacement model and Solis and Longoria’s (SL) shear displacement model. Wong’s model has the simplest form but has difficulty in predicting transient state shear displacement. SL model achieves a better performance in transient state by estimating the shear displacement at each discretized track element with an ordinary differential equation. In order to validate the new model and compare the performance of the three models, a simplified track soil interaction DEM model is developed by using a commercial DEM simulation software ′EDEM′. The shear force obtained from the DEM simulation is used as a reference for the results obtained by using semi-empirical shear force model with different shear displacement models. As proved by the comparison, the new method shows a clear advantage in dynamic shear displacement prediction over Wong’s method. The difference between the SL model and the new model is not as large but also obvious in some particular test cases.

8104 / Determination of the Necessary Indicators of Vehicles and the Movement Surface for Calculating the Mobility Criterion

Alina Markovnina, Umar Vakhidov, Vladimir Belyakov, and Vladimir Makarov

Abstract: This article discusses wheeled vehicles moving on mixed pavements consisting of various types of soil, and indicates the driving conditions typical for such pavements. The main tasks of traffic control are analyzed and it is revealed that the main indicator of the possibility of a vehicle moving in difficult road conditions is the reserve of traction force. The paper presents the general aspects of the methodology for calculating the mobility criterion, according to which it is possible to determine the possibility or impossibility of overcoming a terrain by a given machine. A system of indicators is described that characterizes the possibility of driving a car in given road conditions. The schemes of interaction of the wheel mover with the track bed during rectilinear, curvilinear motion and during maneuvering when bypassing discrete and water obstacles are given. The paper shows various patterns of surface deformation when passing by several axles of a vehicle. The influence of profile obstacles on the cross-country ability and the value of the loss of mobility in terms of cross-country ability are considered. The relationship between the magnitude of the traction force and the magnitude of the obstacle that the vehicle can overcome is determined.

8362 / A High-Fidelity Dynamics Simulation Method for Legged Robot Based on Foot–Terrain Interaction Model

Jianghua Ge, Jianghu Wu, and Xiaofei Zhu

Abstract: The usage of legged robot for extraterrestrial exploration is a research hotspot at present, and the emulation method to simulate the landing and roving process can effectively improve the design efficiency and verify the control, but due to the complex terrain environment on the surface of the celestial body, it is a challenge to realize the unification of simulation accuracy and efficiency, therefore, the foot–terrain interaction process is an important factor to limit the simulation effect. In this paper, an ADAMS-based simulation system for legged robot dynamics is proposed to achieve high precision simulation of legged robot by introducing a foot–terrain interaction model. The main work includes three aspects: the derivation of dynamic impact model, static pressure model and tangential walking model for legged robot rover in complex terrain based on foot–terrain interaction mechanics model, the development of foot–terrain contact interaction solution module for legged robot rover based on the aforementioned three models and the secondary development of foot–terrain interaction model solution module based on ADAMS user subroutine. The simulations of landing and patrols were carried out to verify that the designed legged robot can effectively meet the landing and patrols functions. Compared with the CONTACT simulation method of ADAMS, the ADAMS high-fidelity dynamics simulation system can verify the probe dynamics requirements with higher accuracy and provide reference for foot pad design and motion control optimization of the legged robot.

8685 / Design and Testing of Parallel Embedded Six-Axis Force Sensors for Paddy Walking Wheels

Zaiman Wang, Jianfei He, Erli Zhang, Miao Su, Yue Huang, Wenwu Yang, Minghua Zhang, Weiqin Jia, Yuanli Huang, Yifan Ma, Dongyang Yu, Peizhao Zhong, Zhihao Zeng, and Ziyou Guo

Abstract: The operational load spectrum of paddy power machinery can effectively reflect the machine-paddy soil interactions. In response to the problems of difficult operational performance data measurement and lack of observation means for paddy power machinery under complex working conditions in paddy fields, a paddy wheel parallel embedded six-axis force sensor test system is developed and mounted on the field soil through test device and paddy power chassis for performance testing. The test system consists of six-axis force sensors, angle sensors, signal amplifiers, slip ring components, a data acquisition system, data analysis software, and cables. During operation, the five six-axis force sensors transmit electrical signals from the real-time measurement of six-axis forces to the signal amplifier and slip-ring components, and the slip-ring transmits the signals transmitted by the signal amplifier to the cable, which in turn transmits them to the data acquisition system, where the measurement data are processed and displayed in real-time by the data processing software on the host computer. To verify the accuracy of the measurement data of the parallel embedded six-axis force sensor of the paddy wheel in the complex working environment of a paddy field, the bench performance experiment of the field soil test device was carried out to reveal the relationship between the driving force and rolling resistance of the paddy wheel under different rotational speed and different depth of soil entry conditions, and to investigate the influencing factors affecting the operating performance of the paddy field power chassis. The test results show that the measurement data of the six-axis force sensor test system of the paddy walking wheel is accurate, the relative error is less than 0.2%FS of the measurement accuracy, and the principle and method of the test system are reliable; the six-axis force sensor test system can be used for the load spectrum data acquisition of paddy power machinery field operation.

9092 / Examining the Simulation-to-Reality Gap of a Wheel Loader Interacting with Deformable Terrain

Koji Aoshima, Daniel Lindmark, and Martin Servin

Video: https://vimeo.com/showcase/9710424/video/750588946

Abstract: Simulators are essential for developing autonomous control of off-road vehicles and heavy equipment. They allow automatic testing under safe and controllable conditions, and the generation of large amounts of synthetic and annotated training data necessary for deep learning to be applied. Limiting factors are the computational speed and how accurately the simulator reflects the real system. When the deviation is too large, a controller transfers poorly from the simulated to the real environment. On the other hand, a finely resolved simulator easily becomes too computationally intense and slow for running the necessary number of simulations or keeping realtime pace with hardware in the loop. We investigate how well a physics-based simulator can be made to match its physical counterpart, a full-scale wheel loader instrumented with motion and force sensors performing a bucket filling operation. The simulated vehicle is represented as a rigid multibody system with nonsmooth contact and driveline dynamics. The terrain model combines descriptions of the frictional-cohesive soil as a continuous solid and particles, discretized in voxels and discrete elements. Strong and stable force coupling with the equipment is mediated via rigid aggregate bodies capturing the bulk mechanics of the soil. The results include analysis of the agreement between a calibrated simulation model and the field tests, and of how the simulation performance and accuracy depend on spatial and temporal resolution. The system’s degrees of freedom range from hundreds to millions and the simulation speed up to ten times faster than realtime. Furthermore, it is investigated how sensitive a deep learning controller is to variations in the simulator environment parameters.

9219 / Penetration Dynamics of Asteroid Exploration Landing Based on DEM

Yongbin Wang, Shiqing Wu, Shutong Chen, Xinghua Liu, Huan Liu, He Jia, Xuyan Hou, Shaomin Liang, Shunying Ji, and Zijian Zhang

Abstract: Asteroid exploration is one of the hot areas of deep space exploration in recent years. In this paper, a landing attachment system for asteroid landing detection is proposed, which can complete the landing and attachment of the lander on the asteroid surface. Based on the discrete element method, this paper analyzes the landing impact and penetration dynamics in the process of land penetration. In this paper, the triaxial compression performance of simulated lunar soil is tested, and the physical and mechanical parameters of simulated lunar soil are studied, so as to realize the parameter matching between meso parameters of bulk medium and macro mechanical properties of soil. The penetration process of the detached body is studied. The penetration characteristics under different incident angles, velocities and shapes of the spear body are studied. The characteristics of the depth, velocity and acceleration of the anchor body with time are obtained. The landing impact dynamic characteristics of the anchoring device are obtained. This study provides a new analysis method for the study of asteroid detection anchoring and attachment, and has reference significance for the design of asteroid and other planetary landing and attachment detection.

9260 / Experimental Analysis and Numerical Modeling of Bulldozing Force with Varied Soil Moisture Content

Naohiro Sato and Genya Ishigami

Video: https://vimeo.com/showcase/9710424/video/748991406

Abstract: Bulldozing/earthmoving works on loose soil are typical operations of construction machines. The bulldozing force acting on a machine blade from soil often impedes the machine mobility. The interaction forces depend on the mechanical property and/or bulk characteristics of soil such as moisture content (i.e., dry or wet sand). The Distinct Element Method (DEM) has been widely employed in precise and high-fidelity simulation for varied types of soil particles interacting with machines. However, the DEM requires accurate V&V (verification and validation), and in particular, such a process for a soil having different moisture content remains an open issue. Therefore, the main scope of this paper is to reveal key parameters in the DEM simulation that appropriately corresponds to specific moisture content. First, we conduct bulldozing experiments under different levels of soil moisture and investigate how the bulldozing force varies with moisture content. The experimental results show that the bulldozing force significantly increases as the moisture content saturation exceeds around 60 %. Subsequently, we employ a DEM simulation to calculate the bulldozing force under different moisture content. Here, we focus on the adhesive forces acting between soil particles because past research suggested that the moisture in the soil is largely subject to the adhesive forces. The adhesive force model of the DEM consists of two parameters, i.e., a scaling magnitude w.r.t. particle weight and the maximum adhesive distance of two soil particles. The values for those two parameters are then adjusted such that the maximum bulldozing force calculated in the DEM coincides with that of the experiments in different moisture content. Finally, we reveal the quantitative relationship of the key parameters corresponding to the various moisture content in the soil. The results imply that the DEM simulation is applicable to moist soil environments by carefully adjusting the adhesive force parameters.

9273 / Force China Analysis on Wheel-Soil Interaction Using Photoelastic Method

Yuto Yoshida, Sota Yuasa, and Kenji Nagaoka

Video: https://vimeo.com/showcase/9710424/video/748991406

Abstract: This paper presents a force chain analysis of soil-wheel interaction using a photoelastic method. So far, soil-wheel interaction has been experimentally studied by several approaches; such as an analysis of the empirical relationship between wheel’s mobility performance and wheel design parameters, measurement of soil reaction acting the wheel surface, the soil flow analysis based on a particle image velocimetry method. However, the stress distribution of a soil particle layer is difficult to be directly observed. In this study, we propose a photoelastic method to visualize the stress distributions by which photoelastic particles are used as terrain, simulating soil particles. The photoelastic disks and their layer which simulate soil are shown in Fig. 1. The photo elastic particles can visualize the applied stresses as interference fringes of light (see Fig. 1). As a result, dynamic change of stress distribution and stress propagation in the terrain can be observed. In particular, the fringe and the stress propagate between the particles and form force chains in the terrain. We developed a single-wheel testbed as an experimental apparatus for wheel-soil interaction using the photoelastic method, as shown in Fig. 2. This apparatus can quantitatively provide the relationship between the particle stresses and the interference fringes based on the brightness gradient. Moreover, we investigated the effects of mixing with different filler particles to simulate various wheel slip-sinkage states. In this paper, we first show the force chain analyses of particles beneath a single traveling wheel. Fig. 3 shows an example of the force chain visualization. Then, orientational orders of the force chains are quantitatively evaluated for small and large wheel slippage. The experimental results confirmed the photoelastic method shows the dynamic change of the stress distributions and propagation in terrain.