IAC-13-A1.6.1 THE ANALYSIS OF THE RELATIONSHIP BETWEEN MOTION RESTRAINTS CAUSED BY PRESSURIZED SUITS AND RISK OF FALLING Yasuhiro Akiyama Nagoya University, Japan, akiyama-yasuhiro@mech.nagoya-u.ac.jp Locomotion of astronauts on the planetary surface has a close relationship to performance of space suits and the risk of EVA is related to the performance. Especially, the fall of the astronaut not only disturbs the locomotion but also damages the suit and injures the astronaut. However, pressurized suits interfere the motion of user. Therefore, the relationship between the fall avoidance motion and the feature of the suit such as the relationship between the joint angle and applied torque is focused on in this study. The gait motion and the fall avoidance motion after tripping are recorded in the experiment. As a result, additional torque which mimics the pressurized suit increases speed due to the increase of cadence. However, gait parameters differed from general high speed walking. It also increases the maximum flexion angle of the knee joint. In addition, ground reaction force suggests the risk of slip and trip over. At last, although the fall avoidance strategy is not changed even in case with additional torque, the timing of landing of the tripped leg delayed. This trend means the delay of the fall avoidance motion compared to normal case. I. INTRODUCTION Locomotion of astronauts on the planetary surface is important to increase output of manned space explorations. Astronauts will need to go into the uneven surface where rovers cannot approach. However, the gait on the planetary surface is different from that on the Earth surface because of reduced gravity. In addition, space suits also affect the locomotion of astronauts. It changes the gait motion and walking ability because it interfere the motion of astronauts through resistant torque against the deformation of space suits. On the Earth surface, the fall is one of the most serious risks during gait. Although the shock of the fall is not the serious problem on the planetary surface because of low gravity, contact with sharp object or secondary accident such as falling will be a serious accident. Table. 1 shows the risk of the fall during EVA on the planetary surface. It suggests risks caused by the space suits. Especially, the limitation of the joint motion will decrease the ability of astronauts to avoid the fall. Therefore, the change of the gait motion due to space suits not only disturbs the exploration but also endanger the life of astronauts because it probably increases the risk of fall. Now, most previous studies about EVA are focused on the motion of upper limb 1,2. Although there are some studies about energy consumptions during walking on the planetary surface 3, risk of gait on the planetary surface is not researched enough. Therefore, the gait motion and risk of fall caused by the restraint of the space suits are focused on in this study. Analysis of the mechanism and performance of gait of astronauts will help us to design future space suits. Table 1: Fall risks on the planetary surface Category Hazard source Hazard Process Space suit Limitation of Unsteadily walking field of view Limitation of degree of freedom of joints by limitation of motion Environment Limitation of motion range of joints Extension force with flexion of joint Change of center of gravity caused by devices Slip Failure to adapt to the different gravity Tripping Treacherous ground by limitation of motion by limitation of motion by unexpected force and torque Failure in the function of stance leg by the change of mechanical characteristics by failure of stepping motion by failure of landing of swing leg This matrix exclude hazards of device failure II. METHOD II.I Experimental Approach against the Fall The fall motion of human is related to human cognition deeply. Therefore, an experiment using human subjects is necessary to analyse it. In this study, IAC-13-A1.6.1 Page 1 of 6
Fig. 2: Motor Actuated Lower-limb Orthosis Fig. 1: Overview of experiment the experiment to measure the gait motion and fall motion was done under permission of IRB on Nagoya University. Fig. 1 shows the overview of the experiment. The motion capture system (MAC3D, Motion Analysis) and the force plate (M3D Force plate, Tech Gihan Co., Ltd.) are used to measure the motion of the subject. Force plate is attached to the subject s sole. Using these devices, joint angles, positions and ground reaction force are monitored by 100 Hz. In addition to monitoring devices, the safety harness, protectors and supporters are used for the safety of the subject. The safety harness is connected to the gondola which is placed on the ceiling rail through air spring. Therefore, the subject can walk under the ceiling rail under safe conditions. There is the approach zone of 3 m and the recording zone of 2 m. During overground walking, the subject sometimes tripped in the recording zone. However, the presence or absence of the tripping and the timing of tripping is not informed to the subject. When tripping the subject, the object with 250 mm * 330 mm * 190 mm and 21 kg is placed on the right side of the lane. The subject cannot know whether there is the object or not because the subject s field of view of foot is covered by the goggle. In this study, the gravity is not compensated because this study aims to evaluate the effect of joint restraint at first. II. II Simulated Pressurized Suit Pressurized suits like space suits generate force to keep natural shape when it is deformed due to the inner pressure. Therefore, additional joint torque is needed to flex arms and legs from the natural angle and keep them bending. The relationship between the joint angle and joint torque which is applied by the suit depends on particular suit and it is the important parameter to evaluate it. However, it is inconvenient to use pressurized suits in the experiment because the physical characteristics of it are fixed mechanically and it is difficult to change them. Therefore, the relationship between the joint angle and joint torque is emulated using MALO (Motor Actuated Lower-limb Orthosis) which is developed in Nagoya University. Fig. 2 shows the human with MALO. It has 1DOF on each ankle, 1DOF on each knee and 2DOF on each hip. Flexion-extension of hip and knee joint can be actuated using DC motor (RE 40, Maxon). The pressurized suit generates joint torque due to the decrease of inner volume and deformation of suit surface. Therefore, joint torque becomes nonlinear and it has hysteresis 4. There are some studies which measures the relationship between the joint angle and joint torque 56 of some space suits. Data of such studies is useful to set the parameters to emulate property of the pressurized suit which is installed to MALO. Then, there are some physical models to explain joint torque caused by the pressurized suit. In this study, the tube bending model 7 is used and the parameters are tuned using data of space suits. Joint torque which based on the tube bending model is described in formula [1]. p is the pressure difference between the pressurized suit and outside world. Vc is the volume of bending section which is calculated by considering tube deformation. R is the radios of the tube. Then, φ is bending angle of the joint. d M p VC 2 R d 3 [1] IAC-13-A1.6.1 Page 2 of 6
Fig. 5: Experimental Protocol Fig. 3: Torque of Hip Joint Fig. 4: Torque of Knee Joint In this study, hysteresis is not considered for simplicity. Then, the parameter p is set at 0.3 atm, which is the pressure of EMU suits. At last, the tube radios R is not set as the real radios of pressurized suits because the model does not consider the arm and leg of user and it does not match with the data of the actual suit. Therefore, R is set to fit the moment of the suit joint. Fig. 3 and Fig. 4 show joint torque of MALO. The neutral position of the knee joint is set at 45 deg because the knee flex around 0 70 deg during gait motion in contrast the hip joint flex around -15 30 deg. II. III Experimental Protocol Four young healthy male subjects voluntarily participated in this experiment. Their age was 19-21 and the BMI was 17.4-22.8. Fig. 5 shows the protocol of the experiment. At first, the subject walks on the treadmill for three minutes for adaptation of MALO and torque pattern. Then, he walks overground 10 times (5 with tripping). The subject isn t informed whether he is tripped or not in each case and the order of tripping cases is randomized. After that, same sequence with different control algorithm is repeated. The control algorithms used in this experiment are the space suit emulation algorithm which is described above and the friction compensation algorithm. The order of algorithms is also randomized. III. RESULT At first, the gait parameters of overground walking without tripping are shown in Table. 2. Then, the angles of joints are shown in Fig. 6. Gait parameters shown in Table. 2 differ from that of general gait of healthy adults in some degree. Cadence and step length are lower and the ratio of double support phase is higher than that of general gait. This means that the subject who wears MALO walks slower than usual. There are some reasons to describe this difference. First, the weight of MALO may decrease the walking speed. Second, MALO constrains DOF such as rotation of hip joint. It probably makes the subject more cautious of walking. However, the patterns of joint angles are not strange compared to the general gait. However, these results also suggest that torque which mimicked the pressurized suit increase cadence and double support phase. This tendency does not suit with general gait motion. In general, increase of cadence Sbj no. Control Table 2: Gait Parameters Step Cadence length [/min] [m] Double support phase [%] 1 Normal 41.5 0.40 41.4 Suits 46.3 0.39 47.0 2 Normal 43.9 0.44 37.1 Suits 45.3 0.55 39.2 3 Normal 38.6 0.44 36.2 Suits 47.2 0.41 36.4 4 Normal 48.8 0.42 33.4 Suits 53.7 0.39 39.1 IAC-13-A1.6.1 Page 3 of 6
Fig. 6: Gait of Overground Walking and step length occur at the same time when speed increases. In addition, double support phase generally decreases when speed increases. However, only cadence increases and double support phase increases despite speed increases. This result suggests the acceleration of swing motion in suit case and it is described by applied knee torque. In suit case, MALO apply torque to keep knee angle 45 deg. This torque helps the motion of swing leg when it starts knee flexion, stops knee flexion and starts knee extension. Fig. 6 shows some differences caused by the difference of control algorithm. First, maximum knee angle increases in suit case despite step length does not increase. This difference supports the suggestion that applied torque affect the swing motion of knee. In contrast, range of motion of the hip joint does not change even in suit case. This trend fits with the tendency that step length does not change. Then, the plantar flexion pattern around toe off becomes gentle in suit case. This should be the indirect effect caused by knee or hip torque because the condition of ankle joint is not changed between these cases. The other difference is shown in ground reaction force. According to Fig. 7, normal force around heel strike overshoots in suit case. This suggests that the control of swing leg around landing becomes more Fig. 7: Ground Reaction Force difficult because of applied torque. This will increase the risk of slip and trip over. However, this tendency is highly individual and asymmetric. Therefore, further analysis is required to evaluate this difference. Previous study about fall suggests that trip on early swing phase affect larger than that of the other phases 8. Therefore, trip timing is set at early swing phase of the right leg. Fig. 7 shows the relationship between the knee angle and trip timing. Time series is normalized by average gait time. This graph shows the variation in trip timing due to the variation in gait motion. In addition, one case of suits case is omitted because of the recording error. Generally, there are two strategies that are selected by the tripped human. The lowering strategy aims to land the tripped leg and rebalance soon. In contrast, the elevating strategy aims to step over the object before the landing of the tripped leg 9. The subject seems to flex the knee after tripping in early swing phase in Fig. 8. This is the typical characteristics of the elevating strategy 10. However, the object is too big to step over in this experiment. Therefore, the subject lands the tripped foot before the object and he steps forward with the other leg to stop forward displacement of his body. Timing of IAC-13-A1.6.1 Page 4 of 6
Fig. 8: Knee Angle of Tripping Leg landing of the tripped foot is shown in Fig. 9. This graph suggests that timing of landing tend to be delayed in suit case. One isolated example is the case that the subject pushes the object strongly to step forward with the tripped leg. Delay of landing of the tripped foot means the delay of next step of the other leg. Therefore, it will increase the risk of fall because of the forward displacement of the body. IV. DISCUSSION This experiment does not simulate the environment with reduced gravity. However, it is necessary to think about the effect of gravity because it will affect the gait ability. Previously, gait parameters under reduced gravity are studied using simulated gravity 11. These studies suggest that speed decreases when gravity decreases. In addition, other parameters such as cadence and stance/swing ratio decrease with speed. In this study, same change occurs due to constrain of MALO. However, the effect of MALO should be opposite against that of reduced gravity because MALO probably decreases speed by increasing weight of the subject. EMG will help to describe the difference between them. However, the change of gait motion due to additional torque will occur even under the condition of reduced gravity. This is because reduced mass of the lower thigh helps to flex the knee joint and increase the maximum knee joint angle. The difficulty of verification Fig. 9: Ground Reaction Force of Tripping Leg of this phenomenon comes from the limitation of the method to generate simulated gravity. Previous methods to generate simulated gravity cannot reduce mass of each links of human. Therefore, the lower-limb motion cannot be the same with that under real low gravity environment. In addition, the method which can simulate the mass reduction do not allows fall. The fall motion under reduced gravity is also the interesting point. Reduced gravity will delay timing of landing of the tripped leg because of the reduction of force which lowering the foot. However, it also decreases speed of fall. Therefore, the effect of step timing against the ability to avoid fall should be evaluated more carefully. V. SUMMARY The relationship between joint torque of the pressurized suit and the fall motion was evaluated by the experiment. Torque caused by the deformation of the pressurized suit was mimicked by MALO which is the device controlled by motor actuators. The result of the subjects experiment suggests that torque of the pressurized suit increases cadence and double support phase. As a result, speed increased. However, it differs from the general gait pattern which is found when speed increases. In addition, timing of landing of the tripped leg delayed when torque of the IAC-13-A1.6.1 Page 5 of 6
pressurized suit was applied although the strategy to avoid fall match with the general trend. IV. ACKNOWLEDGEMENT This work was supported by JSPS Grant-in-Aid for Young Scientists B Grant Number 24710184. 1 Kunihiko Tanaka, Chikara Abe, Chihiro Iwata, Kenji Yamagata, Naoko Murakami, Masao Tanaka, Nobuyuki Tanaka and Hironobu Morita, "Mobility of a gas-pressurized elastic glove for extravehicular activity," Acta Astronautica, vol.66, pp. 1039-1043, 2010 2 Richard S. Williams, M.D., "NASA space flight human system standard volume 2: Human factors, habitability, and environmental health," NASA-STD-3001, VOLUME 2, 2011 3 Alberto E. Minetti, "Invariant aspects of human locomotion in different gravitational environments," Acta Astronautica vol. 49, pp. 191-198, 2001 4 P. B. Schmidt, D. J. Newman and E. Hodgson, "Modeling Space Suit Mobility: Applications to Design and Operations," SAE Technical Paper 2001-01-2162, 2001 5 Dana Valish and Karina Eversley, "Space Suit Joint Torque Measurement Method Validation," 42 nd International Conference on Environmental Systems, AIAA 2012-3532, 2012 6 Menendez, V., Diener, M., and Baez, J., "Performance of EVA Suit Soft Flat Pattern Mobility Joints," SAE Technical Paper 941331, 1994 7 P. B. Schmidt, "An Investigation of Space Suit Mobility with Applications to EVA Operations," Doctoral Dissertation, MIT, 2001 8 Mehmet Temel, Katherine S. Rudolph and Sunil K. Agrawal, "Gait Recovery in Healthy Subjects: Perturbations to the Knee Motion with a Smart Knee Brace," Advanced Robotics, 25, pp. 1857 1877, 2011 9 Michael J. Pavol, Tammy M. Owings, Kevin T. Foley and Mark D. Grabiner, "Mechanisms Leading to a Fall From an Induced Trip in Healthy Older Adults," Journal of Gerontology: MEDICAL SCIENCES, vol. 56A, no. 7, pp. 428-437, 2011 10 Janice J. Eng, David A. Winter and Aftab E. Patla, "Strategies for recovery from a trip in early and late swing during human walking," Experimental Brain Research, vol. 102, pp. 339-349, 1994 11 Davis BL and Cavanagh PR., "Simulating reduced gravity: a review of biomechanical issues pertaining to human locomotion," Aviation Space Environmental Medicine, vol. 64, no. 6, pp. 557-66,1993 IAC-13-A1.6.1 Page 6 of 6