Track 17. Biomechanics in Nature

7068 Th, 08:15-08:30 (P39) Walking and running dynamics explained by compliant legs: consequences, general insights, and future directions H. Geyer 1,2, A. Seyfarth 1 . 1Locomotion Lab, Friedrich-Schiller-University Jena, Germany, 2 Biomechatronics Group, MIT Media Lab, Massachusetts Institute of Technology, Cambridge, MA, USA

The basic mechanics of human locomotion are associated with vaulting over stiff legs in walking and rebounding on compliant legs in running; but whereas rebounding legs well explain the stance dynamics observed in running, stiff legs cannot reproduce that of walking. With a simple bipedal spring-mass model we could recently show that not stiff but compliant legs are essential to obtain walking mechanics: incorporating the double-support as an essential part of the walking motion, this bipedal model reproduces the characteristic stance dynamics that result in the observed small vertical oscillation of the body and the observed out-of-phase changes in forward kinetic and gravitational potential energies. This result has immediate consequences. First, it questions the traditional inverted pendulum as proper paradigm for the walking gait and hereby chal- lenges classical views about walking efficiency and the origin of the walk-run transition. Second, it demonstrates that the mechanics of both walking and running can be explained within one mechanical concept that is based on compliant leg behavior. Moreover, further investigating the bipedal model's dynamics potentially yielded general insights into the organizing principles that underlie legged locomotion: toward slow speeds the two known gaits of single force-peak running and double force-peak walking extend into a regular pattern of multiple force-peak gaits suggesting that, similar to the different energy levels electrons can occupy in the atomic shell model, walking and running are just two out of many stable solutions to legged locomotion accessed by energy or speed. Combining walking and running in one simple but compelling gait template, the bipedal spring-mass model can direct future investigations that were not accessible before; for instance, the model can directly address the mechanics of the walk-run transition, or it can assess changes in the dynamics for asymmetric gait patterns resulting from disabilities or disorders.

6969 Th, 08:30-08:45 (P39) Minimising mechanical work in quadrupedal locomotion A. Wilson. Structure and Motion Laboratory, The Royal Veterinary College, Hatfield and Structure and Motion Lab, University College London, London, UK

Legged locomotion usually involves interchange between elastic, kinetic and potential energy. In animals this interchange is likely to have a cost since tendon springs are only about 90% efficient and the serial muscles require energy to resist force. Work therefore needs to undertaken, de novo each stride. This work need not be performed at the same time in the stride as energy is dissipated or indeed in the same limb. For instance it is likely that much of the work of locomotion in horses is performed by the musculature of the hind leg. Minimising mechanical work in locomotion is therefore likely to contribute to heightened efficiency of locomotion. We have investigated these mechanisms in a range of animals during field locomotion. Speed was measured using a GPS data logger, foot-on and foot- off times for each leg were from an accelerometer attached to the dorsal foot and logged into an MP3 recorder and trunk motion determined using one or more six degree of freedom inertial sensors attached over the approximate centre of mass or the attachment points of the limbs. Limb energy storage was calculated from an analytical approximation to planar spring mass mechanics. There were significant differences between gaits in linear and rotational trunk motion as a function of speed. We calculated external work (potential energy and linear kinetic energy changes) through each stride. We modelled the horse trunk as a cylinder to estimate the rotational inertia in each plane and calculated the components of internal work represented rotation of the trunk around the centre of mass. There were significant differences between vertical displacement and pitch and roll amplitude for the different gaits. Cranio- caudal kinetic energy fluctuations were the greatest component of total energy fluctuation for all gaits at all speeds. Elastic energy storage in legs was of relatively minor importance in reducing the mechanical work of locomotion except in trotting. Pitching and rolling movements did not act as an energy store to reduce the mechanical work of locomotion.

7129 Th, 08:45-09:00 (P39) Ants running on inclines: path integration and stability T. Seidl 1 , T. Weihmann 2, R. Blickhan 2, R. Wehner 1 . 1Dept. Neurobiology, Inst. Zoology, University of Zurich, Switzerland, 2Dept. Movement Sciences, Inst. Sports Sciences, FSU Jena, Germany

Foraging ants have an astonishing ability to orient themselves in an unknown environment using path integration. This egocentric orientation behaviour

17.5. Terrestrial Locomotion $361

allows the animal to calculate the shortest way home based on iterative vector addition of the previously run path incorporating heading and length of each single segment. While path integrating the ants can also monitor inclinations and use this information for their distance estimation. The currently most preferred theory states that this distance estimation is mediated by the procession of proprioceptive information i.e. by step counting. Therefore the ants must be able to monitor both step length and inclination of the surface on which they walk. Our aim is to deduce the biomechanic basis for path integration by analyzing the ants' locomotor behaviour. In addressing this question we analyzed the kinematics of desert ants, Cataglyphis fortis, from both lateral and dorsal views. In our field station in Mahar6s/Tunisia the ants were trained to forage to a feeder located at the end of a 4 m aluminium channel. A particular section of this channel could be altered allowing for video recordings (250 Hz) of freely running ants at different inclinations (both uphill and downhill). In order to distinguish between species and task specific effects we compared Cataglyphis fortis with a species of its taxonomic sister genus, the wood ant Formica pratensis. Both ants are morphologically similar, but differ in the relative length of their legs. Cataglyphis shows to be a very fast runner that is focussing on speed rather than on static stability. However the loss of control does not reduce precision of path integration. Meeting such opposing demands Cataglyphis is an ideal study object for the interaction of neural and mechanical systems.

7546 Th, 09:00-09:15 (P39) Frictional characteristics of earthworms in response to ground surfaces H. Fujie, M. Sato, S. Nakajima, K. Motai. Biomechanics Lab, Kogakuin University, Tokyo, Japan

Introduction: Earthworms perform peristaltic movement in which the tribolog- ical affections of their bodies to the ground are effectively utilized for forward motion [1]. To develop a novel and stable movement mechanism for micro- robots, the frictional behaviors of earthworms were determined in the present study. Methods: Earthworms, named "Eisenia fetida (Shima mimizu in Japanese)" were used to determine their frictional behaviors when they moved on several different surfaces with the controlled roughness from 6.4~tm to 172.1 ~tm. The surfaces were fixed to a 3-dimensional micro force plate [2] to measure the frictional force applied between the earthworms and the surfaces. The coefficients of static and dynamic friction were determined. Results: The coefficient of static friction of the earthworm was 11.0 at the roughness of 6.4, and was significantly decreased to 5.2 at the roughness of 172.1. Meanwhile, the coefficient of dynamic friction of the earthworm was 4.2 at the roughness of 6.4, and remained similar value at larger roughness. After the friction test, mucous liquid secreted from the earthworms was observed on the surface of the roughness of 6.4 ~tm. Microscopic observation indicated that earthworms had many spikes called "seta" around their bodies and that they controlled the length of the seta during the movement on an irregular surface. These results suggest that earthworms control their frictional characteristics in response to various surfaces that they contact by the use of their mucous liquid and seta. They also suggest that the mucous liquid plays more important role than the seta in controlling the frictional characteristics in response to relatively smooth surfaces.

References [1] Maeno et al. JSME(C) 1996; 62-60: 142-149. [2] Motai et al. JSME 2004; 12101: 131, 132. Abstract.

5184 Th, 09:15-09:30 (P39) Legs operate different during steady locomotion and escape in a wandering spider T. Weihmann, R. Blickhan. Dept. of Motion Science, Friedrich Schiller University, Jena, Germany

We analyzed kinematic and dynamic parameters of accelerated and steady locomotion in the large South American wandering spider Ancylometes bo- gotensis. The average body length was about 3.4cm, the average body weight was 3.2 g and the maximum velocities where around 1 m/s. The ground reaction forces where normalized on body size and stance duration. During steady slow and fast walking up to 1 m/s ground reaction forces are quite similar despite significant differences in kinematics. The horizontal force vectors are directed towards the COM. Only for the frontal leg pairs the decelerating forces disappear at higher velocities. At all velocities the contribution of this leg pair is small. In contrast, leg dynamics of exercises with accelerations strongly differs with respect to steady locomotion. Here horizontal force maxima are up to 15 times higher. During startles and jumps the distance between the COM and the substratum increases from 8 up to 25mm. The frontal leg pairs generate considerable horizontal forces in the direction of motion. At the same time almost no vertical forces where exerted. Primarily the fourth leg pair generates strong vertical forces. The horizontal force vectors point anteriorly.