Kinematic and Kinetic Analysis
of a Biomechanical Model of Rat Hind Limb
with Biarticular Muscles
Kaiyu Deng 1(B) , Nicholas S. Szczecinski 1 , Alexander J. Hunt 2 , Hillel J. Chiel 1 ,
and Roger D. Quinn 1
1 Case Western Reserve University, Cleveland, OH, USA
kxd194@case.edu
2 Portland State University, Portland, OR, USA
Abstract. This work presents a biomechanical model of rat hind limbs with biarticular muscles and includes kinematic and kinetic analyses. Our previous model
only possessed antagonistic muscle pairs (extensor and flexor) to actuate each
joint in the sagittal plane. In this model, we expanded the number of muscles
in each limb from 6 to 8, including 3 biarticular muscles: BFP (biceps femoris
Posterior), RF (Rectus femoris), and GA (Gastrocnemii). The knee flexor muscle was removed from the previous model. We also developed a new method to
calculate the muscle parameter values, including the shape of the length-tension
curve, which gives the muscle models more biomimetic response properties. In
order to predict muscle stretch and the relationship between passive tension and
step-phases during walking, we formulate the inverse kinematics for the limbs and
muscles.
Keywords: Rat · Biarticular muscles · Synthetic nervous system · Inverse
kinematic analysis
1 Introduction
Roboticists wish to capture the agility of animals in their robots and biologists wish
to better understand how animals move. Unlike robot designs, which are usually minimalist, mammalian legged locomotion actuation is often redundant and highly complex.
Biologists and roboticists alike often construct extremely simplified biomechanical models that capture only certain components of animal walking. This eases analysis, but
makes the model less animal-like. However, a model that replicated all details of animal anatomy may not be appropriate, because it would be too complex to analyze,
very difficult to simulate, and there are still many unknowns about how animals control locomotion. Therefore, balancing simplicity and fidelity is the key to designing an
informative biomimetic robot or walking simulation model.
Previously, we improved the synthetic nervous system (SNS) of a biomechanical rat
hindlimb model by implementing a two-layer neural system with both rhythm generator
and pattern formation networks [1]. This model successfully reproduced rat nominal
© Springer Nature Switzerland AG 2020
V. Vouloutsi et al. (Eds.): Living Machines 2020, LNAI 12413, pp. 55–67, 2020.
https://doi.org/10.1007/978-3-030-64313-3_7
of a Biomechanical Model of Rat Hind Limb
with Biarticular Muscles
Kaiyu Deng 1(B) , Nicholas S. Szczecinski 1 , Alexander J. Hunt 2 , Hillel J. Chiel 1 ,
and Roger D. Quinn 1
1 Case Western Reserve University, Cleveland, OH, USA
kxd194@case.edu
2 Portland State University, Portland, OR, USA
Abstract. This work presents a biomechanical model of rat hind limbs with biarticular muscles and includes kinematic and kinetic analyses. Our previous model
only possessed antagonistic muscle pairs (extensor and flexor) to actuate each
joint in the sagittal plane. In this model, we expanded the number of muscles
in each limb from 6 to 8, including 3 biarticular muscles: BFP (biceps femoris
Posterior), RF (Rectus femoris), and GA (Gastrocnemii). The knee flexor muscle was removed from the previous model. We also developed a new method to
calculate the muscle parameter values, including the shape of the length-tension
curve, which gives the muscle models more biomimetic response properties. In
order to predict muscle stretch and the relationship between passive tension and
step-phases during walking, we formulate the inverse kinematics for the limbs and
muscles.
Keywords: Rat · Biarticular muscles · Synthetic nervous system · Inverse
kinematic analysis
1 Introduction
Roboticists wish to capture the agility of animals in their robots and biologists wish
to better understand how animals move. Unlike robot designs, which are usually minimalist, mammalian legged locomotion actuation is often redundant and highly complex.
Biologists and roboticists alike often construct extremely simplified biomechanical models that capture only certain components of animal walking. This eases analysis, but
makes the model less animal-like. However, a model that replicated all details of animal anatomy may not be appropriate, because it would be too complex to analyze,
very difficult to simulate, and there are still many unknowns about how animals control locomotion. Therefore, balancing simplicity and fidelity is the key to designing an
informative biomimetic robot or walking simulation model.
Previously, we improved the synthetic nervous system (SNS) of a biomechanical rat
hindlimb model by implementing a two-layer neural system with both rhythm generator
and pattern formation networks [1]. This model successfully reproduced rat nominal
© Springer Nature Switzerland AG 2020
V. Vouloutsi et al. (Eds.): Living Machines 2020, LNAI 12413, pp. 55–67, 2020.
https://doi.org/10.1007/978-3-030-64313-3_7