Fluidic Artificial Muscles Research Articles

Investigation of the tradeoffs between tracking performance and energetics in heterogeneous variable recruitment fluidic artificial muscle bundles

In traditional hydraulic robotics, actuators must be sized for the highest possible load, resulting in significant energy losses when operating in lower force regimes. Variable recruitment fluidic artificial muscle (FAM) bundles offer a novel bio-inspired solution to this problem. Divided into individual MUs, each with its own control valve, a variable recruitment FAM bundle uses a switching control scheme to selectively bring MUs online according to load demand. To date, every dynamic variable recruitment study in the literature has considered homogeneous bundles containing MUs of equal size. However, natural mammalian muscle MUs are heterogeneous and primarily operate based on Henneman's size principle, which states that MUs are recruited from smallest to largest for a given task. Is it better for a FAM variable recruitment bundle to operate according to this principle, or are there other recruitment orders that result in better performance? What are the appropriate criteria for switching between recruitment states for these different recruitment orders? This paper seeks to answer these questions by performing two case studies exploring different bundle MU size distributions, analyzing the tradeoffs between tracking performance and energetics, and determining how these tradeoffs are affected by different MU recruitment order and recruitment state transition thresholds. The only difference between the two test cases is the overall force capacity (i.e. total size) of the bundle. For each test case, a Pareto frontier for different MU size distributions, recruitment orders, and recruitment state transition thresholds is constructed. The results show that there is a complex relationship between overall bundle size, MU size distributions, recruitment orders, and recruitment state transition thresholds corresponding to the best tradeoffs change along the Pareto frontier. Overall, these two case studies validate the use of Henneman's Size Principle as a variable recruitment strategy, but also demonstrate that it should not be the only variable recruitment method considered. They also motivate the need for a more complex variable recruitment scheme that dynamically changes the recruitment state transition threshold and recruitment order based on loading conditions and known system states, along with a co-design problem that optimizes total bundle size and MU size distribution.

Hardware-in-the-loop dynamic load emulation of robotic systems actuated by fluidic artificial muscles

Hardware-in-the-loop (HIL) testing is a popular control system testing method because it bridges the gap between modeling/simulation and experiments. Instead of designing a full hardware-based experiment to validate the results of a simulation, the plant hardware can be replaced with an emulator device that responds to exogenous inputs and effectively emulates the dynamic behavior of a system. This approach can be more cost-effective and modular, since the emulated plant system can be modeled in a simulation environment, implemented on a simplified piece of hardware and changed quickly without having to fabricate new parts. This paper develops the hardware and control scheme for a certain type of HIL device called a dynamic load emulator that consists of a 1-DOF linear hydraulic dynamometer equipped with in-line sensing to measure both its own position and the force exerted on it by a device-under-test. This measured force is passed to a real-time model of the emulated dynamic system. The model outputs the emulated system position, and a closed-loop controller is used to emulate this position. The emulator controller incorporates both model-based feedforward and standard feedback PI control. This paper characterizes the dynamometer-based dynamic load emulator and its controller, determining its hardware limitations and validating its capabilities when experiencing a force input from a linear spring with known parameters. Additionally, this paper demonstrates the ability of the emulator to represent the dynamics of a 1-DOF robotic joint when actuated by a pair of fluidic artificial muscles (FAMs). The primary contribution of this work is to allow for more comprehensive testing of FAM configurations, topologies, and controllers for a wide range of parameters, because the same hardware can be used to emulate multiple systems. As a result, this work will lead to more cost-effective, time-efficient, and energy-efficient designs of robotic systems and the FAMs used to actuate them.

Electrohydraulic System Analysis of Variable Recruitment Fluidic Artificial Muscle Bundles With Interaction Effects

Abstract This study investigates the system-level performance of variable recruitment (VR) fluidic artificial muscle (FAM) actuator bundles using a model that incorporates FAM interaction effects. A VR bundle combines multiple FAMs to act as one actuator in which the FAMs are sequentially recruited to increase overall efficiency. In a VR bundle, inactive/low-pressure FAMs are compressed beyond their free strains, exerting resistive forces opposing that of active FAMs. A recent model that captures this behavior is used to simulate sinusoidal contraction of a VR bundle with a hanging mass load. The implications of inter-FAM effects on the force–strain space of a VR bundle are discussed and a method of recruitment state transition required to track a sinusoid is proposed. The dynamics of the electrohydraulic subsystems are presented and used to evaluate its system efficiency and bandwidth limits. Three different electrohydraulic configurations are considered: (1) continuous motor operation with constant pump displacement, (2) intermittent motor operation with constant pump displacement, and (3) continuous motor operation with variable pump displacement. Simulation results show the superior bandwidth capabilities of VR bundles by demonstrating its ability to track sinusoids with amplitudes up to 16% strain at frequencies greater than 0.5 Hz, compared to that of a single equivalent cross section area motor unit (SEMU). In addition to increased bandwidth limit, system efficiencies averaged over a range of amplitudes show up to 170% increase when comparing a VR bundle using variable pump displacement to a SEMU using constant pump displacement.

Correction: Duan, E.; Bryant, M. Implications of Spatially Constrained Bipennate Topology on Fluidic Artificial Muscle Bundle Actuation. Actuators 2022, 11, 82

The authors wish to make the following corrections in Section 3 [...]

Bending Properties of an Extensile Fluidic Artificial Muscle.

Low stiffness, large stroke, and axial force capabilities make Extensile Fluidic Artificial Muscles (EFAMs) a feasible soft actuator for continuum soft robots. EFAMs can be used to construct soft actuated structures that feature large deformation and enable soft robots to access large effective workspaces. Although FAM axial properties have been well studied, their bending behavior is not well characterized in the literature. Static and dynamic bending properties of a cantilevered EFAM specimen were investigated over a pressure range of 5–100 psi. The static properties were then estimated using an Euler-Bernoulli beam model and discrete elastic rod models. The experiments provided data for the determination of bending stiffness, damping ratio, and natural frequency of the tested specimen. The bending stiffness and the damping ratio were found to change fourfold over the pressure range. Experimentally validated bending properties of the EFAM presented insights into structural and control considerations of soft robots. Future work will utilize the data and models obtained in this study to predict the behavior of an EFAM-actuated continuum robot carrying payloads.

Implications of Spatially Constrained Bipennate Topology on Fluidic Artificial Muscle Bundle Actuation

In this paper, we investigate the design of pennate topology fluidic artificial muscle bundles under spatial constraints. Soft fluidic actuators are of great interest to roboticists and engineers, due to their potential for inherent compliance and safe human–robot interaction. McKibben fluidic artificial muscles are an especially attractive type of soft fluidic actuator, due to their high force-to-weight ratio, inherent flexibility, inexpensive construction, and muscle-like force-contraction behavior. The examination of natural muscles has shown that those with pennate fiber topology can achieve higher output force per geometric cross-sectional area. Yet, this is not universally true for fluidic artificial muscle bundles, because the contraction and rotation behavior of individual actuator units (fibers) are both key factors contributing to situations where bipennate muscle topologies are advantageous, as compared to parallel muscle topologies. This paper analytically explores the implications of pennation angle on pennate fluidic artificial muscle bundle performance with spatial bounds. A method for muscle bundle parameterization as a function of desired bundle spatial envelope dimensions has been developed. An analysis of actuation performance metrics for bipennate and parallel topologies shows that bipennate artificial muscle bundles can be designed to amplify the muscle contraction, output force, stiffness, or work output capacity, as compared to a parallel bundle with the same envelope dimensions. In addition to quantifying the performance trade space associated with different pennate topologies, analyzing bundles with different fiber boundary conditions reveals how bipennate fluidic artificial muscle bundles can be designed for extensile motion and negative stiffness behaviors. This study, therefore, enables tailoring the muscle bundle parameters for custom compliant actuation applications.

Motor unit buckling in variable recruitment fluidic artificial muscle bundles: implications and mitigations

Fluidic artificial muscles (FAMs) are a popular actuation choice due to their compliant nature and high force-to-weight ratio. Variable recruitment is a bio-inspired actuation strategy in which multiple FAMs are combined into motor units that can be pressurized sequentially according to load demand. In a traditional ‘fixed-end’ variable recruitment FAM bundle, inactive units and activated units that are past free strain will compress and buckle outward, resulting in resistive forces that reduce overall bundle force output, increase spatial envelope, and reduce operational life. This paper investigates the use of inextensible tendons as a mitigation strategy for preventing resistive forces and outward buckling of inactive and submaximally activated motor units in a variable recruitment FAM bundle. A traditional analytical fixed-end variable recruitment FAM bundle model is modified to account for tendons, and the force–strain spaces of the two configurations are compared while keeping the overall bundle length constant. Actuation efficiency for the two configurations is compared for two different cases: one case in which the radii of all FAMs within the bundle are equivalent, and one case in which the bundles are sized to consume the same amount of working fluid volume at maximum contraction. Efficiency benefits can be found for either configuration for different locations within their shared force–strain space, so depending on the loading requirements, one configuration may be more efficient than the other. Additionally, a study is performed to quantify the increase in spatial envelope caused by the outward buckling of inactive or low-pressure motor units. It was found that at full activation of recruitment states 1, 2, and 3, the tendoned configuration has a significantly higher volumetric energy density than the fixed-end configuration, indicating that the tendoned configuration has more actuation potential for a given spatial envelope. Overall, the results show that using a resistive force mitigation strategy such as tendons can completely eliminate resistive forces, increase volumetric energy density, and increase system efficiency for certain loading cases. Thus, there is a compelling case to be made for the use of tendoned FAMs in variable recruitment bundles.

Design, analysis, and validation of an orderly recruitment valve for bio-inspired fluidic artificial muscles

Biological musculature employs variable recruitment of muscle fibers from smaller to larger units as the load increases. This orderly recruitment strategy has certain physiological advantages like minimizing fatigue and providing finer motor control. Recently fluidic artificial muscles (FAM) are gaining popularity as actuators due to their increased efficiency by employing bio-inspired recruitment strategies such as active variable recruitment (AVR). AVR systems use a multi-valve system (MVS) configuration to selectively recruit individual FAMs depending on the load. However, when using an MVS configuration, an increase in the number of motor units in a bundle corresponds to an increase in the number of valves in the system. This introduces greater complexity and weight. The objective of this paper is to propose, analyze, and demonstrate an orderly recruitment valve (ORV) concept that enables orderly recruitment of multiple FAMs in the system using a single valve. A mathematical model of an ORV-controlled FAM bundle is presented and validated by experiments performed on a proof-of-concept ORV experiment. The modeling is extended to explore a case study of a 1-DOF robot arm system consisting of an electrohydraulic pressurization system, ORV, and a FAM-actuated rotating arm plant and its dynamics are simulated to further demonstrate the capabilities of an ORV-controlled closed-loop system. An orderly recruitment strategy was implemented through a model-based feed forward controller. To benchmark the performance of the ORV, a conventional MVS with equivalent dynamics and controller was also implemented. Trajectory tracking simulations on both the systems revealed lower tracking error for the ORV controlled system compared to the MVS controlled system due to the unique cross-flow effects present in the ORV. However, the MVS, due to its independent and multiple valve setup, proved to be more adaptable for performance. For example, modifications to the recruitment thresholds of the MVS demonstrated improvement in tracking error, albeit with a sacrifice in efficiency. In the ORV, tracking performance remained insensitive to any variation in recruitment threshold. The results show that compared to the MVS, the ORV offers a simpler and more compact valving architecture at the expense of moderate losses in control flexibility and performance.

Modeling of Resistive Forces and Buckling Behavior in Variable Recruitment Fluidic Artificial Muscle Bundles

Fluidic artificial muscles (FAMs), also known as McKibben actuators, are a class of fiber-reinforced soft actuators that can be pneumatically or hydraulically pressurized to produce muscle-like contraction and force generation. When multiple FAMs are bundled together in parallel and selectively pressurized, they can act as a multi-chambered actuator with bioinspired variable recruitment capability. The variable recruitment bundle consists of motor units (MUs)—groups of one of more FAMs—that are independently pressurized depending on the force demand, similar to how groups of muscle fibers are sequentially recruited in biological muscles. As the active FAMs contract, the inactive/low-pressure units are compressed, causing them to buckle outward, which increases the spatial envelope of the actuator. Additionally, a FAM compressed past its individual free strain applies a force that opposes the overall force output of active FAMs. In this paper, we propose a model to quantify this resistive force observed in inactive and low-pressure FAMs and study its implications on the performance of a variable recruitment bundle. The resistive force behavior is divided into post-buckling and post-collapse regions and a piecewise model is devised. An empirically-based correction method is proposed to improve the model to fit experimental data. Analysis of a bundle with resistive effects reveals a phenomenon, unique to variable recruitment bundles, defined as free strain gradient reversal.

Influence of hydraulic versus pneumatic working fluids on quasi-static force response of fluidic artificial muscles

There is currently an increased interest in extending the application of McKibben actuators beyond gaseous environments and working fluids. Extensive efforts have already established force response models for pneumatically driven McKibben actuators that are independent of working fluid material properties, however, the model’s independence to working fluid choice has yet to be validated with empirical evidence. This paper experimentally investigates the effect of working fluid type on the quasi-static pressure dependent force-contraction response of McKibben actuators. Using either air or water as the working fluid, characteristic isobaric force-contraction response curves are compared for both large and small-scale McKibben actuators. To ensure truly isobaric force-contraction characterizations, hydraulic and pneumatic pressure systems were developed to provide precise and accurate control of pressure. Experimental testing proved using air or water as the working fluid resulted in nearly identical isobaric force-contraction response curves, demonstrating that McKibben actuator’s quasi-static force response is independent of working fluid choice. This study establishes that the applicability of existing force response models for pneumatic McKibben actuators can be extended to any practical liquids or gases.

Characterization and Analysis of Extensile Fluidic Artificial Muscles

Extensile fluidic artificial muscles (EFAMs) are soft actuators known for their large ranges of extension, low weight, and blocked forces comparable to those of pneumatic cylinders. EFAMs have yet to be studied in a way that thoroughly focuses on their manufacturing, experimental characterization, and modeling. A fabrication method was developed for production of two EFAMs. The quasi-static axial force response of EFAMs to varying displacement was measured by testing two specimens under isobaric conditions over a pressure range of 103.4–517.1 kPa (15–75 psi) with 103.4 kPa (15 psi) increments. The muscles were characterized by a blocked force of 280 N and a maximum stroke of 98% at 517.1 kPa (75 psi). A force-balance model was used to analyze EFAM response. Prior work employing the force-balance approach used hyper-elastic constitutive models based on polynomial expressions. In this study, these models are validated for EFAMs, and new constitutive models are proposed that better represent the measured stress values of rubber as a function of strain. These constitutive models are compared in terms of accuracy when estimating pressure-dependent stress–strain relationships of the bladder material. The analysis demonstrates that the new hyper-elastic stress models have an error 5% smaller than models previously employed for EFAMs for the same number of coefficients. Finally, the analysis suggests that the new stress functions have smaller errors than the polynomial stress model with the same number of coefficients, guarantee material stability, and are more conservative about the stress values for strains outside of the testing range.

Theoretical Dynamic Modeling and Validation of Braided Pneumatic Artificial Muscles

Abstract The high force-to-weight ratios of braided fluidic artificial muscles (AMs) are ideal for human scale and mobile robot applications. Prior modeling efforts focus on the theoretical static characteristics or empirical dynamic models of these actuators when pressurized. This paper develops a comprehensive high fidelity theoretical dynamic model based on first principles for braided pneumatic AMs and presents experimental validation. A novel theoretical model for the nonlinear stiffness is derived to describe the pressure–displacement behavior. The stiffness model, together with friction, damping, and inertia models, forms an equation of motion (EOM) for braided pneumatic AMs. The EOM is coupled with first-order servopneumatic pressure dynamics, resulting in a third-order system model. System model simulations are compared to experimental results of prototypes with nine different geometries. On average, the system model is able to predict the quasi-static displacement within 7% and the dynamic response within 11%. The theoretical model is also benchmarked against a high fidelity curve fit method, with the empirical method showing a 2% improvement in only quasi-static scenarios. The model promises to be useful for mechanical system and model-based control designs.

An artificial muscle driven by decomposition of gas solution

This note develops an artificial muscle driven by the decomposition of gas solution. It mainly consists of a silicone tube, a resistive wire, a braided mesh shell, and an amount of gas solution which can be decomposed reversibly by temperature variation. It can be operated directly by supplying the resistive wire with controllable electrical power, which is different from traditional fluidic artificial muscles requiring external pressure generating devices. To verify the effectiveness of the artificial muscle, prototypes are fabricated and the process is presented in detail. Ammonia solution is employed in the prototypes due to its high solubility and low decomposition temperature. Experiments under both isometric and isotonic contractions are implemented and the results show that the prototypes can achieve good mechanical performance.

Modeling and analysis of hydraulic piston actuation of McKibben fluidic artificial muscles for hand rehabilitation

Soft robotic actuators are well-suited for interactions with the human body, particularly in rehabilitation applications. The fluidic artificial muscle (FAM), specifically the McKibben FAM, is a type of soft robotic actuator that can be driven either pneumatically or hydraulically, and has potential for use in rehabilitation devices. The force applied by a FAM is well-described by a variety of models, the most common of which is based on the virtual work principle. However, the use of a piston assembly as a hydraulic power source for activation of FAMs has not previously been modeled in detail. This article presents a FAM designed to address the specific needs of a hand rehabilitation device. A syringe pump test bed is used to find and validate a novel volume–strain relationship. The volume–strain relationship remains constant with the coupled piston–FAM system, regardless of load. This confirms a bivariate approach to FAM control which is particularly beneficial in the exoskeleton application as the load varies throughout use. A novel, fixed-end cylindrical model is found to predict the strain of the FAM, given a volume input, regardless of load. For the FAMs tested in this work, the fixed-end cylindrical model improves strain prediction seven-fold when compared with traditional models.

Design and analysis of electrohydraulic pressure systems for variable recruitment in fluidic artificial muscles

This paper investigates the energetics and performance of an electrohydraulic power system with variable recruitment fluidic artificial muscle (FAM) actuators. A coupled dynamic model of the system is developed and applied to study the implications of hydraulic power system architecture for both variable recruitment actuator bundles and equivalent single-muscle actuators. This analysis extends previous FAM literature by considering both actuator recruitment methodology as well as the complete electromechanical circuit and the interactions of these two subsystems. Specifically, for both single-muscle actuators and variable recruitment muscle bundles, hydraulic architectures with a continuously-operating motor and pump are compared with a system in which the motor is intermittently shut down and restarted based on accumulator pressure. The results reveal that variable recruitment offers bandwidth advantages over the single equivalent actuator regardless of the hydraulic power architecture that is selected. However, use of the intermittently-operating motor and pump system allowed the variable recruitment system to achieve efficiency advantages over the other configurations considered. A steady-state analytic solution for the operating envelopes of the variable recruitment and single-muscle systems, including force limits and flow rate limits, was also developed and used to investigate effects of pump displacement on system bandwidth and stroke. The results of these analyses provide tools for the selection of actuator configuration, system architecture, and component design in FAM-actuated electrohydraulic robots.

Bioinspired passive variable recruitment of fluidic artificial muscles

This article presents a novel, passive approach to creating variable actuator recruitment in bundles of fluidic artificial muscles. The passive recruitment control approach is inspired by the functionality of mammalian muscle tissues, in which a single activation signal from the nervous system sequentially triggers contraction of progressively larger actuation elements until the required force is generated. Biologically, this behavior is encoded by differences in electrical resistance properties between smaller and larger muscle-fiber groups. The approach presented here produces analogous behavior using a uniform applied pressure to all fluidic artificial muscles while creating differential pressure responses and threshold pressures among the fluidic artificial muscles via tailored bladder elasticity parameters. A model for using elastic bladder stiffness to control an artificial muscle bundle with a single valve is explored and used to compare a bundle of fluidic artificial muscles with both low and high threshold pressure units to a single fluidic artificial muscle of equivalent displacement and force capability. The results of this analysis indicate the efficacy of using this control method; it is advantageous in cases where a wide range of displacements and forces are necessary and can increase efficiency when the system primarily operates in a low-force regime but requires occasional bursts of high-force capability.

Fluid-driven origami-inspired artificial muscles

Artificial muscles hold promise for safe and powerful actuation for myriad common machines and robots. However, the design, fabrication, and implementation of artificial muscles are often limited by their material costs, operating principle, scalability, and single-degree-of-freedom contractile actuation motions. Here we propose an architecture for fluid-driven origami-inspired artificial muscles. This concept requires only a compressible skeleton, a flexible skin, and a fluid medium. A mechanical model is developed to explain the interaction of the three components. A fabrication method is introduced to rapidly manufacture low-cost artificial muscles using various materials and at multiple scales. The artificial muscles can be programed to achieve multiaxial motions including contraction, bending, and torsion. These motions can be aggregated into systems with multiple degrees of freedom, which are able to produce controllable motions at different rates. Our artificial muscles can be driven by fluids at negative pressures (relative to ambient). This feature makes actuation safer than most other fluidic artificial muscles that operate with positive pressures. Experiments reveal that these muscles can contract over 90% of their initial lengths, generate stresses of ∼600 kPa, and produce peak power densities over 2 kW/kg-all equal to, or in excess of, natural muscle. This architecture for artificial muscles opens the door to rapid design and low-cost fabrication of actuation systems for numerous applications at multiple scales, ranging from miniature medical devices to wearable robotic exoskeletons to large deployable structures for space exploration.

Modeling and analysis of a meso-hydraulic climbing robot with artificial muscle actuation

This paper presents a fully coupled electro-hydraulic model of a bio-inspired climbing robot actuated by fluidic artificial muscles (FAMs). This analysis expands upon previous FAM literature by considering not only the force and contraction characteristics of the actuator, but the complete hydraulic and electromechanical circuits as well as the dynamics of the climbing robot. This analysis allows modeling of the time-varying applied pressure, electrical current, and actuator contraction for accurate prediction of the robot motion, energy consumption, and mechanical work output. The developed model is first validated against mechanical and electrical data collected from a proof-of-concept prototype robot. The model is then employed to study the system-level sensitivities of the robot locomotion efficiency and average climbing speed to several design and operating parameters. The results of this analysis demonstrate that considering only the transduction efficiency of the FAM actuators is insufficient to maximize the efficiency of the complete robot, and that a holistic approach can lead to significant improvements in performance.

Modeling and experiments on the drive characteristics of high-strength water hydraulic artificial muscles

Fluidic artificial muscles are popular in robotics and function as biomimetic actuators. Their pneumatic version has been widely investigated. A novel water hydraulic artificial muscle (WHAM) with high strength is developed in this study. WHAMs can be applied to underwater manipulators widely used in ocean development because of their environment-friendly characteristics, high force-to-weight ratio, and good bio-imitability. Therefore, the strength of WHAMs has been improved to fit the requirements of underwater environments and the work pressure of water hydraulic components. However, understanding the mechanical behaviors of WHAMs is necessary because WHAMs use work media and pressure control that are different from those used by pneumatic artificial muscles. This paper presents the static and dynamic characteristics of the WHAM system, including the water hydraulic pressure control circuit. A test system is designed and built to analyze the drive characteristics of the developed WHAM. The theoretical relationships among the amount of contraction, pressure, and output drawing force of the WHAM are tested and verified. A linearized transfer function is proposed, and the dynamic characteristics of the WHAM are investigated through simulation and inertia load experiments. Simulation results agree with the experimental results and show that the proposed model can be applied to the control of WHAM actuators.

Bio-inspired online variable recruitment control of fluidic artificial muscles

This paper details the creation of a hybrid variable recruitment control scheme for fluidic artificial muscle (FAM) actuators with an emphasis on maximizing system efficiency and switching control performance. Variable recruitment is the process of altering a system’s active number of actuators, allowing operation in distinct force regimes. Previously, FAM variable recruitment was only quantified with offline, manual valve switching; this study addresses the creation and characterization of novel, on-line FAM switching control algorithms. The bio-inspired algorithms are implemented in conjunction with a PID and model-based controller, and applied to a simulated plant model. Variable recruitment transition effects and chatter rejection are explored via a sensitivity analysis, allowing a system designer to weigh tradeoffs in actuator modeling, algorithm choice, and necessary hardware. Variable recruitment is further developed through simulation of a robotic arm tracking a variety of spline position inputs, requiring several levels of actuator recruitment. Switching controller performance is quantified and compared with baseline systems lacking variable recruitment. The work extends current variable recruitment knowledge by creating novel online variable recruitment control schemes, and exploring how online actuator recruitment affects system efficiency and control performance. Key topics associated with implementing a variable recruitment scheme, including the effects of modeling inaccuracies, hardware considerations, and switching transition concerns are also addressed.