Abstract Over the past decade, shape memory alloy (SMA) in the form of wires and cables have been extensively studied for various structural engineering applications. There are numerous application areas where pure compression (or coupled with tension) is the primary load bearing scenario, which requires larger size SMA bars. However, the compression behavior of SMA bars is not well known, and little is reported in the literature. In that perspective, this paper presents an experimental study on large diameter superelastic Nickel-Titanium (NiTi) bars subjected to a cyclic compression load. A total of nine SMA bars having slenderness ratios ranging from 60 to 90 were tested. Hysteretic stress-strain responses are plotted and critical buckling load, energy dissipation and residual strain of SMA bars with different slenderness ratios are presented.

There has been increasing focus in the area of in-situ structural health monitoring since the advent of embedded nano-composites. This experimental research investigates the structural health monitoring abilities of polymer bonded energetics embedded with a uniformly dispersed but randomly oriented carbon nanotube (CNT) sensing network within the polymer binder. A common formulation of the recent solid propellants consists of ammonium perchlorate (oxidizer) and aluminum powder (combustive fuel)-often shaped using a polymer binder, rather than the older techniques of power pressing. Since this study focuses on the structural health of the material and not its thermal properties, monoclinic sugar crystals were used as a substitute for ammonium perchlorate as it has very similar mechanical properties and is much safer in terms of material handling. Thus, a combination of sugar crystals and aluminum powder bound by a Polydimethylsiloxane (PDMS) binder is fabricated in varying concentrations to simulate actual solid rocket propellants. The main focus of this study lies in characterizing the mechanical and electrical properties of the CNT embedded energetic material through subjecting it under mechanical loads; followed by a detailed observation and study of the real time electro-mechanical response under tension and compression. The addition of carbon nanotubes to the polymer binder, thus translates to a real time sensing technique for detection of multi-scale damage in polymer bonded energetics. The results of this study aim to establish a proof of concept for CNT embedded structural health monitoring systems.

Additive manufacturing is an emerging method to produce customized parts with functional materials without big investments. As one of the common additive manufacturing methods, fused deposition modeling (FDM) uses thermoplastic-based feedstock. It has been recently adapted to fabricate composite materials too. Acrylonitrile butadiene styrene (ABS) is the most widely used material as FDM feedstock. However, it is an electrically insulating polymer. Carbon Nanotubes (CNTs) on the other hand are highly conductive. They are attractive fillers because of their high aspect ratio, and excellent mechanical and physical properties. Therefore, a nanocomposite of these two materials can give an electrically conductive material that is potentially compatible with FDM printing. This work focuses on the investigation of the relationships between the FDM process parameters and the electrical conductivity of the printed ABS/CNT nanocomposites. Nanocomposite filaments with CNT contents up to 10wt% were produced using a twin-screw extruder followed by 3D printing using FDM method. The starting material was pellets from a masterbatch containing 15 wt% CNT. Compression-molded samples of ABS/CNT were also prepared as the bulk baselines. The effects of CNT content and nozzle size on the through-layer and in-layer electrical conductivity of the printed nanocomposites were analyzed. Overall, a higher percolation threshold was observed in the printed samples, compared to that of the compression-molded counterparts. This resulted in the conductivity of the printed samples that is at least one order of magnitude lower. Moreover, at CNT contents up to 5 wt%, the in-layer conductivity of the printed samples was almost two orders of magnitudes higher than that in the through-layer direction. In ABS/3 wt% CNT samples, the through-layer conductivity continuously decreased as the nozzle diameter was decreased from 0.8 mm to 0.35 mm. These variations in the electrical conductivity were explained in terms of the CNT alignment, caused by the extrusion process during the print, quality of interlayer bonding during deposition, and the voids created due to the discrete nature of the printing process.

Localized reinforcement of composites employed to manufacture parts for the transport industries is making possible the lightweighting of components that have a much sought-after effect in the reduction of CO 2 and NO x emissions. However, its realization, through the removing of mass where it is not required and reinforcement added to areas more prone to stress from working loads, relies on the development of novel manufacturing processes that can create structures whose performance is on a par with their solid counterparts, but at a fraction of the weight and at an affordable production cost. In this work we exploit the use of a very weak and safe magnetic field to control the location and orientation of functionalized discontinuous carbon fibers within a polymeric structural (polyurethane) foam to create performance-optimized composites. Two wet-chemistry methods (i.e. in-situ precipitation-deposition and amine-co-adjuvated electrodeposition of magnetite) to transform commercial carbon fiber into a magnetically active form were explored. The resulting fibers were analyzed and characterized through a set of physico-chemical tests. The functionalized fibers were then embedded at 3 different %vol contents in the polymeric matrix at given locations and with a desired alignment. Their mechanical performance (incl. compression, tension) was assessed and benchmarked against both a similar %volumetric content but non-functionalized-reinforcement (i.e. randomly distributed) composites and to non-reinforced matrices. In the two sets of reinforced composites (random and aligned) there is a positive correlation between stiffness, yield strength and strain with increasing %vol content. Both sets outperformed the non-reinforced matrix, demonstrating good fiber adhesion within the matrix and successful load transfer from matrix to fiber. The magnetically aligned composites generally outperformed the non-functionalized ones in terms of stiffness and strength at yield.

In the new data intensive world, predictive maintenance has become a central issue for the modern industrial plants. Monitoring of electric machinery is one of the most important challenges in predictive maintenance. Adaptive manufacturing processes/plants may be possible through the monitored conditions. In this respect, several attempts have been made to utilize deep learning algorithms for rotating machinery fault detection and diagnosis. Among them, deep autoencoders are very popular, because of their denoising effect. They are also implemented in electric machinery fault diagnostics in order to obtain lower order representation of signals. However, none of these efforts regard the autoencoders as compression units. Bearing in mind that spectra of vibration and current signals that are collected from electric machinery are critical instruments for detection and diagnosis of their faults, we propose that deep stacked autoencoder can be utilized as spectrum compression units. The performance of the proposed strategy are assessed using a bearing data set in three ways: (1)Rule-based classifiers are implemented on raw and compressed-decompressed spectrum and their performance are compared. (2) It is shown that the several machine learning classifiers such as support vector machines, artificial neural networks and k-nearest neighbour classifiers on compressed-decompressed spectrum achieves the performance of them on raw data. (3) A multi-layer perceptron (MLP) classifier is implemented on the low dimensional representation and it is demonstrated that the strategy of employing the same autoencoder as pretraining of feature extraction module cannot outperform the performance of this MLP classifier.

This paper presents a method to measure gripping force of a bipedal wall-climbing robot (WCR) with spiny toe pads. The spiny toe pad is designed based on inspiration of an insect’s tarsal system. Each foot of the robot consists of a pair of opposed linear spiny arrays. The foot employs a pulley system to actuate the arrays via four pairs of tension and compression springs. Two Hall effect sensors are embedded into the robot feet to sense the gripping force by detecting the linear deformation of the springs. The two Hall effect sensors are calibrated and the relationship between the voltage signal output of the sensors and displacement is established before measuring gripping force. Then the consistency and accuracy of Hall effect sensor measurement method are verified by comparing with a commercial force sensor. A horizontal crawling test of the WCR is carried out and the gripping force verse time when the WCR moves. The experimental results show that the measured force history is in accordance with the actual movement states.

Energy harvesting from low frequency cyclic motion is possible in a variety of applications, but generating power with piezoelectric stacks at low, off-resonance frequencies is challenging. In this study, Compliant Layer Adaptive Composite Stacks (CLACS) were investigated as a toughened piezoelectric generator to increase efficiency at low frequencies and match the compliance of many commercial devices. CLACS were manufactured with PZT discs, interdigitated epoxy layers of varying thicknesses, and encapsulated in epoxy. Energy production of each CLACS type as a function of compliant layer thickness was characterized. Power amplification of CLACS was modeled assuming discs remain planar, volume of epoxy was conserved, and total epoxy deformations were small. Shear lag theory demonstrated increases in positive in-plane strains induced by external through-thickness compression. This amplified sensitivity of the entire stack to through-thickness compressions, substantially increases power generation capability. Experimental data showed that increases in compliant layer thickness resulted in increased power generation in all loading conditions. The shear lag structural mechanics model showed good correlation with theoretical predictions, assuming small deformation of the compliant layer. In addition to reducing composite stiffness, the CLACS generated 61% more power than conventional stack actuators with the same PZT volume via lateral strain amplification effects.

Elastocaloric cooling is a novel environment-friendly alternative to vapor compression-based cooling systems. This solid-state cooling technology uses NiTi shape memory alloys (SMAs) as cooling medium. SMAs are well known for lightweight actuator systems and biomedical applications, but in addition, these alloys exhibit excellent cooling properties. Due to the high latent heats activated by mechanical loading/unloading, large temperature changes can be generated in the material. Accompanied by a small required work input, this also leads to a high coefficient of performance superior to vapor compression-based systems. In order to access the potential of these alloys, the development of suitable thermodynamic cooling cycles and an efficient system design are required. This paper presents a model-based design process of an elastocaloric air-cooling device. The device is divided into two parts, a mechanical system for continuously loading and unloading of multiple SMA wire bundles by a rotary motor and a heat transfer system. The heat transfer system enables an efficient heat exchange between the heat source and the SMA wires as well as between the SMA wires and the environment. The device operates without any additional heat transfer medium and cools the heat source directly, which is an advantage in comparison to conventional cooling systems. The design of this complex device in an efficient manner requires a model approach, capable of predicting the system parameters cooling power, mechanical work and coefficient of performance under various operating conditions. The developed model consists of a computationally efficient, thermo-mechanically coupled and energy based SMA model, a model of the system kinematics and a heat transfer model. With this approach, the complete cooling system can be simulated, and the required number of SMA wires as well as the mechanical power can be predicted in order to meet the system requirements. Based on the simulation results a first prototype of the elastocaloric cooling system is realized.

Strain in solid materials under external loads cannot be visualized until they reach a high value or failure occurs; and the common measuring method of using strain sensors is effective but limited to wiring or power supply. In this study, we introduce a new concept of self-sensing solid materials by designing thin surface circular delamination regions on a material body to sense and predict the elastic global strain through controlled elastic local buckling. Delamination buckling is an undesirable failure occurrence in laminated composites under compression. However, it can translate imperceptible small global strains on the main material body to a visible large deformation in the surface of the delaminated region due to buckling. We analytically studied the buckling and post-buckling response of a clamped circular thin plate with unilateral constraint using an energy method to obtain the critical buckling loads, the buckling configurations, and the center out-of-plane displacement under uniaxial and biaxial loading conditions. The results show that for a given buckling configuration in the local region, the global strain condition of the main material body can be predicted. The study thus explores and proves a feasible way to design self-sensing materials through controlled delamination buckling.

Limitations on neuron firing rates restrict the frequency bandwidth of many biological sensory systems. The cochlea overcomes these limitations to hear high frequency sounds through its tonotopic structure and non-synchronous sampling. The cochlea’s tapering basilar membrane serves as a filter bank decomposing an applied sound into its frequency components. Auditory hair cells produce neural impulses at the peaks in the local basilar membrane oscillations resulting in an event-driven, sub-Nyquist rate sampling strategy. These two effects extend the human hearing range to about 20 kHz despite maximum neuron firing rates of just a few hundred hertz. Inspired by this, this paper presents a concept demonstration of an accelerometer and signal compression strategy for high-rate impact events using a similar filter bank approach. A series of clamped-clamped beams will serve as analog analysis filters much like the cochlea’s basilar membrane. This paper focuses on the design and simulation of such a beam array and how the natural frequencies and damping ratios of the beams’ first modes affect measuring a broadband impact excitation.

The present study concerns with the performance of a skid landing gear (SLG) system of a rotorcraft impacting the ground at a vertical sink rate of 5.0 m/s. The impact attitude is per chapter 527 of the Airworthiness Manual (AWM) of Transport Canada Civil Aviation and FAR Part 27 of the U.S. Federal Aviation Regulation. A single degree of freedom helicopter model is investigated under two rotor lift factors 0.67 and 1.0. Three Configurations are evaluated: a) A conventional SLG; b) SLG equipped with a passive viscous damper and c) SLG incorporated with a magnetorheological energy absorber. The non-dimensional solutions of the helicopter model show that the passive damper system could reduce the maximum acceleration experienced by the helicopter occupants by 21% and 19.8% in comparison to the undamped system for the above rotor lift factors, respectively. However, the passive damper fails to constrain the non-dimensional energy absorption stroke of the damper within the given 18 cm maximum stroke and a bottoming out of the damper piston was noticed. Therefore, the alternative and successful choice was to employ a magnetorheological energy absorber (MREA). To improve the MREA controllability and to resettle the payload with no oscillations, i.e. in one cycle, two different Bingham numbers for compression stroke and rebound stroke were defined in the non-dimensional solution. Several simulations were conducted for different values of Bingham numbers. Among these numerical simulation results, the solution that implemented the optimum Bingham numbers was found to be the only one feasible solution. In this case the MREA with optimum Bingham number for compression could utilize the full energy absorption stroke to attain soft landing. In the rebound stroke, the generated optimal on-state damping force successfully controls the bounce of the payload until the payload settles down to its original equilibrium position with no oscillations.

Postbuckling response, long considered mainly as a failure limit state is gaining increased interest for smart applications, such as energy harvesting, frequency tuning, sensing, actuation, etc. Cylindrical shells have received less attention as structural form to harness elastic instabilities due to their increased modeling complexity and high imperfection sensitivity. Yet, preliminary experimental and computational evidence indicates that the elastic postbuckling response of cylindrical shells can be controlled and potentially managed. Further, cylindrical shells offer desirable features for the design of mechanical devices and adaptive structures that other forms cannot attain without additional external constraints. This paper presents a study on tailoring the elastic postbuckling response of thin-walled cylindrical shells under compression by means of non-uniform wall stiffness distributions. The pattern of stiffness distribution was designed by discretizing the shell surface into cells and thickening selected cells with respect to a baseline uniform wall thickness. Diverse patterns were characterized in the way of how they affect the postbuckling response through numerical simulations using the finite element method. Results show that the elastic postbuckling response can be tailored into three response types: softening, sustaining, and stiffening; and that number, sequence/time and location/space of localized buckling events can be designed. This work provides new knowledge on the means to design the cylindrical shells with controlled elastic postbuckling behavior for applications in smart materials, mechanical devices, and adaptive structures.

The ability to functionalize droplet interface bilayers (DIBs) with the MscL channel and its mutants has been demonstrated. In previous work, the V23T gain of function mutant of MscL produced consistent activation when harmonic axial compressions were applied to the aqueous droplets supporting the lipid bilayer, where the channels settle. The deformation of the droplets results, at maximum compression, in an increase in surface area, and thus an increase in tension at the water-lipid-oil interface. This increase in monolayer tension was found to be the product of the relative change in surface area of each of the droplets and the compressibility modulus of the DPhPC monolayer (∼120 mN/m). The tension increase at the water-lipid-oil interface almost doubles to make up the increase in tension in the bilayer interface, resulting in activation of the incorporated MscL channels. However, it was found that the application of a relatively high transmembrane potential (∼100 mV), from an external power source, is a requirement for the activation of the V23T-MscL channels. Here, we investigate and analyze the impact of transmembrane potential on the activity of MscL channels in both a droplet interface bilayer system and E. coli spheroplast via patch-clamp. We demonstrate that the channels became more susceptible to gating upon the application of a negative potential, compared to when a positive potential is applied, proving their sensitivity to voltage polarity.

Growing demands for Total Knee Arthroplasty (TKA) and also total knee revision surgery combined with the cost, risk, and complication of the surgery have led to numerus attempts to improve surgical techniques, implant design, and postoperative orthopedic therapies. Although abundant information about knee function and reaction forces and moments have been provided by researchers through biomechanical models, cadaver testing, in-vitro testing, and limited in-vivo measurements, rigorous real-time in-vivo data from knee implants is still required to improve the performance of TKAs. Instrumented knee implants using piezoelectric (PZT) transducers have promising potential to satisfy clinical needs in terms of continuous in-vivo data acquisition, self-powered operation, and retention of prevalent implant design, which can ultimately lead to improved patient satisfaction. In this study, a simplified Ultra High Molecular Weight (UHMW) Polyethylene TKA bearing geometry with an embedded PZT on the bottom surface is proposed and investigated to analyze sensing and power harvesting, and longevity of the conceptual design. As a result, this work is separated into two distinct sections. The first part provides an evaluation study on the performance of the design in terms of output voltage and power using both simulations and experimental tests. Finite element analysis (FEA) is employed to model the stress-strain behavior of the system and to develop effective force reaction on the PZT transducer. An analytical model is used to describe the electromechanical behavior of the PZT transducer under the effective force predicted by FEA, and the output voltage and power of the system are simulated. Furthermore, results obtained from modeling are validated through experimental compression testing using simulated gait conditions. Embedding a PZT element in the knee bearing may cause changes in stress distribution in UHMW and as a result the variation in the fatigue life of the bearings with encapsulated PZTs is considered as a remarkable factor to investigate. Therefore, in the second part of the work, a parametric study on the effect of dimensional parameters on the longevity and electromechanical performance of the design is performed. High cycle stress life of the polyethylene component with embedded PZT transducer as well as transferred force to the PZT and generated voltage under periodic knee load are studied. . The diameter and depth of the pocket machined in the UHMW bearing, the thickness ratio of the PZT element to the UHMW component, and modification of the contact edges inside the PZT pocket and PZT are considered as effective geometrical parameters on the fatigue life of the UHMW bearing and are studied individually. Two designs are investigated; the initial design with sharp corners and a revised design with filleted corners. The results show a significant fatigue life improvement by adding a fillet radius modification on the sharp corners of the UHMW and PZT components accompanied by a slight reduction in output voltage. The effect of pocket diameter is dependent on the geometry and for the initial design the fatigue life and output voltage increase when diameter increases. For the revised design, fatigue life decreases for large fillet radii and increases for small fillet radii and converge as diameter is increased, whereas the output voltage slightly increases with large pocket diameters. Pocket depth has a significant reverse effect on fatigue life and output voltage of the PZT, such that a 0.05 mm deeper pocket results in no force transfer and no voltage but improved fatigue life.

This paper presents a structural concept that exploits elastic instabilities in novel periodic lattice structures for shape adaptation purposes. The nonlinear behaviour resulting from the occurrence of local buckling is utilised to achieve significant variations in the global structural response of the lattice. For the proposed structural concept, a unit cell is identified and utilised to investigate the mechanical characteristics for the load cases of uniaxial compression, shear, and rotation, conducting nonlinear finite element simulations. The results of the unit cell characterization are compared to the mechanical response of lattice structures under equivalent loading and convergence is achieved for all considered load cases. This paper therefore introduces a novel design concept to achieve selective compliance, especially beneficial for shape adaptation of wing structures.

This study will present a proof-of-concept non-contact strain sensor, utilizing a prototype magnetostrictive (Fe-Ga alloy, Galfenol) strip on a steel plate; coupled mechanical-magnetostrictive equations will be used to evaluate sensor performance prediction. In contrast with typical piezo-ceramic materials, Galfenol is ductile and has an excellent ability to withstand mechanical shock. Galfenol also changes its permeability in response to applied stress. This stress-based permeability change is not time dependent and can measure static loads. The variation of magnetic induction with stress depends strongly on bias magnetic field. Typically, the performance of Galfenol is measured in a compressional load region because it has higher response there. However, in this study, we are aiming to develop a sensor for tensile stress measurement. To achieve a compression load in the sensing element, a Galfenol strip is aligned perpendicular to a tension bar, so that tension in the bar creates compression in the strip, via the Poisson effect. The experimental setup in this study consists of a polycrystalline Galfenol strip bonded in the horizontal direction of a steel dog-bone shaped tension specimen. Two permanent magnets are attached at both ends of the Galfenol strip to provide a magnetic bias field through the strip. The magnetic flux through the Galfenol strip is measured with a non-contact Hall sensor during the tensile load test. The design reported here aims at low frequency applications, such as static and dynamic tension monitoring.

NiTi alloys are interesting materials for biomedical implants since they offer unique characteristics such as superelastic behavior, low stiffness (I.e., modulus of elasticity) close to that of the cortical bone, and shock absorption. Thermal treatments are the most common and practical ways to improve the superelasticity of these alloys. In addition to the superelastic behavior of the metallic implants, it is important for the implants to have a stiffness similar to that of cortical bone in order to reduce the risk of failure caused by stress shielding. The cortical bone has a stiffness ranging from 12 to 31 GPa for different patients (e.g., sex, age, mechanical behavior of bone) and various bone locations (e.g., jaw implant, hip implant), while the untreated Ni-rich NiTi has the stiffness equal to 41.37 GPa. One recently used technique to lower the stiffness of NiTi implant is to introduce porosity into the implant. The major problem associated with the imposed porosity is stress concentration on the pore walls and the subsequent implant failure. In this work, the purpose is to tune the stiffness via changing the post-heat treatment conditions, i.e., aging time and aging temperature. In this study, several bulk specimens of Ni-rich NiTi (SLM Ni50.8Ti49.2) were additively manufactured using selective laser melting (SLM) technique. The samples were solution annealed (950 °C, 5.5 h) and subsequently water quenched to provide equilibrium state in the samples. Subsequently, different aging conditions (350 °C and 450 °C for 5 to 18 hours) were applied to the samples. Mechanical testing (compression) was conducted on the samples and the stiffness of each sample was defined to investigate the effect of aging on the stiffness. Our results indicate that the range of 29.9 to 43.7 GPa for stiffness can be achieved through the implant via different time period and temperatures for aging. The modulus of 43.7 GPa is attributed to 10 hours heat treatment under 450 °C and the modulus of 29.9 GPa is attributed to 18 hours heat treatment under 350 °C.

Solid state refrigeration processes, such as magnetocaloric and electrocaloric refrigeration, have recently shown to be a promising alternative to conventional compression refrigeration. A new solid state elastocaloric refrigeration process using the latent heats within Shape Memory Alloys (SMA) could also hold potential in this field. This work investigates the elastocaloric effects in Ni-Ti-based superelastic Shape Memory Alloy (SMA) systems for use in an elastocaloric cooling processes. Ni-Ti alloys exhibits large latent heats and a small mechanical hysteresis, which may potentially lead to the development of an efficient environmentally friendly solid-state cooling system, without the need for ozone-depleting refrigerants. A systematic investigation of the SMA is conducted using a novel custom-built scientific testing platform specifically designed to measure cooling process related phenomena. This testing system is capable of performing tensile tests at high rates as well as measuring and controlling the solid-state heat transfer between SMA and heat source/heat sink. Tests are conducted following a cooling process related training cycle where the material has achieved stabilized behavior. First, a characterization of the elastocaloric material properties is performed followed by an investigation of the material under cooling process conditions. A comprehensive monitoring of the mechanical and thermal parameters enables the observation of temperature changes during mechanical cycling of the SMA at high strain rates. These observations can be used to study the rate dependent efficiency of the elastocaloric material. The measurement of the temperature of both the heat source/heat sink and the SMA itself, as well as the required mechanical work during a running cooling process, reveals the influence of the operating conditions on the elastocaloric effect of the material. Furthermore investigations of the process efficiency at different thermal boundary conditions (temperature of heat source/heat sink), indicates that the process is dependent on the boundary conditions which have to be controlled in order to optimize the efficiency.

Magnetic shape memory alloys are a relatively new class of materials that are suitable for actuation, sensing, and power harvesting. The power harvesting capability comes from the change in magnetization that the material exhibits when internal martensitic variants change orientation. In typical power harvesting tests, the material is loaded with axial compression in the presence of a bias magnetic field applied normal to the compressive loading direction. However, previous results suggest that having a component of the bias magnetic field applied axially, parallel to the compressive stress, can increase the power output of MSMAs. Furthermore, most of the MSMAs power harvesting results reported to date focused on the open circuit voltage that the material can generate during cyclic loading. However, this information is not indicative of the true power harvesting capability of the material and one has to focus on the power output of the material instead. This paper presents voltage trends and power output data for a MSMA sample exposed simultaneously to a cyclic compressive stress and bi-axial magnetic field.

Actuators in the form of a helical spring made from shape memory alloy are attractive due to light weight, large recoverable deformation, high energy density and manufacturing simplicity. For their optimal design and control detailed information on evolution of phase and stress distribution within the material during operation is advantageous. In this work a constitutive model tailored for non-proportionally loaded shape memory alloys exhibiting R-phase transition, transformation strain anisotropy, tension-compression asymmetry is employed to reveal and interpret relation between macroscopic response of such an actuator and microscopic state within the shape memory material. Numerical simulations confirm good predictive capability of the model and demonstrate that because of naturally non-proportional loading mode, phase and stress distributions within cross-section of the wire may be rather complex and counterintuitive.