MEMS Strain Sensors Research Articles

Mechanical and electrical characterization of resonant piezoelectric microbridges for strain sensing

This work reports on a resonantly operated piezoelectric MEMS strain sensor based on a microbridge structure. The resonator is excited with a sputtered aluminum nitride layer and was fabricated with standard CMOS processing techniques. The mechanical frequency spectra and the mode shapes were studied with a laser Doppler vibrometer. Additionally, an electrical read out was realized by measuring the conductance spectra with an impedance analyzer. Several sensor devices varying in width have been fabricated and the differences in the resonance frequency, the mechanical displacement and the electrical conductance spectra have been evaluated, with a focus on the first two Euler-Bernoulli modes. The buckling behavior of the devices under high compressive external strain was analyzed. Furthermore, the mode veering phenomenon could be observed and is analyzed in detail. To classify this new type of strain sensor a frequency-dependent gauge factor has been defined. It shows that the sensor devices exhibit exceptionally high gauge factors, one to two orders of magnitude higher than conventional strain gauges based on the piezoresistive effect. • Resonant piezoelectric microbridges for strain sensing provide exceptional high gauge factor. • Mode veering influences mode shapes and mode amplitude. • Increased microbridge width increases electrical signal without reducing high responsivity. • Lower order modes have higher responsivity than higher order modes.

Vacuum Packaged Low-Power Resonant MEMS Strain Sensor

This paper describes a technical approach toward the realization of a low-power temperature-compensated micromachined resonant strain sensor. The sensor design is based on two identical and orthogonally-oriented resonators where the differential frequency is utilized to provide an output proportional to the applied strain with temperature compensation achieved to first order. Interface circuits comprising of two front-end oscillators, a mixer, and low-pass filter are designed and fabricated in a standard 0.35- $\mu \text{m}$ CMOS process. The characterized devices demonstrate a scale factor of 2.8 Hz/ $\mu \varepsilon $ over a strain range of 1000 $\mu \varepsilon $ with excellent linearity over the measurement range. The compensated frequency drift due to temperature is reduced to 4% of the uncompensated value through this scheme. The total continuous power consumption of the strain sensor is 3 $\mu \text{W}$ from a 1.2 V supply. This low power implementation is essential to enable battery-powered or energy harvesting enabled monitoring applications. [2015-0328]

MEMS sensor fusion: Acoustic emission and strain

In this paper, Micro-Electro-Mechanical-Systems (MEMS) based device accommodating acoustic emission (AE) and strain sensors is presented for damage detection in structures. The MEMS AE sensors have the transduction principle of capacitance change while the MEMS strain sensors have the transduction principle of piezoresistivity. Two aluminum 7075 coupon samples with existing notches are tested under fatigue loading. The performance of MEMS sensors is compared with conventional piezoelectric AE sensors, and strain gauges. The MEMS AE sensors detect the damage initiation and evolution while their sensitivities are lower than piezoelectric AE sensors. The MEMS strain sensors can detect the fatigue cycles while they are influenced by thermal drift, which affects the measurement of true strain. The relevant AE signals released by active flaw occur at high strain levels. The combination of AE and strain sensors on the same device can be utilized to define active and idle modes of AE data acquisition using strain as controlling variable.

Design and Modeling of a Chevron MEMS Strain Sensor With High Linearity and Sensitivity

Two new physical designs for sensors that can be used for strain measurement in large structures, such as bridges, wind turbines, or airplanes, are presented. While the proposed sensor designs focus on high sensitivity, they are based on simple operating principle of comb-drive differential variable capacitances and chevron displacement amplification. The chevron beams convert small amount of applied strains to measurable changes in capacitance of comb fingers. The design of the structures enables simple fabrication methods for the realization of the sensors. Two designs are proposed with the first design can also be used as a sensitive resonant strain sensor. Device performances are validated both by analytical solutions and also by finite-element method simulations. The obtained nominal capacitance is 25 fF, with sensitivities of 13 and 2.7 aF per microstrain ( $\mu \varepsilon$ ) while demonstrating a maximum strain range of ±1000 and $\pm 1800~\mu \varepsilon $ , respectively, for the first and second designs. As a resonant strain sensor, the first design exhibits a sensitivity of $\sim 8.6$ Hz/ $\mu \varepsilon $ .

Strain Transfer Analysis of Surface-Bonded MEMS Strain Sensors

The transmission of strain fields in adhesively bonded MEMS strain sensors is analyzed. In strain sensors that are attached to host structures using adhesive layers such as epoxy, complete strain transfer to the sensor is hindered due to the influence of the adhesive layer on the transfer. This paper presents an analytical model, validated by finite element method simulation, to provide insight and accurate formulation for strain transfer mechanism for bonded sensors. A shear-lag parameter has been introduced to account for the component geometry and properties. The model is capable of predicting the strain transmission ratio through a sensor gauge factor, and it clearly establishes the effects of the flexibility, length, and thickness of the adhesive layer and MEMS sensor substrate. Finally, modifications to sensor substrates, by the implementation of micromachined tapered edges and trench etching, are proposed in order to increase the strain transmission ratio. It has been found, for the selected case study, that the sensor sensitivity could be enhanced by up to 30%. This work on bonding analysis is applicable for performance prediction and calibration in sensor systems design.

Development and Experimental Evaluation of a Novel Piezoresistive MEMS Strain Sensor

This paper presents the experimental evaluation of a new piezoresistive microelectromechanical systems strain sensor. The sensing chip is highly capable of measuring biaxial state of strain/stress. The sensing elements are p-type piezoresistors on (100) single crystal silicon aligned along [110] and its in-plane transverse. The concept of introducing geometric features to enhance the sensor sensitivity is investigated. The results of experimental evaluation and finite-element analysis (FEA) proved the viability of this concept to improve the sensor sensitivity. The microfabrication process utilizes five doping concentrations to explore the effect of doping level on the sensor performance. The sensor is developed considering applications under varying temperature conditions. Therefore, high doping concentration (more than 1 ×1019 atoms/cm3) is favorable to reduce the sensor thermal drift. As a result, the sensor sensitivity is significantly reduced. Hence, geometric features are introduced in the sensor silicon carrier to compensate for the signal loss through stress concentration effect, which magnified the strain field in the proximity of the sensing elements. In addition, the use of full-bridge configuration reduced the overall temperature coefficient of resistance (TCR). At doping concentration of ~5 ×1019 atoms/cm3, the measured strain sensitivity is 0.035 mV/μe for input voltage of 5 volts, which corresponds to an effective gauge factor of ~7 and piezoresistive gauge factor of ~44. The effective gauge factor includes all the signal losses and the effect of bonding adhesive. Design and analysis, prototyping, and experimental evaluation are presented. Finally, guidelines to select the bonding adhesive and packaging scheme are provided.

High-performance piezoresistive MEMS strain sensor with low thermal sensitivity.

This paper presents the experimental evaluation of a new piezoresistive MEMS strain sensor. Geometric characteristics of the sensor silicon carrier have been employed to improve the sensor sensitivity. Surface features or trenches have been introduced in the vicinity of the sensing elements. These features create stress concentration regions (SCRs) and as a result, the strain/stress field was altered. The improved sensing sensitivity compensated for the signal loss. The feasibility of this methodology was proved in a previous work using Finite Element Analysis (FEA). This paper provides the experimental part of the previous study. The experiments covered a temperature range from −50 °C to +50 °C. The MEMS sensors are fabricated using five different doping concentrations. FEA is also utilized to investigate the effect of material properties and layer thickness of the bonding adhesive on the sensor response. The experimental findings are compared to the simulation results to guide selection of bonding adhesive and installation procedure. Finally, FEA was used to analyze the effect of rotational/alignment errors.

Fabrication and packaging techniques for the application of MEMS strain sensors to wireless crack monitoring in ageing civil infrastructures

We report on the development of a new technology for the fabrication of Micro-Electro-Mechanical-System (MEMS) strain sensors to realize a novel type of crackmeter for health monitoring of ageing civil infrastructures. The fabrication of micromachined silicon MEMS sensors based on a Silicon On Insulator (SOI) technology, designed according to a Double Ended Tuning Fork (DETF) geometry is presented, using a novel process which includes a gap narrowing procedure suitable to fabricate sensors with low motional resistance. In order to employ these sensors for crack monitoring, techniques suited for bonding the MEMS sensors on a steel surface ensuring good strain transfer from steel to silicon and a packaging technique for the bonded sensors are proposed, conceived for realizing a low-power crackmeter for ageing infrastructure monitoring. Moreover, the design of a possible crackmeter geometry suited for detection of crack contraction and expansion with a resolution of 10 and very low power consumption requirements (potentially suitable for wireless operation) is presented. In these sensors, the small crackmeter range for the first field use is related to long-term observation on existing cracks in underground tunnel test sections.

1A1-H14 ゴルフクラブに働く撃力の計測(触覚と力覚)

We propose a method to detect the impact force which acts on a golf club head with strain sensors. This method is capable of measuring compression force and shear force and estimating the impact point. We made the large scale model for the impact measurement of a golf club and a ball. We also embedded in a golf club head two MEMS strain sensor units which size was 2 mm x 2 mm x 0.8 mm. This unit detected the strains of 10 kHz on impact in 0.4 ms. The result by the sensor was superior to that by a high speed camera in time resolution.

Fly's proprioception-inspired micromachined strain-sensing structure: idea, design, modeling and simulation, and comparison with experimental results

A new strain-sensing structure inspired from insect's (especially the Fly) propricoception sensor is devised. The campaniform sensillum is a strain-sensing microstructure with very high sensitivity despite its small dimension (diameter ∼10 µm in a relatively stiff material of insect's exocuticle (E = ∼109 Pa). Previous work shows that the high sensitivity of this structure towards strain is due to its membrane-in-recess- and strainconcentrating- hole- features. Based on this inspiration, we built similar structure using silicon micromachining technology. Then a simple characterisation setup was devised. Here, we present briefly, finite-element modeling and simulation based on this actual sample preparation for the characterisation. As comparison and also to understand mechanical features responsible for the strain-sensitivity, we performed the modeling on different mechanical structures: bulk chunk, blind-hole, thorugh-hole, surface membrane, and membrane-in-recess. The actual experimental characterisation was performed previously using optical technique to membranein- recess micromachined Si structure. The FEM simulation results confirm that the bending stress and strain are concentrated in the hole-vicinity. The membrane inside the hole acts as displacement transducer. The FEM is in conformity with previous analytical results, as well as the optical characterisation result. The end goal is to build a new type MEMS strain sensor.

Calibration of MEMS strain sensors fabricated on silicon: Theory and experiments

A calibration technique for measuring MEM's strain sensor performance is presented. For resistance based sensors, calibration entails determining a relationship between change in resistance of the sensor and strain (the gauge factor). A modification to the standard calibration method employed for metal foil, resistance strain gauges is presented. The approach entails constructing two nearly identical test specimens: a specimen with the MEM's sensor mounted with adhesive and a specimen with a strain gauge on silicon mounted with adhesive. Data from the strain gauge specimen provide the basis for evaluating the strain at the sensor. Test data are presented which show that strain at the wafer is 52% to 55% of the strain applied to the specimen. A theoretical basis for this strain transfer relationship is presented. Finally, a dimensionless geometry factor, based on shear lag theory, is derived. As the sensor cross section (width and length) and thickness changes, the strain transfer between the specimen and sensor vary linearly with the geometry factor. This result emphasizes the importance in considering the overall sensor geometry when employing semiconductor strain gauges.

Two Demonstrations of MEMS in Civil Engineering

testing of a MEMS strain sensor for rail fatigue studies and provides a comprehensive view of an overall demonstration project. The second is titled ‘‘Design of Piezoresistive MEMS-Based Accelerometer for Integration with a Wireless Sensing Unit for Structural Monitoring,’’ and is authored by Lynch, Partridge, Law, Kenny, Kiremidjian, and Carryer, from Stanford University. The paper reports on the development of an innovative MEMS accelerometer for structural monitoring, compares its performance to capacitive-type MEMS accelerometers, and describes the demonstration of a system for wireless sensing. Both papers illustrate improvements in civil engineering sensing practices that can be achieved using MEMS devices. They also illustrate steps by which researchers from different disciplines can find collaborative opportunities. Potentially, aerospace and civil engineering applications are an important market for new sensing approaches, and I wish to thank these authors for presenting their pioneering work with MEMS technologies. Irving J. Oppenheim Carnegie Mellon University

Simulation and fabrication of piezoresistive membrane type MEMS strain sensors

Two piezoresistive ( n-polysilicon) strain sensors on a thin Si 3N 4/SiO 2 membrane with improved sensitivity were successfully fabricated by using MEMS technology. The primary difference between the two designs was the number of strips of the polysilicon patterns. For each design, a doped n-polysilicon sensing element was patterned over a thin 3 μm Si 3N 4/SiO 2 membrane. A 1000×1000 μm 2 window in the silicon wafer was etched to free the thin membrane from the silicon wafer. The intent of this design was to fabricate a flexible MEMS strain sensor similar in function to a commercial metal foil strain gage. A finite element model of this geometry indicates that strains in the membrane will be higher than strains in the surrounding silicon. The values of nominal resistance of the single strip sensor and the multi-strip sensor were 4.6 and 8.6 kΩ, respectively. To evaluate thermal stability and sensing characteristics, the temperature coefficient of resistance [TCR=(Δ R/ R 0)/Δ T] and the gage factor [GF=(Δ R/ R 0)/ ε] for each design were evaluated. The sensors were heated on a hot plate to measure the TCR. The sensors were embedded in a vinyl ester epoxy plate to determine the sensor sensitivity. The TCR was 7.5×10 −4 and 9.5×10 −4/°C for the single strip and the multi-strip pattern sensors. The gage factor was as high as 15 (bending) and 13 (tension) for the single strip sensor, and 4 (bending) and 21 (tension) for the multi-strip sensor. The sensitivity of these MEMS sensors is much higher than the sensitivity of commercial metal foil strain gages and strain gage alloys.

Experimental evaluation of MEMS strain sensors embedded in composites

Micromechanical in-plane strain sensors were fabricated and embedded in fiber-reinforced laminated composite plates. Three different strain sensor designs were evaluated: a piezoresistive filament fabricated directly on the wafer; a rectangular cantilever beam; and a curved cantilever beam. The cantilever beam designs were off surface structures, attached to the wafer at the root of the beam. The composite plate with embedded sensor was loaded in uniaxial tension and bending. Sensor designs were compared for repeatability, sensitivity and reliability. The effects of wafer geometry and composite plate stiffness were also studied. Typical sensor sensitivity to a uniaxial tensile strain of 0.001 (1000 /spl mu//spl epsi/) ranged from 1.2 to 1.5% of the nominal resistance (dR/R). All sensors responded repeatably to uniaxial tension loading. However, for compressive bending loads imposed on a 2-3-mm-thick composite plate, sensor response varied significantly for all sensor designs. This additional sensitivity can be attributed to local buckling and subsequent out of plane motion in compressive loading. The curved cantilever design, constructed with a hoop geometry, showed the least variation in response to compressive bending loads. All devices survived and yielded repeatable responses to uniaxial tension loads applied over 10 000 cycles.