A multi-station compression detection device and method for thermistors

Abstract

The application relates to the fields of electronic component detection equipment and electromechanical control technology, in particular to a multi-station pressure resistance detection device and method for a thermistor; the device comprises a main trunk lower pressing plate with a built-in self-pressure-equalizing fluid network, a test pressure head provided with a magneto-rheological elastomer layer and an internal through-conducting micro-wire array and the like modules; the device combines a fluid static pressure chamber and a throttling micro-channel to dynamically redistribute the local overload pressure of the multi-station, and dynamically adjusts the stiffness of the pressure head through a controller and an excitation coil; the core is to adaptively balance the stress of the multi-station based on the fluid network, and to realize low-stiffness buffering at the initial testing stage and high-stiffness force transmission at the pressure applying stage in cooperation with the magneto-rheological material; the application overcomes the uneven stress caused by the component size tolerance, avoids the component bruising and contact point micro-slippage caused by the traditional rigid compression, and effectively guarantees the stability of the electrical measurement.

Description

Technical Field

[0001] This invention relates to the fields of electronic component testing equipment and electromechanical control technology, specifically to a multi-station voltage resistance testing device and method for thermistors. Background Technology

[0002] In the current electronic component manufacturing and quality control environment, brittle ceramic electronic components such as thermistors typically require piezoelectric performance testing. Multi-station synchronous loading structures are often used for batch pressing tests and to simultaneously collect the resistance changes of components under stress. To obtain the electrical response under pressure, existing solutions generally employ a rigid drive architecture with single-sided contact electrodes. This involves a motor driving the main pressing plate downwards, causing the metal indenter to make direct rigid contact with each thermistor on the support platform. During the pressing process, the resistance data fed back by the circuit is continuously measured to determine performance. Although this solution has limitations in ideal height... While it possesses a certain batch testing capability in demanding scenarios, its high dependence on the dimensional accuracy of the test components and the absolute flatness of the base, coupled with the lack of an adaptive pressure compensation mechanism between workstations, makes it prone to load distribution imbalance when the pressure head contacts components with different dimensional tolerances, directly converting the tolerances into localized destructive impact loads. Simultaneously, rigid single-sided metal electrodes are prone to severe fluctuations in contact resistance under pressure and micro-slip conditions. Furthermore, relying solely on single-dimensional resistance monitoring methods suffers from high response delays and numerous false alarms and missed alarms in the early stages of microcrack initiation within components, making it difficult to support non-destructive protection and accurate condition assessment of components under high-frequency testing conditions.

[0003] Therefore, how to balance local load adaptive balancing and contact resistance stability in multi-station synchronous piezoelectric performance testing, and improve the timeliness and accuracy of micro-damage risk warning and control, has become an urgent technical problem to be solved. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides a multi-station withstand voltage testing device and method for thermistors. Specifically, the technical solution of the present invention is as follows:

[0005] A multi-station withstand voltage testing device for thermistors includes:

[0006] The test base, which serves as the main frame supporting the overall structure, includes a top plate, side walls, and a bottom plate.

[0007] A drive mechanism is disposed on the top plate of the test base, and the drive mechanism includes a servo motor and a ball screw driven by the servo motor;

[0008] The main pressure plate is connected to the ball screw and slidably connected to the side wall of the test base. The main pressure plate is provided with a self-equalizing fluid network. When the servo motor is working, the main pressure plate moves in the vertical direction. The self-equalizing fluid network includes interconnected fluid static pressure chambers and throttling microchannels.

[0009] The test head is located below the main pressure plate, corresponding to the fluid static pressure chamber. A magnetorheological elastomer layer is provided at the bottom of the test head, wherein a conductive microfilament array is provided through the inside of the magnetorheological elastomer layer, and an excitation coil is wound around the periphery of the test head.

[0010] The lower electrode support stage is disposed on the base plate of the test base corresponding to the test pressure head;

[0011] The controller is connected to control the operation of the drive mechanism and the excitation coil.

[0012] In one possible implementation, the hydrostatic chamber is fitted into a blind hole pre-machined on the bottom surface of the main pressure plate, and a corrugated elastic diaphragm is sealed to the bottom opening of the hydrostatic chamber, wherein the lower surface of the corrugated elastic diaphragm is connected to the test pressure head.

[0013] In one possible implementation, the hydrostatic chamber and the throttling microchannel are filled with incompressible dimethyl silicone oil.

[0014] In one possible implementation, the top end of the conductive microfilament array is electrically connected to a test lead inside the test indenter, wherein the bottom end of the conductive microfilament array is exposed on the lower surface of the magnetorheological elastomer layer.

[0015] In one possible implementation, the matrix of the magnetorheological elastomer layer is silicone rubber, wherein carbonyl iron powder particles are uniformly dispersed inside the matrix.

[0016] In one possible implementation, the output shaft of the servo motor is directly connected to the ball screw via a flexible coupling, wherein the nut seat of the ball screw is fixedly connected to the main pressure plate.

[0017] In one possible implementation, the two sides of the main pressure plate are slidably connected to the side wall via linear guide rails.

[0018] A multi-station withstand voltage detection control method for thermistors includes:

[0019] S1. Control the servo motor to drive the main lower pressure plate downward to approach the thermistor placed on the lower electrode support platform, and control the excitation coil to be in the de-energized state.

[0020] S2. Control the servo motor to apply the rated pressure-resistant feed amount and control the injection of peak current into the excitation coil;

[0021] S3. Extract the nonlinear resistance change rate of the thermistor over time in real time, and extract the subharmonic spectrum characteristics of the servo motor drive current to obtain the servo current distortion rate.

[0022] S4. Input the extracted resistance nonlinear change rate and the servo current distortion rate into the electrodynamic micro-damage calculation model to calculate the equivalent micro-damage index.

[0023] S5. Determine the relationship between the equivalent micro-damage index and the preset safety threshold. If the equivalent micro-damage index is greater than the preset safety threshold, control the reduction of the injection current of the corresponding excitation coil. If the equivalent micro-damage index is less than or equal to the preset safety threshold, maintain the injection current of the corresponding excitation coil unchanged.

[0024] S6. Based on the average value of the equivalent micro-damage index of all workstations, control and adjust the feed pulse frequency of the servo motor; repeat steps S2 to S6 above.

[0025] In one possible implementation, the test head is internally provided with test leads connected to a conductive microfilament array; the step S3 is preceded by:

[0026] S0. The resistance value of the thermistor is continuously recorded through the test lead, and the difference between the resistance value at the current moment and the resistance value at the previous sampling moment is calculated. The difference is divided by the sampling time interval to obtain the nonlinear change rate of the resistance.

[0027] In one possible implementation, the electrodynamic micro-damage calculation model is built into the controller; the step between S4 and S5 includes: calculating the cumulative integral over time of the deviation of the nonlinear rate of change of resistance from the normal piezoresistive effect linear reference; multiplying the cumulative integral by a pre-calibrated material brittleness coefficient to calculate the equivalent micro-damage index.

[0028] The present invention has the following beneficial effects:

[0029] 1. This invention, by setting a self-equalizing fluid network containing a hydrostatic chamber and throttling microchannels inside the main pressure plate, can achieve dynamic redistribution of local overload pressure when multiple workstations are under asynchronous pressure, effectively overcoming the problem of uneven force caused by component size tolerances; at the same time, the magnetorheological elastomer layer at the bottom of the test head, together with the controller and excitation coil, maintains low stiffness to buffer mechanical impact when approaching the thermistor in the power-off state, and increases stiffness to ensure force transmission when the rated pressure feed is applied and peak current is injected; together with the conductive micro-wire array running through its interior, it avoids component damage and micro-slippage at contact points caused by traditional compression, ensuring the stability of electrical measurements;

[0030] 2. This invention provides a control method incorporating an electrodynamic micro-damage calculation model, capable of extracting the resistance nonlinearity rate of change of a thermistor and the servo current distortion rate of the servo motor drive current in real time. By inputting the resistance nonlinearity rate of change and the servo current distortion rate into the electrodynamic micro-damage calculation model for joint analysis, an equivalent micro-damage index is calculated. The system can achieve real-time perception of micro-damage such as microcracks in components during the detection process. By comparing this index with a preset safety threshold, the controller can independently adjust the current injected into the excitation coil of the corresponding station to release local stress. Based on the average value of the equivalent micro-damage index of all stations, the feed pulse frequency of the servo motor is controlled and adjusted, effectively preventing irreversible crushing damage to brittle materials. Attached Figure Description

[0031] Figure 1 This is a schematic diagram of the overall structure of the device;

[0032] Figure 2 This is a schematic diagram of the drive mechanism of the device;

[0033] Figure 3 This is a schematic diagram of the self-equalizing fluid network structure of the device;

[0034] Figure 4 This is a schematic diagram of the device's hydrostatic chamber and throttling microchannel structure.

[0035] Figure 5 This is a flowchart of the method of the present invention.

[0036] In the diagram: 1. Test base; 2. Top plate; 3. Side wall; 4. Base plate; 5. Drive mechanism; 6. Servo motor; 7. Ball screw; 8. Main lower pressure plate; 9. Self-equalizing fluid network; 10. Fluid static pressure chamber; 11. Throttling microchannel; 12. Test head; 13. Magnetorheological elastomer layer; 14. Conductive microfilament array; 15. Excitation coil; 16. Lower electrode support platform; 17. Blind hole; 18. Corrugated elastic diaphragm; 19. Dimethyl silicone oil; 20. Test lead wire; 21. Silicone rubber; 22. Carbonyl iron powder particles; 23. Flexible coupling; 24. Nut seat; 25. Linear guide rail; 26. Thermistor. Detailed Implementation

[0037] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0038] Example 1:

[0039] Combination Figure 1 and Figure 2 A multi-station pressure testing device for a thermistor 26, comprising:

[0040] Test base 1, which serves as the main frame supporting the overall structure, includes a top plate 2, side walls 3 and a bottom plate 4;

[0041] The drive mechanism 5 is located on the top plate 2 of the test base 1. The drive mechanism 5 includes a servo motor 6 and a ball screw 7 driven by the servo motor 6.

[0042] like Figure 3 As shown, the main pressure plate 8 is connected to the ball screw 7 and slidably connected to the side wall 3 of the test base 1. The main pressure plate 8 is provided with a self-equalizing fluid network 9. When the servo motor 6 is working, the main pressure plate 8 moves in the vertical direction. The self-equalizing fluid network 9 includes interconnected fluid static pressure chambers 10 and throttling microchannels 11.

[0043] The test head 12 is located below the main lower pressure plate 8, corresponding to the fluid static pressure chamber 10. A magnetorheological elastomer layer 13 is provided at the bottom of the test head 12. A conductive microfilament array 14 is provided inside the magnetorheological elastomer layer 13. An excitation coil 15 is wound around the periphery of the test head 12.

[0044] The lower electrode support platform 16 is set on the base plate 4 of the test base 1, corresponding to the test pressure head 12;

[0045] The controller connects and controls the operation of the drive mechanism 5 and the excitation coil 15.

[0046] In this embodiment, the multi-station pressure testing device for the thermistor 26 is configured as a four-station synchronous testing structure. The four stations are arranged in a rectangular array along the bottom surface of the main pressure plate 8. The center distance between the stations can be set to 40mm to 80mm to adapt to the batch testing needs of common chip or leaded thermistors 26.

[0047] The test base 1 adopts a steel or aluminum alloy frame structure. The test base 1 includes a top plate 2, side walls 3 and a bottom plate 4. The top plate 2 is used to install the drive mechanism 5, the side walls 3 are used to provide vertical guide support, and the bottom plate 4 is used to install the lower electrode support platform 16 and bear the reaction force. The overall flatness of the test base 1 can be controlled within 0.05mm to reduce the one-time impact of multi-station installation errors on the distribution of the pressing force.

[0048] The drive mechanism 5 is located on the top plate 2 of the test base 1. The drive mechanism 5 includes a servo motor 6 and a ball screw 7 driven by the servo motor 6. The servo motor 6 can be an AC servo motor with a rated power of 200W to 750W, and the ball screw 7 can be a precision ball screw with a lead of 5mm to 10mm. The function of the drive mechanism 5 is to convert the rotational motion of the servo motor 6 into the vertical linear motion of the main lower pressure plate 8 so as to apply anti-pressure synchronously to each station.

[0049] Compared to relying solely on a rigid pressing structure, this drive mechanism 5 is only responsible for providing the total feed and total driving force. Its output force is redistributed to the test head 12 through the self-equalizing fluid network 9 (described later), thereby avoiding the direct conversion of dimensional tolerances into local impact loads.

[0050] The main pressure plate 8 is connected to the ball screw 7 and slidably connected to the side wall 3 of the test base 1. The main pressure plate 8 has a self-equalizing fluid network 9 inside. The main pressure plate 8 can be formed by processing a metal plate with a thickness of 20mm to 40mm. It has interconnected fluid static pressure chambers 10 and throttling microchannels 11 inside.

[0051] The so-called self-equalizing fluid network 9 refers to a hydraulic equalization structure that, under the condition that multiple workstations are subjected to pressure at different times or with different pressure displacements, uses the flow of incompressible fluid between connected cavities to redistribute the pressure corresponding to local overload among multiple workstations.

[0052] In the technical context of this invention, this structure does not refer to a traditional hydraulic actuator, but rather to a pressure redistribution network located inside the main lower pressure plate 8, used for passively adjusting the load of multi-station pressure heads.

[0053] When the servo motor 6 is working, the main lower pressure plate 8 moves in the vertical direction. The translational stroke can be set from 5mm to 30mm, the approach stage speed can be set from 10mm / s to 50mm / s, and the detection stage speed can be set from 0.1mm / s to 5mm / s.

[0054] The test heads 12 are respectively located below the main lower pressure plate 8 and correspond to the fluid static pressure chamber 10. Each test head 12 has a magnetorheological elastomer layer 13 at its bottom, and a conductive microfilament array 14 is arranged through the inside of the magnetorheological elastomer layer 13. An excitation coil 15 is wound around the periphery of the test head 12.

[0055] The so-called magnetorheological elastomer layer 13 refers to a composite layer in which magnetic particles are dispersed in an elastic polymer matrix and its apparent mechanical properties can be changed under the action of an external magnetic field. Its role in this invention is not simply a buffer pad, but to provide surface compliance in the initial contact of the indenter and to improve stiffness through the magnetic field in the detection loading stage, so as to take into account contact protection, force transmission and electrical conductivity stability.

[0056] The so-called conductive microfilament array 14 refers to multiple slender conductors distributed along the thickness direction of the indenter and capable of passing through the magnetorheological elastomer layer 13 to achieve vertical conductivity; its role in this invention is to maintain a stable electrical signal transmission path at the bottom of the indenter in both compliant contact and high stiffness states, and to reduce the contact resistance fluctuation of a single-sided contact electrode under micro-slip conditions.

[0057] The excitation coil 15 is wound around the periphery of the test head 12 and is used to apply a penetrating magnetic field to the magnetorheological elastomer layer 13. The excitation coil 15 can be wound with enameled copper wire, with 200 to 1200 turns and a working current of 0.1A to 2A.

[0058] The lower electrode support platform 16 is set on the base plate 4 of the test base 1, corresponding to the test pressure head 12. The lower electrode support platform 16 is used to support the thermistor 26 to be tested and forms an electrical contact support corresponding to the lower surface electrode of the thermistor 26. The lower electrode support platform 16 can be made of copper alloy or gold-plated conductive metal, and the flatness of the platform surface can be controlled within 0.02mm to ensure the consistency of electrical contact.

[0059] The controller connects to control the operation of the drive mechanism 5 and the excitation coil 15; the controller can be an industrial control computer, a programmable logic controller with a data acquisition card or an embedded control unit; the controller outputs servo motion commands to control the displacement, speed and feed pulse frequency of the main lower platen 8;

[0060] On the other hand, the controller outputs a drive signal to the excitation coil 15 to control the change of the magnetic field strength of each test head 12; the controller can also receive the sampling signals of the resistance value of the thermistor 26 and the drive current of the servo motor 6, and process them according to the subsequent defined control method.

[0061] Thus, the device provides structural support through the test base 1, provides total load through the drive mechanism 5, adjusts the load of each station through the self-equalizing fluid network 9 in the main lower pressure plate 8, realizes the switching between contact state and stiffness state through the test head 12, provides lower end support and electrical connection through the lower electrode support platform 16, and coordinates motion and magnetic field control through the controller, so that the thermistor 26 can take into account damage resistance, load consistency and electrical measurement stability in multi-station synchronous piezoelectric performance testing.

[0062] Combination Figure 4 The fluid static pressure chamber 10 is assembled in the blind hole 17 machined on the bottom surface of the main lower pressure plate 8. A corrugated elastic diaphragm 18 is sealed at the bottom opening of the fluid static pressure chamber 10, wherein the lower surface of the corrugated elastic diaphragm 18 is connected to the test pressure head 12.

[0063] In this embodiment, the fluid static pressure chamber 10 is assembled in the blind hole 17 machined on the bottom surface of the main lower pressure plate 8; the blind hole 17 can be formed by CNC machining, the hole diameter can be set to 12mm to 30mm, the hole depth can be set to 8mm to 25mm, and each blind hole 17 is connected to the throttling microchannel 11.

[0064] The fluid static pressure chamber 10 can be directly formed by the blind hole 17 body, or a matching cylindrical cavity component can be installed in the blind hole 17 to improve the consistency of the inner wall roughness and assembly and maintenance. This arrangement makes the fluid static pressure chamber 10 located inside the main lower pressure plate 8 and correspond one-to-one with the pressure head of each station, which helps to shorten the fluid pressure transmission path and reduce the volume increase and response lag caused by the pipeline external connection structure.

[0065] A corrugated elastic diaphragm 18 is sealed at the bottom opening of the hydrostatic chamber 10. The corrugated elastic diaphragm 18 is a thin-walled elastic isolation member with a concentric corrugated or annular pleated geometry, which can provide repeatable elastic deformation under small axial displacement and maintain the sealed isolation between the fluid inside the chamber and the external environment.

[0066] Compared to flat diaphragms, this corrugated structure can achieve greater axial compliant displacement and lower boundary tensile stress with the same material thickness, thereby reducing the fatigue risk of the diaphragm under repeated loading; the corrugated elastic diaphragm 18 can be made of stainless steel film, nickel-based alloy film or polymer composite elastic diaphragm, with a thickness of 0.05mm to 0.5mm and an effective axial displacement of 0.1mm to 2mm.

[0067] The lower surface of the corrugated elastic diaphragm 18 is connected to the test head 12; the connection method can be threaded connection, screw plate connection, welding connection or adhesive connection; when using threaded connection, a connecting post can be set at the upper end of the test head 12 and cooperate with the metal connecting seat on the lower surface of the diaphragm;

[0068] When welding is used, a reinforcing shim can be placed in the center of the diaphragm to disperse the heat effect of welding; this connection allows the vertical displacement of the test head 12 to be directly transmitted to the corrugated elastic diaphragm 18, causing changes in the liquid pressure inside the hydrostatic chamber 10.

[0069] When the thickness of the thermistor 26 at a certain station is too large or its clamping position is relatively high, the test head 12 at that station will contact the component earlier. After the test head 12 is subjected to a reaction force, it pushes the corrugated elastic diaphragm 18 upward, causing the internal pressure of the corresponding fluid static pressure chamber 10 to rise.

[0070] Since the hydrostatic chamber 10 is interconnected with other hydrostatic chambers 10 through the throttling microchannel 11, the locally increased pressure will drive the liquid to flow to the work station that is not in sufficient contact or has a low contact load, thereby causing other corrugated elastic diaphragms 18 to move downward, so that the corresponding test head 12 obtains a compensating downward load.

[0071] Therefore, the corrugated elastic diaphragm 18 not only plays a sealing role in this invention, but also plays a role in converting mechanical displacement and fluid pressure, so that the total displacement of the main lower pressure plate 8 can achieve adaptive load distribution between different work positions; this structure can reduce the single-work position overload problem caused by component size deviation, bearing platform installation deviation and pressure head initial height deviation.

[0072] The fluid static pressure chamber 10 and the throttling microchannel 11 are filled with incompressible dimethyl silicone oil 19;

[0073] In this embodiment, the fluid static pressure chamber 10 and the throttling microchannel 11 are filled with incompressible dimethyl silicone oil 19; the term incompressible means that, under the technical conditions of this invention, the volume compressibility of the selected fluid is negligible within the range of detection load and operating temperature, so that it can convert local displacement changes into pressure responses in other connected chambers in a timely manner.

[0074] Dimethyl silicone oil 19 can be selected from models with a kinematic viscosity of 50 cSt to 1000 cSt, preferably 100 cSt to 500 cSt, in order to achieve a balance between flow damping and response speed; the operating temperature range can be 20°C to 80°C, within which dimethyl silicone oil 19 has relatively stable viscosity-temperature characteristics, low volatility and good insulation.

[0075] The reason for using dimethyl silicone oil 19 is as follows:

[0076] Firstly, dimethyl silicone oil 19 exhibits high chemical stability under repeated loading conditions and is not easily altered in bulk modulus and viscosity due to oxidation or moisture absorption.

[0077] Secondly, dimethyl silicone oil 19 has good compatibility with metal cavities, corrugated elastic diaphragms 18 and common sealing materials, which can reduce the risk of sealing aging and corrosion during long-term use.

[0078] Third, dimethyl silicone oil 19 is an electrically insulating medium, so even if the fluid network is located close to the electrical measurement channel, it is not easy to create parasitic conductive paths for the conductive microfilament array 14 and the test circuit.

[0079] After the throttling microchannel 11 and the hydrostatic chamber 10 are filled with dimethyl silicone oil 19, a controlled flow connection is formed between each station. The so-called throttling microchannel 11 refers to a fluid channel with a relatively small cross-sectional size and relatively large flow resistance. Its function is to limit the flow rate of fluid between each chamber, so that instantaneous local overload will not directly turn into violent overall oscillation, but will complete the pressure redistribution within a set time constant.

[0080] The cross-sectional width of the throttling microchannel 11 can be set from 0.2 mm to 2 mm, the height can be set from 0.2 mm to 1.5 mm, and the length can be set from 5 mm to 50 mm. By selecting the viscosity of dimethyl silicone oil 19 and the geometric parameters of the throttling microchannel 11, the pressure equalization time constant can be set in the range of 10 ms to 500 ms to adapt to the size compensation requirements when the thermistor 26 is synchronously loaded.

[0081] During the testing process, when a single workstation makes premature contact, causing the pressure inside the corresponding chamber to rise, dimethyl silicone oil 19 flows to adjacent or other connected chambers through the throttling microchannel 11 under the pressure difference.

[0082] Because dimethyl silicone oil 19 has low compressibility, this flow process can convert the upward displacement of this station into the downward displacement of the diaphragm in other stations, so that the overall load tends to be balanced.

[0083] If a gas or a highly compressible liquid is used, the local displacement will be absorbed by the fluid itself, resulting in insufficient force transmission across workstations. Therefore, in this embodiment, dimethyl silicone oil 19 is selected as the medium of the self-equalizing fluid network 9.

[0084] The top end of the conductive microfilament array 14 is electrically connected to the test wire 20 inside the test head 12, wherein the bottom end of the conductive microfilament array 14 is exposed on the lower surface of the magnetorheological elastomer layer 13.

[0085] In this embodiment, the top end of the conductive microfilament array 14 is electrically connected to the test wire 20 inside the test head 12, and the bottom end of the conductive microfilament array 14 is exposed on the lower surface of the magnetorheological elastomer layer 13. The test wire 20 refers to a conductive connector disposed inside the test head 12 and connected to an external measurement circuit or controller, which is used to lead out the electrical signal of the electrode on the upper surface of the thermistor 26.

[0086] The test lead 20 can be made of multi-strand flexible wire, fine-diameter shielded wire or deposited metal trace structure; the conductive microfilament array 14 can be made of gold-plated copper wire, stainless steel microfilament, tungsten wire or nickel wire, the diameter of a single wire can be set to 20μm to 200μm, and the array density can be set to 10 to 200 wires per mm².

[0087] The top end of the conductive microfilament array 14 can be electrically connected to the test lead 20 by welding, crimping, conductive adhesive connection or sintering; the connection area should preferably be set in the rigid support cavity inside the test head 12 to reduce the influence of external bending on the resistance of the connection point; the bottom end of the conductive microfilament array 14 is exposed on the lower surface of the magnetorheological elastomer layer 13, and the exposed length can be controlled within the range of 5μm to 200μm;

[0088] This exposed structure allows each conductive microfilament to form a microscale independent contact point on the surface of the magnetorheological elastomer layer 13. When the test head 12 contacts the electrode coating of the thermistor 26, multiple microfilaments can simultaneously establish parallel conductive paths.

[0089] The technical problem solved by this structure is that a single metal planar electrode is prone to unstable contact resistance during crimping testing due to surface roughness, local protrusions or slight changes in contact posture. The multi-point conductive microfilament array 14 can adapt to the micro-surface morphology during the flexible dielectric coating process.

[0090] Since the conductive microfilament array 14 is embedded inside the magnetorheological elastomer layer 13, its root is elastically supported and can bend to a limited extent with the surface morphology in the initial contact stage, thereby reducing electrode damage caused by hard contact; after the excitation coil 15 is energized, the stiffness of the magnetorheological elastomer layer 13 increases and the substrate around the conductive microfilament array 14 hardens synchronously, which improves the positional stability of the conductive microfilaments in the crimping state and reduces the resistance fluctuation caused by micro-slippage at the contact point.

[0091] During testing, the thermistor 26 can be contacted by the lower electrode support platform 16 and the conductive microwire array 14 to its lower electrode and upper electrode respectively, forming a double-ended measurement circuit; the controller or external measurement module can continuously collect the resistance value of the thermistor 26 through the test wire 20.

[0092] Since the bottom end of the conductive microfilament array 14 is directly exposed on the lower surface of the magnetorheological elastomer layer 13, the current path does not need to pass through the insulating substrate, but is transmitted only through the conductive microfilaments themselves, thus achieving both compliant contact and stable conductivity.

[0093] The matrix of the magnetorheological elastomer layer 13 is silicone rubber 21, wherein carbonyl iron powder particles 22 are uniformly dispersed inside the matrix.

[0094] In this embodiment, the matrix of the magnetorheological elastomer layer 13 is silicone rubber 21, and carbonyl iron powder particles 22 are uniformly dispersed inside the matrix. Silicone rubber 21 refers to a polymer material with organosilicon polymer as the main body and an elastic network formed after cross-linking. Its selection in this invention is based on its good elastic recovery, temperature stability and molding and covering properties, which can adapt to the small thickness composite layer structure at the bottom of the test head 12.

[0095] The carbonyl iron powder particles 22 are soft magnetic particles with a particle size that can be set from 1 μm to 20 μm and a volume fraction that can be set from 10% to 45%, preferably from 20% to 35%. The so-called uniform dispersion means that the carbonyl iron powder particles 22 have no obvious agglomeration areas in the thickness direction and in-plane direction of the silicone rubber 21 matrix, so that the resulting magnetorheological elastomer layer 13 has approximately consistent magnetic and mechanical responses in different regions.

[0096] The magnetorheological elastomer layer 13 can be prepared by mixing liquid silicone rubber 21 prepolymer, crosslinking agent and carbonyl iron powder particles 22 in a set ratio, removing internal air bubbles by vacuum degassing, prepositioning conductive microfilament array 14 in a mold, pouring the mixture into the bottom mold cavity of the test head 12 and curing it.

[0097] The curing temperature can be set to 25°C to 120°C, and the curing time can be set to 30 minutes to 6 hours; after curing, a magnetorheological elastomer layer 13 with a thickness of 0.1 mm to 2 mm is formed; the thickness of this layer should not be too large to avoid excessive influence on the transmission accuracy of compression deformation in the absence of a magnetic field; nor should it be too small to ensure sufficient surface compliance and conductive microfilament encapsulation and fixation capabilities.

[0098] In the absence of a magnetic field, the silicone rubber 21 matrix mainly determines the low modulus characteristics of the magnetorheological elastomer layer 13, enabling it to undergo limited coating deformation when in contact with the surface of the thermistor 26; after applying a magnetic field, the carbonyl iron powder particles 22 tend to form chain-like or columnar aggregation along the direction of the magnetic field lines, resulting in an increase in the equivalent shear modulus and compressive modulus of the composite layer.

[0099] This change is not achieved by replacing materials, but by adjusting the particle arrangement state through an external magnetic field. Therefore, it is suitable for rapid state switching in multi-station synchronous testing. Compared with ordinary elastic pads, this structure can maintain low contact stress concentration during the contact establishment stage and high load transfer stiffness during the compressive strength testing stage, thereby reducing the pressure distortion problem caused by long-term energy absorption of flexible materials.

[0100] The output shaft of the servo motor 6 is directly connected to the ball screw 7 via the flexible coupling 23, wherein the nut seat 24 of the ball screw 7 is fixedly connected to the main lower pressure plate 8;

[0101] In this embodiment, the output shaft of the servo motor 6 is directly connected to the ball screw 7 via the flexible coupling 23, and the nut seat 24 of the ball screw 7 is fixedly connected to the main lower pressure plate 8. The so-called flexible coupling 23 refers to a connecting piece that can compensate for a small amount of coaxiality error, angular error or axial micro-displacement between the motor output shaft and the ball screw 7 input shaft while transmitting torque.

[0102] The flexible coupling 23 can be a plum blossom-shaped flexible coupling 23, a diaphragm coupling or a bellows coupling, preferably a plum blossom-shaped flexible coupling 23 with high torsional stiffness and moderate compensation capacity; its rated torque can be set from 1 N·m to 20 N·m according to the output capacity of the servo motor 6.

[0103] By using a flexible coupling 23 to directly connect the ball screw 7, the backlash and additional elastic deformation caused by intermediate transmission components such as gears and belts can be reduced, so that the output angle of the servo motor 6 and the axial displacement of the ball screw 7 can maintain a clear correspondence.

[0104] Since the method of the present invention requires adjusting the feed pulse frequency of the servo motor 6 according to the equivalent micro-damage index, if there is a large backlash or unstable transmission ratio in the transmission chain, the correspondence between the distortion rate of the drive current and the actual end force change will be reduced, which is not conducive to subsequent micro-damage mapping.

[0105] Therefore, the output shaft of the servo motor 6 is directly connected to the ball screw 7 through the flexible coupling 23, which has the effect of improving the displacement control accuracy and the interpretability of the drive signal;

[0106] The nut seat 24 of the ball screw 7 is fixedly connected to the main lower pressure plate 8; the fixed connection can be a flange connection, a screw connection or an integrated machining connection; when the ball screw 7 rotates, the nut seat 24 moves axially, driving the main lower pressure plate 8 to make a vertical translation.

[0107] The lead of the ball screw 7 can be determined according to the required pressing resolution. For example, when the lead is 5mm and the encoder resolution of the servo motor 6 is 10,000 pulses per revolution, the theoretical displacement resolution can reach 0.5μm.

[0108] Through this transmission structure, the frequency of the feed pulse output by the controller can be directly converted into the feed speed change of the main lower pressure plate 8, enabling the device to stably load at the mm / s level or even lower speed level according to the requirements of the detection algorithm.

[0109] The two sides of the main lower pressure plate 8 are slidably connected to the side wall 3 via linear guide rails 25;

[0110] In this embodiment, the two sides of the main lower pressure plate 8 are slidably connected to the side wall 3 by linear guide rails 25; the linear guide rails 25 can be ball linear guide rails 25, auxiliary guide rails with sliders, or cross roller guide rails; one or two guide rails can be provided on each side, and one or more sliders can be provided on each guide rail to improve the resistance to eccentric loads.

[0111] The parallelism between the guide rail mounting reference surface and the axis of the ball screw 7 can be controlled within 0.03mm / 300mm, and the relative parallelism between the left and right guide rails can be controlled within 0.02mm / 300mm. This guide structure is used to limit the lateral displacement and rotational freedom of the main lower pressure plate 8, so that it can maintain vertical translation under the drive of the servo motor 6.

[0112] During multi-station testing, even if the self-equalizing fluid network 9 can redistribute the local pressure to a certain extent, the main lower pressure plate 8 will still be subject to instantaneous eccentric load torque caused by the incomplete synchronous contact between stations. If the linear guide rails 25 on both sides are not constrained, the main lower pressure plate 8 may pitch or tilt, exacerbating the uneven contact of a single station.

[0113] By setting linear guide rails 25 on both sides of the main lower pressure plate 8 and sliding them with the side wall 3, the movement trajectory of the main lower pressure plate 8 can be made more stable, and the overturning tendency caused by uneven load can be limited to the small gap range allowed by the guide rail.

[0114] The side wall 3 can be a vertical plate of the overall frame of the test base 1, or it can be a column-type lateral support that is independently installed between the bottom plate 4 and the top plate 2; the connection between the linear guide rail 25 and the side wall 3 can be fixed with bolts, and the connection between the main lower pressure plate 8 and the slider can be fixed with countersunk screws or pressure plates.

[0115] With this structure, the main pressure plate 8 has high guiding stiffness when driven by the ball screw 7, and the internal self-equalizing fluid network 9 can further improve the attitude stability and load output consistency during the multi-station pressure process.

[0116] Example 2:

[0117] like Figure 5 As shown, a multi-station withstand voltage detection control method for a thermistor 26 includes:

[0118] S1. Control the servo motor 6 to drive the main lower pressure plate 8 downward to approach the thermistor 26 placed on the lower electrode support platform 16, and control the excitation coil 15 to be in a de-energized state.

[0119] S2. Control the servo motor 6 to apply the rated pressure-resistant feed amount and control the injection of peak current into the excitation coil 15;

[0120] S3. Extract the nonlinear resistance change rate of the thermistor 26 over time in real time, and extract the subharmonic spectrum characteristics of the drive current of the servo motor 6 to obtain the servo current distortion rate.

[0121] S4. Input the extracted resistance nonlinearity change rate and servo current distortion rate into the electrodynamic micro-damage calculation model to calculate the equivalent micro-damage index.

[0122] S5. Determine the relationship between the equivalent micro-damage index and the preset safety threshold. If the equivalent micro-damage index is greater than the preset safety threshold, control the reduction of the injection current of the corresponding excitation coil 15. If the equivalent micro-damage index is less than or equal to the preset safety threshold, maintain the injection current of the corresponding excitation coil 15 unchanged.

[0123] S6. Based on the average value of the equivalent micro-damage index of all workstations, control and adjust the feed pulse frequency of servo motor 6; repeat steps S2 to S6 above.

[0124] In this embodiment, the control method is applied to the aforementioned multi-station pressure testing device for thermistor 26 and is executed by the controller; the nonlinear resistance change rate in this method refers to the component of the rate of change of the resistance value of thermistor 26 relative to time during the pressure process that deviates from the steady-state linear piezoresistive response.

[0125] Servo current distortion rate refers to the degree of subharmonic component change of the drive current of servo motor 6 relative to the reference drive waveform, used to characterize the transient resistance fluctuation caused by mechanical transmission chain and pressure contact interface; equivalent micro damage index refers to a dimensionless or normalized index that characterizes the degree of micro stress concentration inside the thermistor 26, calculated based on the combined characteristics of electrical and driving mechanics.

[0126] This method uses different indenter stiffness states for the contact and loading stages, and combines resistance signals with drive signals for damage prediction, thereby reducing the risk of irreversible pressure damage while the device reaches the target detection depth.

[0127] In S1, the servo motor 6 is controlled to drive the main lower pressure plate 8 downward to approach the thermistor 26 placed on the lower electrode support platform 16, and the excitation coil 15 is controlled to be de-energized; at this time, the magnetorheological elastomer layer 13 maintains a low modulus state, and the test head 12 can produce surface conformity in the initial contact with the thermistor 26.

[0128] The approach speed can be set from 10 mm / s to 50 mm / s, and the contact criterion can be any one or a combination of displacement threshold, drive current sudden change threshold or contact resistance formation threshold.

[0129] The reason for using the power-off state is that if the pressure head directly contacts the thermistor 26 in a high-rigidity state, it is easy to turn the micro-protrusions on the surface of the component into local high-stress points, which is not conducive to the protection of brittle ceramic materials.

[0130] In S2, the servo motor 6 is controlled to work and apply the rated compressive feed amount, and the peak current is injected into the excitation coil 15. The rated compressive feed amount can be set from 0.01mm to 1mm according to the specifications of the thermistor 26. The peak current can be set from 0.2A to 2A according to the design of the excitation coil 15 and the thickness of the magnetorheological elastomer layer 13. The energization set-up time can be set from 1ms to 100ms.

[0131] In this step, the servo motor 6 continues to advance, and after the peak current is injected into the excitation coil 15, the stiffness of the magnetorheological elastomer layer 13 is increased, and the spatial position stability of the conductive microfilament array 14 is simultaneously improved, so that the main driving force can be transmitted to the thermistor 26 body more directly; this method makes the initial detection stage and the main detection stage correspond to the pressure head state with low stiffness and high stiffness, respectively.

[0132] In S3, the nonlinear resistance change rate of the thermistor 26 over time is extracted in real time, and the subharmonic spectrum characteristics of the drive current of the servo motor 6 are extracted to obtain the servo current distortion rate; the sampling frequency of the resistance signal can be from 1kHz to 100kHz, and the sampling frequency of the servo current signal can be from 5kHz to 200kHz.

[0133] Within the ideal elastic compression range, the resistance of the thermistor 26 typically exhibits a calibrable continuous change; however, when internal microcracks initiate or grain boundaries are subjected to abnormal stress, short-term fluctuations, micro-steps, or local slope changes may occur.

[0134] The drive current of the servo motor 6 has a predetermined waveform when it is driving stably. When the end pressure state changes abruptly, the change in mechanical transmission resistance will be fed back as a change in the subharmonic component in the drive current spectrum. Extracting these two types of signals at the same time can reduce the probability of misjudgment caused by relying on only a single electrical signal or a single mechanical signal.

[0135] In S4, the extracted resistance nonlinearity change rate and servo current distortion rate are input into the electrodynamic micro-damage calculation model to calculate the equivalent micro-damage index. The electrodynamic micro-damage calculation model can be an empirical model, a lookup table model, a linear weighted model, or a nonlinear regression model based on calibration data.

[0136] Its inputs include at least the resistance nonlinearity change rate and servo current distortion rate of the current station, and the output is the equivalent microscopic damage index of the station; the model can be pre-calibrated in combination with parameters such as the material type, nominal resistance, and size specifications of the thermistor 26.

[0137] In S5, the relationship between the equivalent micro-damage index and the preset safety threshold is determined; if the equivalent micro-damage index is greater than the preset safety threshold, the injection current of the corresponding excitation coil 15 is reduced; if the equivalent micro-damage index is less than or equal to the preset safety threshold, the injection current of the corresponding excitation coil 15 is kept unchanged.

[0138] The controlled object in this step is the excitation coil 15 corresponding to the workstation, so the stiffness of each workstation can be adjusted differently.

[0139] After reducing the current in the excitation coil 15, the local stiffness of the magnetorheological elastomer layer 13 decreases, and the bottom of the pressure head regains a certain degree of compliance, which can buffer the stress peak on the contact interface; the safety threshold here can be set to a normalized value range of 0.6 to 0.95 based on the material brittleness, the upper limit of target detection, and historical calibration data.

[0140] In S6, the feed pulse frequency of the servo motor 6 is controlled and adjusted according to the average value of the equivalent micro-damage index of all workstations. If the average value is close to the safety threshold, the feed pulse frequency is reduced to decrease the pressing speed of the main pressure plate 8. If the average value is significantly lower than the safety threshold, the original pulse frequency is maintained or the pulse frequency is appropriately increased within the set range.

[0141] Specifically, the adjustment of the feed pulse frequency adopts segmented proportional control logic. The controller calculates the difference between the average value of the equivalent micro-damage index of all stations and the preset safety threshold. When the difference is greater than zero, the feed pulse frequency is reduced according to the set first proportional coefficient to achieve rapid deceleration and risk avoidance.

[0142] When the difference is less than or equal to zero, the increment of the feed pulse frequency is calculated based on the set second proportional coefficient; to ensure that the response of deceleration avoidance is faster than the response of acceleration recovery, the first proportional coefficient is set to be significantly greater than the second proportional coefficient.

[0143] To implement the above proportional control logic, the controller internally uses the following piecewise function formula to calculate the adjustment amount of the feed pulse frequency. and the updated target pulse frequency :

[0144]

[0145]

[0146] in, This represents the average equivalent micro-damage index of all workstations within the current control cycle. To preset a safety threshold, The first proportional coefficient is set. This is the second proportional coefficient that is set. This represents the current feed pulse frequency of servo motor 6. This is the adjustment amount for the feed pulse frequency. The updated target pulse frequency;

[0147] The following quantitative calculation example illustrates the adjustment and control of speed increase and decrease: Assume that the preset safety threshold is 0.80, the set proximity margin is -0.05, and the current feed pulse frequency of servo motor 6 is 10000Hz;

[0148] If the average equivalent micro-damage index of all stations is calculated to be 0.85 within a certain control cycle, and the difference between this average value and the threshold is positive 0.05, and if the first proportional coefficient is set to 20000Hz, then the feed pulse frequency decrease is 0.05 multiplied by 20000 equals 1000Hz, and the system control outputs a new pulse frequency of 9000Hz.

[0149] If the average value in another cycle is 0.72, the difference is calculated to be -0.08. Since it is less than zero and completely exceeds the range close to the margin of -0.05, if the second proportional coefficient is set to 5000Hz, the increment is calculated by multiplying the absolute value of the difference 0.08 by 5000, which equals 400Hz. The pulse frequency is then increased to 10400Hz.

[0150] This calculation process achieves quantitative closed-loop negative feedback control; the average value in this step is used to reflect the overall risk level of the tested batch at the current moment.

[0151] Since the drive mechanism 5 is a shared main drive structure, the feed speed is adjusted by the average value of all stations, which can avoid the overload of individual stations being continuously amplified, while maintaining the synchronous detection cycle of multiple stations.

[0152] The above steps S2 to S6 are executed repeatedly. In actual execution, S1 mainly corresponds to the approach and initial contact phase, while S2 to S6 are mainly performed in a periodic loop during the loading detection phase. The control period can be set from 1ms to 20ms.

[0153] Through this method, the device uses a low stiffness state to reduce the initial impact when contact is established, and a high stiffness state to ensure load and conductivity stability when continuous pressure is applied. It also performs damage warning and control adjustment based on the nonlinear change rate of resistance and the distortion rate of servo current, thereby realizing the piezoelectric performance detection of the multi-station thermistor 26.

[0154] In this embodiment, the data flow and judgment basis of the above parameters in the control logic are as follows: The input source of the resistance nonlinear change rate comes from the real-time resistance value sequence collected by the test wires 20 at each station, which represents the abnormal change intensity of the thermistor 26 relative to the normal piezoresistive response under the current pressure state.

[0155] The input source of the servo current distortion rate comes from the phase current or bus current sampling sequence fed back by the servo driver, which represents the degree of disturbance of the end contact state change to the stability of the drive load; the equivalent micro damage index is the intermediate state quantity used by the controller to perform protection judgment. The larger the value, the higher the possibility of microcrack initiation, abnormal grain boundary slip, or local stress concentration continuing to develop inside the thermistor 26.

[0156] The preset safety threshold is the comparison threshold for triggering softening protection and speed reduction control. This threshold is preferably determined through sample calibration test, that is, select the same specification thermistor 26, and statistically analyze the equivalent micro damage index range before irreversible electrical performance drift occurs under different loading conditions, and then write it into the controller after leaving a safety margin.

[0157] The processing flow of S3 can be executed in the following order: The controller synchronously reads the resistance sampling value and servo current sampling value of the current station in each control cycle; filters and differentials the resistance sampling value to obtain the resistance change rate at the current moment, and then compares it with the normal piezoresistive change benchmark corresponding to the current displacement segment or load segment to obtain the nonlinear change rate of resistance.

[0158] The servo current sampling value is subjected to time window interception, and the amplitude of the subharmonic component other than the fundamental frequency is extracted within the time window. The amplitude of the subharmonic component is compared with the pre-stored reference amplitude under the same speed, no-load or standard load conditions to obtain the servo current distortion rate.

[0159] Specifically, the calculation process for the servo current distortion rate, which compares the amplitude of the harmonic component with a pre-stored reference amplitude, is performed according to the following formula:

[0160]

[0161] in, The servo current distortion rate extracted at the current moment. This represents the total amplitude of the subharmonic components extracted within the current time window. This is a pre-stored reference amplitude under the same speed and standard load conditions; through this ratio difference calculation, the abstract spectral amplitude variation is transformed into a standardized digital characteristic quantity; the total amplitude of the subharmonic components. The specific calculation method is to extract the first time window. Next to The root mean square of the amplitude of the second major harmonic component, i.e.:

[0162]

[0163] in, For the first The amplitude of the second harmonic. The highest harmonic order for analysis is preset; these two results are fed into the electrodynamic micro-damage calculation model in S4 as parallel inputs; thus, the output of S3 is not the original electrical signal itself, but two state variables that have completed feature extraction and can be used for logical judgment.

[0164] The processing sequence from S4 to S6 can be understood as follows: The controller first determines whether there is an abnormal electrical response inside the material based on the nonlinear change rate of resistance, and then determines whether the abnormality is accompanied by mechanical resistance or abrupt change in contact state based on the servo current distortion rate. The two are combined to form an equivalent microscopic damage index.

[0165] When the equivalent micro-damage index of a certain station exceeds the preset safety threshold, it indicates that there is a high risk of crushing damage if the station continues to advance while maintaining the current pressure head stiffness. Therefore, the controller prioritizes reducing the injection current of the excitation coil 15 of the station, so that the stiffness of the magnetorheological elastomer layer 13 at the bottom of the pressure head of the station decreases and the local stress peak is released first.

[0166] After the single-station protection is completed, the controller then checks the average value of the equivalent micro-damage index of all stations to determine whether to reduce the overall pushing speed of the main lower pressure plate 8. In other words, S5 is responsible for local protection of a single station, and S6 is responsible for the overall cycle adjustment of multiple stations. The two correspond to local risk suppression and system-level risk suppression, respectively.

[0167] To facilitate project implementation, the preset safety threshold can be set to two or three levels. When using two levels of thresholds, an early warning threshold and a protection threshold can be set: when the equivalent microscopic damage index reaches the early warning threshold but does not reach the protection threshold, only the anomaly is recorded and the current loading is maintained.

[0168] When the equivalent micro-damage index reaches the protection threshold, the excitation current reduction action in S5 is executed; when a three-level threshold is used, a shutdown threshold can also be added. When any station exceeds the shutdown threshold for multiple consecutive control cycles, the controller stops the servo motor 6 from continuing to press down and lifts the main pressure plate 8.

[0169] The above thresholds are still a refinement of the implementation of the preset safety threshold. The purpose is to make the threshold not only have a numerical range, but also have a clear logical function and execution consequences.

[0170] To make the causal relationship between each step clearer, the execution logic of S1 to S6 in a detection cycle can be limited as follows: S1 first completes the initial bonding of multiple stations in a low-rigidity contact state to reduce the local impact caused by component thickness difference and surface micro-protrusion.

[0171] Then, under the premise that the contact has been established, S2 increases the stiffness of the pressure head so that the subsequent feed amount is more converted into a stable load on the thermistor 26 body; S3 and S4 respectively identify whether there is an anomaly from the electrical response side and the drive load side, and summarize the identification results into a comparable equivalent micro damage index; S5 and S6 respectively apply to the single station stiffness and the global feed speed.

[0172] Since the adjustment of the single-station excitation coil 15 changes the local contact compliance, while the adjustment of the feed pulse frequency of the servo motor 6 changes the overall propulsion rhythm of the shared main drive, executing S5 first and then S6 can first reduce the stress peak locally and then suppress the continued accumulation of risk as a whole, so that the control logic corresponds one-to-one with the local execution unit and the global execution unit of the device in terms of structure.

[0173] To avoid the interpretation that the repeated execution of steps S2 to S6 is a complete re-execution of the approach action in each cycle, in this embodiment, the meaning of cycle is limited to: S1 is executed to establish initial contact at the beginning of a test; after entering the rated compressive strength test range, S2 to S6 are repeated according to the control cycle until the target feed rate is reached, the target test load is reached, the stop threshold is triggered at any station, or the test completion condition is met.

[0174] In other words, S1 mainly occurs at the beginning of a single detection, while S2 to S6 constitute the closed-loop adjustment body in the loading process; the above steps are coordinated with the actual continuous loading process of the device, and the main pressure plate 8 remains continuously pressed down in each control cycle without needing to retract and then contact again;

[0175] The electrodynamic micro-damage calculation model can be logically divided into three parts: an anomaly extraction unit, a credibility correction unit, and a damage output unit. The anomaly extraction unit receives the nonlinear change rate of resistance, which is used to characterize whether a continuous anomaly deviating from the normal piezoresistive effect has appeared inside the thermistor 26 material.

[0176] The credibility correction unit receives the servo current distortion rate to characterize whether the above electrical anomalies are accompanied by mechanical resistance fluctuations or contact state abrupt changes; the damage output unit receives the results of the first two and combines them with the calibration parameters corresponding to the sample specifications to output a standardized equivalent micro-damage index.

[0177] The model as a whole represents the coupled causal relationship between thermistor 26 and the internal structural anomaly caused by local stress concentration during the compression process, which is simultaneously fed back to the resistance change and the driving load change. Therefore, its role is not to replace the direct measurement of cracks, but to indirectly characterize the risk of micro-damage by the combined evidence of electrical and mechanical anomalies when it is impossible to observe the micro-cracks inside the component in real time.

[0178] The test head 12 is internally provided with test wires 20 that are connected to the conductive microfilament array 14;

[0179] The steps preceding S3 include:

[0180] S0. The resistance value of the thermistor 26 is continuously recorded through the test lead 20, and the difference between the resistance value at the current moment and the resistance value at the previous sampling moment is calculated. The difference is divided by the sampling time interval to obtain the nonlinear change rate of the resistance.

[0181] In this embodiment, the test head 12 is provided with a test wire 20 connected to the conductive microfilament array 14; one end of the test wire 20 is connected to the top of the conductive microfilament array 14, and the other end is connected to the resistance measurement circuit; the resistance measurement circuit can adopt a constant current source voltage acquisition method or a Wheatstone bridge method.

[0182] For the thermistors 26 with resistance values ​​ranging from tens of ohms to hundreds of kilohms, a detection current of 10μA to 5mA can be selected to balance measurement resolution and self-heating effect control; the test leads 20 should preferably be wired with shielded wire or twisted pair to reduce electromagnetic interference generated by the switching on and off of the excitation coil 15 and the drive of the servo motor 6.

[0183] Before step S3, step S0 is included; in S0, the resistance value of the thermistor 26 is continuously recorded through the test lead 20, and the difference between the resistance value at the current moment and the resistance value at the previous sampling moment is calculated. The difference is then divided by the sampling time interval to obtain the nonlinear change rate of the resistance. Specifically, the controller reads the resistance value at the current moment at a set sampling frequency. Resistance value at the previous sampling time Calculate the difference :

[0184]

[0185] Then Divide by the sampling time interval The instantaneous rate of change is obtained. :

[0186]

[0187] Among them, subscript Indicates the discrete time node number of the current sampling period, subscript This indicates the discrete time node number of the immediately preceding sampling period; The time interval between two adjacent samples;

[0188] To distinguish between normal piezoresistive changes and nonlinear abnormal changes, we can... The deviation is obtained by comparing the deviation with the calibrated normal piezoresistive response slope reference; the specific comparison calculation is performed using the following formula: Let the normal piezoresistive response slope reference corresponding to the current state be... Then the instantaneous rate of change deviates by The preliminary calculation formula is as follows:

[0189]

[0190] Further introduce the noise threshold set by the system. Perform anomaly confirmation and judgment: If Then force set ;like Then the The deviation is confirmed as a valid deviation; this valid deviation is the nonlinear rate of change of resistance subsequently input into the micro-damage model.

[0191] The deviation or its filtered result can be used as the input for the nonlinear resistance change rate in subsequent S3 and S4. To improve operability, digital filtering can be added to S0. The filtering method can be moving average filtering, median filtering or first-order low-pass filtering. The moving average window length can be 3 to 20 sampling points to suppress high-frequency electrical noise and retain the short-term anomaly corresponding to the microcrack initiation.

[0192] If there is significant power frequency interference in the measurement circuit, a 50Hz or 60Hz notch filter can be added to the controller; through the above continuous recording and differential calculation, the time change of the resistance value can be converted into a derivative that is more directly related to the internal state change of the material, thereby improving the sensitivity to short-term anomalies;

[0193] When this embodiment is used in conjunction with the aforementioned device structure, the conductive microfilament array 14 provides the upper electrode contact, the test lead 20 provides the internal lead-out path, and the lower electrode support platform 16 provides the lower electrode contact, forming a continuous electrical measurement circuit.

[0194] Since the resistance value recording is continuous, the controller can synchronously acquire resistance change information throughout the entire process of the main lower platen 8 being pushed forward, providing basic data for subsequent judgment of internal micro-damage of the thermistor 26;

[0195] In this embodiment, the nonlinear rate of change of resistance in S0 does not simply refer to the instantaneous rate of change dRdt itself, but rather to the deviation of the instantaneous rate of change from the normal piezoresistive effect linear reference. Its processing logic can be executed in the following order: First, the normal piezoresistive effect linear reference of the same type of thermistor 26 is pre-stored in the controller. This reference can be stored in segments according to displacement interval, load interval, or time interval.

[0196] Then, based on the displacement value, feed rate, or estimated load under the current control cycle, the reference rate of change corresponding to the current state is called; the measured instantaneous rate of change is compared with the reference rate of change. If the difference between the two is within the noise tolerance, the current change is determined to be a normal piezoresistive response; if the difference exceeds the noise tolerance, the excess part is recorded as the nonlinear rate of change of resistance and sent to the subsequent steps.

[0197] The specific numerical deduction example of the internal processing of this feature determination is as follows: Assume that the normal piezoresistive response linear change rate benchmark of the thermistor 26 of the same size and specification is set to -1.5Ω / ms in this load segment;

[0198] When the system acquires a measured resistance value of 100.0Ω at the current moment and a measured value of 102.5Ω at the previous moment, the system calculates that the measured instantaneous change rate is -2.5Ω / ms.

[0199] The measured value was compared with the benchmark and the difference was calculated to be -1.0Ω / ms. At this time, assuming that the noise tolerance set by the system is ±0.5Ω / ms, since the absolute value of the difference of 1.0 directly exceeds the predetermined tolerance band, it indicates that this change is not a reasonable disturbance such as transient high-frequency noise.

[0200] The system then extracts and confirms this deviation of 1.0Ω / ms as the effective value of the resistance nonlinear change rate, and directly pushes it to S3 cascade; thus, the output of S0 is the abnormal characteristic quantity after benchmark comparison;

[0201] In the above processing flow, the input of S0 includes at least three types of data: the first type is the voltage or resistance data collected in real time by the test lead 20; the second type is the sampling clock signal recorded by the controller, which is used to determine the time interval between adjacent sampling points.

[0202] The third category is the normal piezoresistive effect linear benchmark corresponding to the current sample specification; the output of S0 is the nonlinear resistance change rate of each station in the current sampling period. This result is sent to the feature fusion and damage assessment links corresponding to S3 and S4, and is used as the electrical input for calculating the equivalent micro-damage index.

[0203] To avoid amplifying quantization noise during differential calculation, an anomaly confirmation rule can be added to S0. Specifically, when the deviation obtained in a certain sampling period occurs only once and recovers to within the noise tolerance in the next sampling period, the controller marks it as an instantaneous disturbance instead of directly treating it as a damage symptom.

[0204] When the deviation exceeds the limit for two or more consecutive sampling periods, it is used as an effective resistance nonlinear change rate input into the subsequent model. The purpose of this rule is to distinguish between single-point spikes caused by transient contact resistance changes and electromagnetic interference and continuous deviations caused by continuous anomalies inside the material, thereby improving the stability of subsequent damage assessment.

[0205] In this embodiment, S0 is mainly responsible for converting the original resistance sampling sequence into usable electrical anomaly features, and S3 is mainly responsible for organizing the electrical anomaly features and the mechanical anomaly features on the servo current side into parallel inputs.

[0206] In other words, S0 completes the preprocessing of resistance data and the initial screening of anomalies, while S3 completes the synchronous extraction and parallel output of multi-source features. Since S0 first performs differential, filtering, benchmark comparison and anomaly confirmation on the original resistance value, S3 will not repeat the same resistance differential process. Instead, it will directly call the resistance nonlinear change rate output by S0 and send it into the model in S4 together with the servo current distortion rate extracted in the same period, thereby avoiding duplicate calculations or overlapping responsibilities between steps.

[0207] In this embodiment, the normal piezoresistive effect linear reference refers to the expected response trajectory of the resistance change rate of the thermistor 26 sample as displacement, load, or time progresses when no micro-damage occurs and the pressure state changes stably; after the controller calls this reference, it actually judges whether the measured change deviates from the normal pressure behavior.

[0208] S0 uses the normal piezoresistive effect of the thermistor 26 under pressure as a reference, and extracts the part of the change that continues to exceed the reference as an electrical anomaly caused by internal stress concentration, abnormal grain boundary slip, or microcrack initiation. Through this reference-based processing, the input received by the subsequent model has a more direct causal relationship with the internal state changes of the thermistor 26.

[0209] The electrodynamic micro-damage calculation model is built into the controller;

[0210] The steps between S4 and S5 include:

[0211] Calculate the cumulative integral over time of the deviation of the nonlinear rate of change of resistance from the normal piezoresistive effect linear reference.

[0212] The equivalent micro-damage index is calculated by multiplying the cumulative integral by a pre-calibrated material brittleness coefficient.

[0213] In this embodiment, the electrodynamic micro-damage calculation model is built into the controller. The controller stores the normal piezoresistive effect linear reference and material brittleness coefficient corresponding to the thermistors 26 of different specifications. The so-called normal piezoresistive effect linear reference refers to the reference response relationship that the nonlinear change rate of the resistance of the thermistor 26 should satisfy in the undamaged elastic compression range.

[0214] This benchmark can be obtained through pre-calibration. During calibration, multiple batches of undamaged samples are selected, and the relationship between their resistance change rate and pressure or displacement is measured under low-risk loading conditions. The results are then fitted to form a benchmark curve, piecewise linear function, or lookup table.

[0215] The so-called material brittleness coefficient refers to the calibration coefficient that reflects the difference in sensitivity of different thermistor material systems to the same deviation amount. Its value can be obtained based on the statistical analysis of sample destructive tests and non-destructive tests.

[0216] To ensure that the equivalent micro-damage index obtained by mapping is a dimensionless index, the material brittleness coefficient has a dimension that cancels out the dimension of the cumulative integral of resistance, and its actual dimension is the reciprocal of resistance.

[0217] Between steps S4 and S5, the controller calculates the cumulative integral over time of the deviation of the resistance nonlinearity rate of change from the normal piezoresistive effect linear reference. Specifically, in each sampling period, the controller reads the current resistance nonlinearity rate of change and compares it with the reference rate of change corresponding to the current load or displacement to obtain the deviation. To implement and rigorously define the above noise removal logic using mathematical form, an effective deviation is introduced. It satisfies the following piecewise function:

[0218]

[0219] in, The set noise threshold; cumulative integral value Update according to the following discrete integral formula:

[0220]

[0221] in, This represents the cumulative integral value at the end of the current sampling period. For the aforementioned effective deviation, This represents the cumulative integral value retained at the end of the previous sampling period. The time interval between the two adjacent samplings mentioned above; to avoid over-response caused by occasional interference, an upper limit can be set for the deviation, or the integral can be included only after the abnormal conditions are met in multiple consecutive sampling periods.

[0222] The equivalent micro-damage index is calculated by multiplying the cumulative integral by a pre-defined material brittleness coefficient; if the material brittleness coefficient is denoted as... The accumulated points are When there is a joint correction for servo current distortion rate, the controller calculates according to the following explicit rules:

[0223] Based on the low, medium, or high level of the servo current distortion rate, corresponding mechanical correction coefficients are matched. For example, low-level correspondence It is 1.0, corresponding to the medium level. The value is 1.2, corresponding to higher levels. The value is 1.5; based on the equivalent microscopic damage index. The calculation formula is as follows:

[0224]

[0225] Perform calculations; correction terms This is used to reflect mechanical resistance anomalies that cannot be fully revealed by simple resistance changes, and amplifies high-risk conditions through a product relationship; the calculated equivalent micro-damage index The standardized range of 0 to 1 or 0 to 100 is linearly converted and truncated, which facilitates the unified threshold judgment between different batches and different specifications of components.

[0226] The calculation logic outputs results quickly within each control cycle, meeting the real-time requirements of the system. The purpose of this implementation is to convert the short-term abnormal electrical performance of the thermistor 26 during the pressure process into a damage characterization quantity with cumulative significance. The deviation at a single moment may come from noise, contact disturbance or local non-uniformity of materials, while the integral value accumulated over time can more stably reflect the continuous abnormal trend.

[0227] After mapping by combining the material brittleness coefficient, the damage risk corresponding to the same integral value under different material systems can be distinguished; the controller outputs the equivalent micro-damage index to the threshold judgment module in S5, and adjusts the injection current of the corresponding excitation coil 15 accordingly. At the same time, in conjunction with the servo feed pulse frequency adjustment in S6, active protection and detection depth control are achieved during the pressure resistance detection process of the thermistor 26.

[0228] The specific processing flow of the electrodynamic micro-damage calculation model includes the following sequence;

[0229] Step 1: The model reads the resistance nonlinearity rate of change, servo current distortion rate, current displacement segment number, and sample specification number of the current station in the current control cycle from the controller buffer.

[0230] Step 2: The model calls the pre-stored normal piezoresistive linear reference, material brittleness coefficient, and noise threshold corresponding to the sample specification number.

[0231] Step 3: The model compares the current nonlinear rate of change of resistance with the called linear benchmark to obtain the abnormal deviation of the current cycle, and first completes noise removal.

[0232] Step 4: The model performs cumulative integration on the abnormal deviations of each consecutive period to obtain the integral quantity reflecting the degree of continuous abnormality.

[0233] Step 5: The model reads the servo current distortion rate of the current cycle and determines whether the mechanical abnormality is at a low, medium, or high level.

[0234] Step 6: The model combines the accumulated integral with the material brittleness coefficient to obtain the basic damage value, and then corrects the basic damage value by increasing, keeping it unchanged, or slightly amplifying it according to the level of servo current distortion rate.

[0235] Step 7: The model restricts the corrected results to a preset standardized range and outputs them as the final equivalent micro-damage index to S5 and S6.

[0236] The model completes the data flow and output through the above steps; the material brittleness coefficient can be determined by offline calibration; specifically, it can be done by selecting 26 thermistor samples of the same material system and specifications, and synchronously recording the abnormal resistance deviation, servo current distortion characteristics, and whether irreversible electrical performance drift or cracking occurs under low-speed progressive loading conditions; a representative interval is statistically calculated according to the integral value corresponding to the damage of the sample.

[0237] Then, the range is matched with the actual damage results, and the coefficient value that makes the equivalent micro-damage index of the undamaged sample mostly fall below the safety threshold and the equivalent micro-damage index of the damaged sample mostly exceed the safety threshold is selected as the material brittleness coefficient of the sample of this specification.

[0238] The logical function of this coefficient is to convert the same abnormal integral into a unified damage scale suitable for comparison of different material systems.

[0239] The role of the servo current distortion rate in the model is not to replace the resistance anomaly integral, but to correct its mechanical side credibility.

[0240] Specifically, when the abnormal deviation of the resistance is large while the servo current distortion rate remains at a low level, the controller can determine that the abnormality is more likely to come from electrical measurement contact disturbance or transient noise, and therefore outputs only the basic damage value.

[0241] When the abnormal deviation of resistance and the distortion rate of servo current increase simultaneously, it indicates that electrical abnormality and mechanical resistance abnormality occur at the same time, and the two corroborate each other. At this time, the basic damage value is amplified and corrected.

[0242] When the abnormal deviation of the resistance is small but the distortion rate of the servo current suddenly increases, the controller can first record this situation as a mechanical side warning and increase the attention of subsequent sampling, instead of directly identifying it as high damage; the above logic makes the joint judgment of the model have a clear causal chain;

[0243] To facilitate engineering implementation, the servo current distortion rate can be extracted using a fixed time window comparison method. The controller extracts a segment of servo current sampling data in each control cycle, performs spectral decomposition or digital filtering on this segment of data, and extracts the total amplitude of the subharmonic components. Then, this total amplitude is compared with the reference total amplitude pre-recorded under the conditions of the same rotational speed, the same feed pulse frequency, and no workpiece or standard workpiece.

[0244] If the difference between the two falls within the baseline fluctuation range, the distortion rate is determined to be low; if it exceeds the baseline fluctuation range but does not reach the obvious abrupt change range, it is determined to be medium; if it exceeds the range significantly for multiple consecutive periods, it is determined to be high. The model uses the classification result for correction, maintaining clear calculation rules.

[0245] The electrodynamic micro-damage calculation model is used to infer whether the current loading has approached the risk of irreversible crush damage when the micro-cracks, local instability of grain boundaries or stress concentration propagation process inside the thermistor 26 cannot be directly observed in real time. It utilizes online-obtainable resistance anomaly information and driving load anomaly information.

[0246] The model logically comprises three interconnected processing layers:

[0247] The first processing layer is the electrical anomaly accumulation layer, which receives the nonlinear resistance change rate after processing by S0 and S3 and outputs the integral of the anomaly deviation.

[0248] The second processing layer is the mechanical anomaly correction layer, which receives the servo current distortion rate level and outputs the credibility correction result of the aforementioned integral quantity.

[0249] The third processing layer is the damage standardization output layer, which receives the material brittleness coefficient, sample specification parameters and the aforementioned correction results, and outputs an equivalent micro-damage index that can be directly compared with a preset safety threshold.

[0250] Correspondingly, the data flow is as follows: the original resistance sampling value is first converted into an abnormal deviation, and then the cumulative integral is formed; the original servo current sampling value is first converted into a distortion level, and then the cumulative integral is corrected; combined with the material brittleness coefficient and the standardization range limit, a control quantity is formed for S5 and S6 to call.

[0251] The overall physical relationship represented by the model can be defined as follows: when the thermistor 26 is under pressure, if the internal material is still in the stable elastic response range, the resistance change should evolve along the normal piezoresistive effect linear reference, and the reaction force on the driving mechanism 5 changes relatively smoothly.

[0252] If microscopic damage begins to occur in a local area, on the one hand, the resistance change will deviate from the normal reference, and on the other hand, the load fluctuation of the servo motor 6 will increase due to changes in contact state and local load path, thus enhancing the subharmonic characteristics of the drive current.

[0253] The model is an abstract representation of the causal relationship between this internal stress anomaly and the electrical and mechanical sides. Since the electrical anomaly reflects the change in the internal conductive path or piezoresistive response of the material, and the mechanical anomaly reflects the change in the end resistance and contact state, the combination of the two can better reflect the real damage risk than a single channel.

[0254] The following supplementary quantitative derivation examples illustrate the model data flow process: Let the deviation for a certain sampling period be... And the sampling time interval is The deviation increment contribution calculated in this period is: Assume that the cumulative integral value has reached 0.05 after continuous accumulation;

[0255] The material brittleness coefficient of the target batch of thermistors 26, as pre-calibrated, is also known. for Furthermore, the servo current distortion rate extracted within the same period falls into the medium-level feature delineation range that requires key early warning. At this time, the mechanical correction coefficient M is preset to 1.2. Substituting it into the calculation logic, the basic damage value is 12.0 multiplied by 0.05 to calculate 0.6. Constrained by the mechanical correction link, it is multiplied by 1.2 again, and finally the equivalent micro damage index D is calculated to reach 0.72.

[0256] Considering the lower limit of the S5 preset safety threshold of 0.80, this proves that the current D value has not yet exceeded the threshold to trigger the system protection action; however, if the accumulated integral continues to grow to the level of 0.06 in the next cycle, the calculation process will immediately update the D value to 12.0 multiplied by 0.06 multiplied by 1.2, which equals 0.864. Greater than the threshold The controller issues a command to reduce the excitation current in order to perform a protection action;

[0257] In this embodiment, the accumulated integral is reset or frozen in the following situations: when the current detection cycle ends, the main pressure plate 8 is lifted and the next test piece is entered, the integral quantity corresponding to the previous test piece is cleared to zero.

[0258] When a certain workstation returns to the noise tolerance and remains stable within multiple consecutive control cycles, the integral value of that workstation can remain unchanged and not continue to increase; when any workstation reaches the shutdown threshold and triggers the protection shutdown, the integral value of that workstation is frozen to the current value, which is used as the risk record result of this detection.

[0259] With the above limitations, the cumulative integral corresponds one-to-one with a single test piece and a single loading process, so that its numerical meaning is consistent with the degree of abnormal accumulation during the current pressure process, avoiding logical confusion caused by inheritance across workpieces or across detection cycles.

[0260] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of this application without departing from the spirit and scope of the claims, and all of these forms are within the protection scope of this application.

Claims

1. A multi-station withstand voltage testing device for thermistors, characterized in that, include: The test base (1) serves as the main frame supporting the overall structure. The test base (1) includes a top plate (2), side walls (3), and a bottom plate (4). The drive mechanism (5) is located on the top plate (2) of the test base (1). The drive mechanism (5) includes a servo motor (6) and a ball screw (7) driven by the servo motor (6). The main pressure plate (8) is connected to the ball screw (7) and slidably connected to the side wall (3) of the test base (1). The main pressure plate (8) is provided with a self-equalizing fluid network (9). When the servo motor (6) is working, the main pressure plate (8) moves in the vertical direction. The self-equalizing fluid network (9) includes interconnected fluid static pressure chambers (10) and throttling microchannels (11). The test head (12) is respectively located below the main lower pressure plate (8) corresponding to the fluid static pressure chamber (10). A magnetorheological elastomer layer (13) is provided at the bottom of the test head (12). A conductive microfilament array (14) is provided inside the magnetorheological elastomer layer (13). An excitation coil (15) is wound around the periphery of the test head (12). The lower electrode support platform (16) is set on the bottom plate (4) of the test base (1) corresponding to the test head (12). The controller is connected to control the operation of the drive mechanism (5) and the excitation coil (15).

2. The multi-station pressure testing device for a thermistor (26) according to claim 1, characterized in that, The hydrostatic chamber (10) is assembled in the blind hole (17) pre-machined on the bottom surface of the main pressure plate (8). A corrugated elastic diaphragm (18) is sealed at the bottom opening of the hydrostatic chamber (10). The lower surface of the corrugated elastic diaphragm (18) is connected to the test pressure head (12).

3. The multi-station pressure testing device for a thermistor (26) according to claim 2, characterized in that, The fluid static pressure chamber (10) and the throttling microchannel (11) are filled with incompressible dimethyl silicone oil (19).

4. The multi-station pressure testing device for a thermistor (26) according to claim 1, characterized in that, The top end of the conductive microfilament array (14) is electrically connected to the test wire (20) inside the test head (12), wherein the bottom end of the conductive microfilament array (14) is exposed on the lower surface of the magnetorheological elastomer layer (13).

5. The multi-station pressure testing device for a thermistor (26) according to claim 1, characterized in that, The matrix of the magnetorheological elastomer layer (13) is silicone rubber (21), wherein carbonyl iron powder particles (22) are uniformly dispersed inside the matrix.

6. The multi-station pressure testing device for a thermistor (26) according to claim 1, characterized in that, The output shaft of the servo motor (6) is directly connected to the ball screw (7) via a flexible coupling (23), wherein the nut seat (24) of the ball screw (7) is fixedly connected to the main pressure plate (8).

7. The multi-station pressure testing device for a thermistor (26) according to claim 1, characterized in that, The two sides of the main lower pressure plate (8) are slidably connected to the side wall (3) via linear guide rails (25).

8. A control method, applied to the multi-station pressure testing device for thermistor (26) as described in claim 1, characterized in that, include: S1. Control the servo motor (6) to drive the main lower pressure plate (8) downward to approach the thermistor (26) placed on the lower electrode support platform (16), and control the excitation coil (15) to be in the de-energized state. S2. Control the servo motor (6) to work and apply the rated pressure feed amount, and control the injection of peak current into the excitation coil (15); S3. Extract the nonlinear resistance change rate of the thermistor (26) over time in real time, and extract the subharmonic spectrum characteristics of the drive current of the servo motor (6) to obtain the servo current distortion rate. S4. Input the extracted resistance nonlinear change rate and the servo current distortion rate into the electrodynamic micro-damage calculation model to calculate the equivalent micro-damage index. S5. Determine the relationship between the equivalent micro-damage index and the preset safety threshold. If the equivalent micro-damage index is greater than the preset safety threshold, control the reduction of the injection current of the corresponding excitation coil (15). If the equivalent micro-damage index is less than or equal to the preset safety threshold, maintain the injection current of the corresponding excitation coil (15) unchanged. S6. Based on the average value of the equivalent micro-damage index of all workstations, control and adjust the feed pulse frequency of the servo motor (6); repeat the steps S2 to S6 above.

9. The control method according to claim 8, characterized in that, The test head (12) is provided with a test wire (20) connected to the conductive microfilament array (14); before step S3, the following steps are included: S0, continuously recording the resistance value of the thermistor (26) through the test wire (20), and subtracting the resistance value at the current moment from the resistance value at the previous sampling moment, and dividing the difference by the sampling time interval to obtain the nonlinear change rate of resistance.

10. The control method according to claim 8, characterized in that, The electrodynamic micro-damage calculation model is built into the controller; The step between S4 and S5 includes: calculating the cumulative integral of the deviation of the nonlinear rate of change of resistance from the normal piezoresistive effect linear reference over time. The equivalent micro-damage index is calculated by multiplying the cumulative integral by a pre-calibrated material brittleness coefficient.

Images