Non-Contact Degradation Evaluation for IGBT Modules Using Eddy Current Pulsed Thermography Approach

energies

Article Non-Contact Degradation Evaluation for IGBT Modules Using Eddy Current Pulsed Thermography Approach

Xingliang Liu 1 , Guiyun Tian 2, Yu Chen 1, Haoze Luo 1, Jian Zhang 1,* and Wuhua Li 1 1 College of Electrical Engineering, Zhejiang University, Hangzhou 310027, China; [email protected] (X.L.); [email protected] (Y.C.); [email protected] (H.L.); [email protected] (W.L.) 2 College of Electrical and Electronic Engineering, Newcastle University, Newcastle upon Tyne NE1 7RU, UK; [email protected] * Correspondence: [email protected]

Received: 10 April 2020; Accepted: 19 May 2020; Published: 21 May 2020

Abstract: In this paper, a non-contact degradation evaluation method for insulated gate bipolar transistor (IGBT) modules is proposed based on eddy current pulsed thermography approach. In non-contact heat excitation procedures, a high-power induction heater is introduced to generate heat excitation in IGBT modules. The thermographs of the whole temperature mapping are recorded non-invasively by an IR camera. As a result, the joint degradation of IGBT modules can be evaluated by the transient thermal response curves derived from the recorded thermographs. Firstly, the non-destructive evaluation principle of the eddy current pulsed thermography (ECPT) system for an IGBT module with a heat sink is introduced. A 3D simulation module is built with physical parameters in ANSYS simulations, and then thermal propagation behavior considering the degradation impact is investigated. An experimental ECPT system is set up to verify the eﬀectiveness of the proposed method. The experimental results show that the delay time to peak temperature can be extracted and treated as an eﬀective indicative feature of joint degradation.

Keywords: eddy current pulsed thermography (ECPT); joint degradation; non-contact evaluation; transient thermal response curves

1. Introduction With continuous innovation and development of power semiconductor techniques, the insulated gate bipolar transistors (IGBTs) are representative of a fully controlled electronic device, which is widely used in motor drives, renewable energy generation systems, etc. [1,2]. The main trends of modern IGBT modules are higher switching frequency, smaller chip footprint, and higher operating temperature [3]. These trends challenge the reliability design of IGBT modules. Based on the classic bathtub curve of failure rate of power devices over their lifetime, the failure period consists of three main stages: early failure, middle random failure, and the wear-out failure period [4]. In the ﬁrst two stages, the failure factors are mainly concerned with manufacturing (e.g., edge emitter design) and unexpected excessive stress (e.g., short circuit current). Owing to the features of slowing propagation and aging regularity, wear-out failure is most likely to be monitored and predicted during long term service [5]. Since joint degradation failures are aﬀected by the internal thermal stress (due to power loss) and external temperature changes, these failures are of academic and industrial interests. In terms of power conversion systems, typical joint degradation failures include delamination and random voids, particularly in the solder and thermal grease layer [6,7]. Generally, the main reasons for joint

Energies 2020, 13, 2613; doi:10.3390/en13102613 www.mdpi.com/journal/energies Energies 2020, 13, 2613 2 of 14 degradation failures are mismatched coefficients of thermal expansion (CTE) and dry-out of the thermal grease [8,9]. Joint degradation failure is a significant long-term failure mode and negatively impacts the effective connection area between layers and the thermal emission path. As a result, the maximum operating temperature and peak output power decline after a long-running period [10]. Traditional detection methods for joint degradation are mainly based on monitoring the degradation of sensitive electrical parameters (DSEPs) and junction temperature (Tj). Parameter variation caused by degradation can be reflected and extracted from external electrical parameters. An inspected IGBT module in a power circuit can be taken as a resistor-inductor-capacitor (RLC) network during switching transitions, where the equivalent RLC network is dependent on the joint conditions of the multilayer structure. It has been reported that collector voltage vce,on under small current [11–13] and low order harmonics of the output voltage [14] are effective DSEPs for RLC degradation detection. However, additional auxiliary and sampling circuits for DSEPs measurement are needed during testing. The surface Tj measurement is a more direct way of degradation detection. Degradation failures like delamination, cracks and voids can worsen heat emission ability. The measured Tj increases with degradation under the same test conditions. In general, the accurate Tj measurement methods are mainly divided into the temperature sensitive electrical parameter (TSEP) methods [15,16] and optical methods [17,18]. For the TSEP method, high accuracy Tj can be extracted non-invasively. However, every TSEP needs a delicate calibration procedure and relevant calibration circuits. Furthermore, inside the package, parasitic parameter changes caused by the aging effects, introduce measurement errors. The advantage of an IR camera is the possibility to obtain the global temperature of the power device. Nevertheless, the accuracy of an IR camera is affected by the emissivity of the chip surface. In order to get more accurate results, the silica gel within the inspected IGBT module should be removed, and thermal black paint needs to be deposited on the chip surface [19]. As mentioned, the two critical steps in degradation detection are the excitation process and the measurement process. However, the excitation and measuring processes are mostly physical contact methods. These physical contact methods are limited by field-environment and maintenance time. The ideal defect identification method should ensure the normal operation is not interrupted. Recently, the idea of non-destructive and non-contact evaluation test methods for field application has been reported [20,21]. This paper proposes a non-contact test method for joint degradation detection, based on an eddy current pulsed thermography (ECPT) technique. Because of the high conductivity of the metal elements in the heat-sink, copper, and solder, the inspected IGBT module can be non-contact heated by using an inductor heater. Then, the global thermography can be recorded by an IR camera and analyzed by pulsed thermography (PT) technique. As a result, the degradation process characteristics can be extracted and evaluated from pulse thermal transient response curves. This paper is organized as follows. In Section2, the thermal resistor-capacitor (RC) network degradation in IGBT modules is analyzed. In Section3, the ECPT system operational principle and defect identification method is proposed to evaluate the thermal RC network defect. In Section4, an ECPT platform is built for simulation and experimental verification. The last section summarizes the conclusions drawn from the investigation.

2. Thermal RC Network Degradation and Detection

2.1. Characteristic of Degradation in an IGBT Module The emerging base-plate free module is studied in this section. Because it is free from base plate, it can save cost and reduce the thermal resistance. In Figure1, the cross-section of a base-plate free module with heat-sink is shown. The IGBT module consists of several layers of diﬀerent materials, and the silicon chip and baseplate are soldered together by the so-called direct bonded copper (DBC) method. In order to obtain better heat conduction, it is necessary to add a thin and even layer of thermal grease between the IGBT module and the heat sink. Energies 2020, 13, x FOR PEER REVIEW 3 of 14 random voids in the multi-layer structure, as illustrated in Figure 2. Owing to the accumulative thermo-mechanical stresses of power and thermal cycling, the outmost corner is the most stressed region [23]. The identified lifetime limiting delamination is expected to start at the corners and then spread along the entire boundary. More seriously, they will Energiescontinue 2020 to, 13 ,enlarge x FOR PEER with REVIEW continued operation cycling. In general, this degradation reduces3 of the 14 effective connection area between two layers and can be represented by lumped thermal RC network randomEnergies 2020 voids, 13, 2613in the multi-layer structure, as illustrated in Figure 2. 3 of 14 parameters [24]. Owing to the accumulative thermo-mechanical stresses of power and thermal cycling, the outmost corner is the most stressed region [23]. The identified lifetime limiting delamination is expected to start at the corners and thenabcde spread along the entired f boundary. More seriously, they will continue to enlarge with continued operation cycling. In general, this degradation reduces the effective connection area between two layers and can be represented by lumped thermal RC network parameters [24].

abcded f d g h

Figure 1. Cross-sectional view of base-plate free insulated gate bipolar transistor (IGBT) module with a Figure 1. Cross-sectional view of base-plate free insulated gate bipolar transistor (IGBT) module with heat-sink: (a) Aluminum metallization, (b) Chip, (c) Bond wire, (d) Copper pad,d (e) Solder, (f) Ceramic a heat-sink: (a) Aluminum metallization, (b) Chip, (c) Bond wire, (d) Copper pad, (e) Solder, (f) subtracts, (g) Thermal grease, and (h) Heat-sink. Ceramic subtracts, (g) Thermal grease, and (h) Heat-sink. g Operational IGBT modules suﬀer from various thermal stresses, not onlyh internal power losses but also external temperature cycling. The identiﬁed lifetime limiting degradation failures are associated withFigure the solder 1. Cross-sectional joints. Also, view dry-out of base-plate failure free mechanisms insulated gate are bipolar associated transistor with (IGBT) the module thermal with grease layera [ 22heat-sink:]. The most (a) Aluminum common characteristicmetallization, degradations(b) Chip, (c) Bond with delaminationwire, (d) Copper are pad, internal (e) Solder, random (f) voids in theCeramic multi-layer subtracts, structure, (g) Thermal as illustrated grease, and in ( Figureh) Heat-sink.2.

(a) (b)

Figure 2. Joint characteristic degradation in multi-layer structure: (a) defect free connection; (b) rough surfaces with gradation defect.

2.2. Degradation Thermal Modeling (a) (b) IGBT model thermal behavior can be determined by geometrical and physical approaches as follow:Figure 2. JointJoint characteristic characteristic degradation degradation in in multi-layer multi-layer structure: structure: ( (aa)) defect defect free free connection; connection; ( (b)) rough rough surfaces with gradation defect. =⋅ Rdth H/(λ thS A )  =⋅⋅ ⋅ 2.2. DegradationOwing to the Thermal accumulative Modeling thermo-mechanical  Ccthρ stressesdS H A of power and thermal cycling, the outmost(1) corner is the most stressed region [23]. The identifiedτ =⋅RC lifetime limiting delamination is expected to IGBT model thermal behavior can be determinedth th thby geometrical and physical approaches as start at the corners and then spread along the entire boundary. More seriously, they will continue to follow: enlargein which with Rth continuedand Cth are operation the associated cycling. thermal In general, resistance this degradation and thermal reduces capacitance the effective of each connection layer. dH areaand betweenSA are the two thickness layers andand cancross-section be represented area=⋅ of by ea lumpedch layer. thermal In order RC to network obtain the parameters values of [ 24Rth]. and Rdth H/(λ thS A ) Cth, the material property of each layer, like thermal conductivity λth, specific heat c, and material  Cc=⋅⋅ρ dS ⋅ (1) 2.2.density Degradation ρ, are necessary. ThermalModeling τth is the thermal timeth constant H of A a layer.  =⋅ The studied module is SK75GB12T4T fromτth Semikron,RC th th where the layout of the unpackaged IGBT IGBT model thermal behavior can be determined by geometrical and physical approaches as follow: module is seen in Figure 3. There are four 75 A silicon chips (T1, T2 are IGBT dies and D1, D2 are diode in which Rth and Cth are the associated thermal resistance and thermal capacitance of each layer. dH dies) soldered onto two copper plates with =different( areas.) The transparent encapsulation (silicone and SA are the thickness and cross-section R areath dofH /eachλth layer.SA In order to obtain the values of Rth and gel) on top of the chips can protect from moisture, powder,· and besmirch in the operation and it also Cth, the material property of each layer, likeCth =thermalc ρ d conductivityH SA λth, specific heat c, and material(1) provides electrical insulation. Because of its positi· on· in ·the outermost layer and the relatively small density ρ, are necessary. τth is the thermal timeτ constant= R C of a layer. th th · th thermalThe conductivitystudied module of silicone is SK75GB12T4T gel (0.2 W/m·K), from Semikron, the encapsulation where the is layout not considered of the unpackaged in the following IGBT in which R and C are the associated thermal resistance and thermal capacitance of each layer. d and module is seenth in Figureth 3. There are four 75 A silicon chips (T1, T2 are IGBT dies and D1, D2 are diodeH dies)SA are soldered the thickness onto two and copper cross-section plates areawith of different each layer. areas. In orderThe tr toansparent obtain the encapsulation values of Rth (siliconeand Cth, gel)the materialon top of property the chips of can each protect layer, from like thermalmoisture, conductivity powder, andλth besmirch, specific heatin thec, operation and material and density it also providesρ, are necessary. electricalτth insulation.is the thermal Because time of constant its positi ofon a layer.in the outermost layer and the relatively small thermalThe conductivity studied module of silicone is SK75GB12T4T gel (0.2 W/m·K), fromSemikron, the encapsulation where the is not layout considered of the unpackaged in the following IGBT module is seen in Figure3. There are four 75 A silicon chips (T 1,T2 are IGBT dies and D1,D2 are diode

Energies 2020, 13, 2613 4 of 14 dies) soldered onto two copper plates with diﬀerent areas. The transparent encapsulation (silicone gel) on top of the chips can protect from moisture, powder, and besmirch in the operation and it also providesEnergies 2020 electrical, 13, x FOR insulation. PEER REVIEW Because of its position in the outermost layer and the relatively small4 of 14 thermal conductivity of silicone gel (0.2 W/m K), the encapsulation is not considered in the following · simulations. AsAs a a result, result, the the thermal thermal network network degradation degradation caused caused by the by encapsulation the encapsulation layer is ignored,layer is andignored, almost and all almost the temperature all the temperature mapping changes mapping can changes be attributed can be to theattributed solderlayer to the degradation solder layer in thisdegradation work. in this work.

Figure 3. Image of the IGBT module under study.

The properties and thickness of each module layer and the thermal grease grease are are listed listed in in Table Table 11,, whichwhich mainlymainly referrefer toto thethe manualmanual [[25]25] obtainedobtained fromfrom thethe oofficialﬃcial websitewebsite ofof Semikron.Semikron.

Table 1.1. Layer thickness and material propertiesproperties ofof IGBTIGBT inin simulation.simulation.

Designation and Thickness Density Thermal ConductivityThermal Specific Heat DesignationMaterial and Thickness(µm) (kgDensity/m3) (W/m K) CapacitySpecific (J/kg K)Heat Conductivity· · Material (µm) (kg/m3) Capacity (J/kg·K) Silicon chip 300 2330(W/m·K) 148 712 SiliconDie attach chip solder 300 50 7400 2330 57 148 226 712 Up copper layer 300 8960 390 390 Die attach solder 50 7400 57 226 Ceramic-Al2O3 700 3780 24 830 Up Downcopper copper layer layer 300 300 8960 8960 390 390 390 390 Ceramic-AlThermal2O grease3 700 50 2000 3780 0.67 24 300 830 Down copper layer 300 8960 390 390 ThermalThe RC grease thermal network 50 values can be 2000 calculated by the 0.67lumped RC thermal method. 300 Hence, since most of the heat flows to the heat-sink, the studied module can be built as a one-dimensional model.The The RC thermal thermal RC network network values values can are alteredbe calculat aftered long by termthe lumped service. RC The thermal variable method. valued thermalHence, RCsince network, most of considering the heat flows the delaminationto the heat-sink, degradation the studied factors, module is plotted can be in built Figure as 4a, one-dimensional where the major degradationmodel. The thermal parts are RC considered network tovalues be the are Die-DBC altered solderafter long and term the thermal service. greaseThe variable layer. Thevalued RC networksthermal RC are network, corresponding considering to the the definitions delamination shown de ingradation Figure1 and factors, the values is plotted are extractedin Figure from4, where the materialthe major properties. degradation The parts effective are interfaceconsidered area to decreases be the Die-DBC with increased solder and fatigue the damagethermal accumulation. grease layer. TheThe RC thermal networks capacitance are corresponding is related toto materialthe definitions volume, shown whilst in Figure thermal 1 and resistance the values is related are extracted to the interfacefrom the areamaterial between properties. layers. ForThe jointeffective degradation interface like area delamination decreases with and randomincreased voids, fatigue the damage change inaccumulation. thermal capacitance The thermal is significantly capacitance less is than related the thermalto material resistance volume, change. whilst thermal resistance is relatedIn Figureto the5 interface, the thermal area time between constant layers. under For di jointfferent degradation connection like area delamination of the thermal and grease random layer andvoids, die-attach the change solder in thermal layer, arecapacitance plotted. Theis significantly thermal time less constant than the increasesthermal resistance with the reductionchange. in the connection area. The thermal time constant of thermal grease is higher than that of the die-attach solder layer.

Energies 2020, 13, x FOR PEER REVIEW 5 of 14

Reduction trends of effective connection area

Rth1=0.0029(K/W) Energies 2020, 13, 2613 Cth1=0.0251(J/W) 5 of 14 Tj Rth2=0.0015(K/W) Energies 2020, 13, x FOR PEER REVIEW Tc 5 of 14 Cth2=0.5242(J/W) Rth3=0.0194(K/W) Rth1 Rth2 Rth3 Rth4 Rth5 Cth3=3.2943(J/W) Cth1 Cth2 Cth3 Cth4 Cth5 R =0.0005(K/W) Ps th4 Cth4=1.5725(J/W) Troom Rth5=0.0498 (K/W) Cth5=0.0450 (J/W) Solder Copper Ceramic Copper Thermal Die-DBC subtracts grease (100% Connection) Reduction trends of effective connection area Figure 4. Equivalent non-constant thermal resistor-capacitor (RC) network considering the Rth1=0.0029(K/W) degradation effect. Cth1=0.0251(J/W) Tj Rth2=0.0015(K/W) Tc In Figure 5, the thermal time constant under different connectionC areath2=0.5242(J/W) of the thermal grease layer R Rth3=0.0194(K/W) and die-attach solder layer,Rth1 are plRotted.th2 TheRth3 thermalRth4 time constantth5 increases with the reduction in the Cth3=3.2943(J/W) Cth1 Cth2 Cth3 Cth4 Cth5 R =0.0005(K/W) connection area. PThes thermal time constant of thermal grease is higherth4 than that of the die-attach solder layer. Cth4=1.5725(J/W) Troom Rth5=0.0498 (K/W) As a result, the thermal time constant of the complete thermal RCCth5=0.0450 networ (J/W)k increases with joint Solder Copper Ceramic Copper Thermal degradation. Accordingly,Die-DBC the transientsubtracts thermal responsegrease curves of (100%the inspected Connection) die are varied with the different degradation stages. As long as the temperature propagation of inspected die can be extractedFigure and 4.4. comparedEquivalentEquivalent with non-constant non-constant the intact die,thermal thermal the agresistor resistor-capacitoring degree-capacitor of thermal (RC) (RC) network RC networks considering can be detected.the degradation effect. effect.

In Figure 5, the thermal time constant under different connection area of the thermal grease layer and die-attach solder layer, are plotted. The thermal time constant increases with the reduction in the connection area. The thermal time constant of thermal grease is higher than that of the die-attach solder layer. As a result, the thermal time constant of the complete thermal RC network increases with joint degradation. Accordingly, the transient thermal response curves of the inspected die are varied with the different degradation stages. As long as the temperature propagation of inspected die can be extracted and compared with the intact die, the aging degree of thermal RC networks can be detected.

(a) (b)

FigureFigure 5.5. Thermal time constant under didifferentﬀerent connectionconnection area:area: ((aa)) reductionreduction ofof thermalthermal greasegrease areaarea

andand ((bb)) reductionreduction ofof die-attachdie-attach connectionconnection area.area.

3. ECPTAs a Based result, Non-Contact the thermal timeDetection constant of Thermal of the complete RC Network thermal Degradation RC network increases with joint degradation. Accordingly, the transient thermal response curves of the inspected die are varied with the3.1. diThermographyﬀerent degradation Technologies stages. As long as the temperature propagation of inspected die can be extracted and compared with the intact die, the aging degree of thermal RC networks can be detected. The thermal propagation rate is changed by the deformation and metamorphism in the multi- 3.layer ECPT structure. Based Non-ContactThermography Detection technology of Thermalhas been RCgradually Network applied Degradation in electronics devices in recent years. Pulse thermography(a) (PT), lock-in thermography (LIT), and the pulse-phase(b) thermography

3.1.(PPT) Thermography are proposed Technologies as the effective algorithms for the analysis of transient thermal response [26–28]. A comprehensiveFigure 5. Thermal comparison time constant of underthese differenttechniques connection has been area: analyzed (a) reduction in terms of thermal of transient grease area thermal The thermal propagation rate is changed by the deformation and metamorphism in the multi-layer responseand (b in) reduction [29]. The of thermography die-attach connection analysis area. methods have been applied to the failure detection in structure. Thermography technology has been gradually applied in electronics devices in recent electronics packaging like the printed circuit boards (PCBs), cracks in the coppers and random voids 3.years. ECPT Pulse Based thermography Non-Contact (PT),Detection lock-in of thermographyThermal RC Network (LIT), and Degradation the pulse-phase thermography (PPT) are proposed as the effective algorithms for the analysis of transient thermal response [26–28]. 3.1.A comprehensive Thermography Technologies comparison of these techniques has been analyzed in terms of transient thermal responseThe thermal in [29]. Thepropagation thermography rate is analysischanged methodsby the deformation have been appliedand metamorphism to the failure in detection the multi- in layerelectronics structure. packaging Thermography like the printed technology circuit has boards been(PCBs), gradually cracks applied in the in coppers electronics and devices random in voids recent in years.soldered Pulse metal-oxide-semiconductor thermography (PT), lock-in field thermogr effect transistorsaphy (LIT), (MOSFETs) and the pulse-phase [29–31]. In this thermography work, a fast (PPT)degradation are proposed detection as methodthe effective based algorithms on the combination for the analysis of inductive of transient eddy currentthermal heating response and [26–28]. the PT Atechnique comprehensive is studied. comparison of these techniques has been analyzed in terms of transient thermal response in [29]. The thermography analysis methods have been applied to the failure detection in electronics packaging like the printed circuit boards (PCBs), cracks in the coppers and random voids

Energies 2020, 13, x FOR PEER REVIEW 6 of 14 in soldered metal-oxide-semiconductor field effect transistors (MOSFETs) [29–31]. In this work, a fast degradationEnergies 2020, 13 detection, 2613 method based on the combination of inductive eddy current heating and6 ofthe 14 PT technique is studied.

3.2. Operation Operation Principle Principle of ECPT System The eddyeddy currentcurrent pulsed pulsed thermography thermography (ECPT) (ECPT) system system is shownis shown in Figure in Figure6, where 6, where the four the main four mainparts parts are: non-contact are: non-contact inductive inductive excitation excitation source, source, IGBT IGBT based based power power converter, converter, infrared infrared camera camera for fortemperature temperature map map measurement, measurement, and and data data processing. processing.

Figure 6. EddyEddy current current pulsed pulsed thermography thermography ( (ECPT)ECPT) system diagram for thermal detection within of IGBT modules.

In power converter systems, the heat-sink material is mainlymainly aluminumaluminum alloy.alloy. The The heat sink volume is significantly signiﬁcantly more than the volume of the metal layers (copper and solder layers) in an IGBT module.module. WhenWhen thethe inspected inspected IGBT IGBT sample sample is placedis placed above above the the excitation excitation coil, coil, most most induced induced heat heatpower power (Ph) comes(Ph) comes from beneathfrom beneath the module the module assembly. assembly. In practice, In practice,Ph is related Ph is torelated material to propertymaterial propertyand induction and induction heater parameters heater parameters and is governed and is governed by: by:

r μf σ ∝ 2 µ f = 0 PIhe2 , σ σ0 (2) Ph Ie σ, σ = 1(+−α TT ) (2) ∝ σ 1 + α(T T0 ) − 0 where Ie is the high-frequency coil excitation current, f is the alternating current (AC) frequency, μ is where Ie is the high-frequency coil excitation current, f is the alternating current (AC) frequency, the permeability of material under test, electrical conductivity σ is the temperature-dependent µ is the permeability of material under test, electrical conductivity σ is the temperature-dependent parameter, and σ0 is the conductivity of material at temperature T0 and α is the temperature coefficient parameter, and σ0 is the conductivity of material at temperature T0 and α is the temperature coefficient of resistivity [27]. The applied AC frequency is related to the heating power and the material property of resistivity [27]. The applied AC frequency is related to the heating power and the material property of the sample. of the sample. The excitation power generated by the induction heater is a series of high, fixed frequency AC The excitation power generated by the induction heater is a series of high, fixed frequency AC pulses and is controlled by adjusting the frequency and duration of AC. The duration of high- pulses and is controlled by adjusting the frequency and duration of AC. The duration of high-frequency frequency AC and the recording time are both controlled and triggered synchronously by the pulse AC and the recording time are both controlled and triggered synchronously by the pulse generator. generator. The induced eddy currents and corresponding resistive heat Q are induced into the The induced eddy currents and corresponding resistive heat Q are induced into the conductive conductive material of the inspected sample. Finally, the whole thermography is captured and the material of the inspected sample. Finally, the whole thermography is captured and the transient transient temperature response of the key area is extracted by the relevant software and MATLAB. temperature response of the key area is extracted by the relevant software and MATLAB. The recorded The recorded thermal videos are then processed for visualization and post-processing, including thermal videos are then processed for visualization and post-processing, including defect identification, defect identification, and electrical/thermal feature parameters extraction. and electrical/thermal feature parameters extraction.

3.3. Heat Heat Conduction Conduction Analysis During the AC excitation period, period, the the Joule Joule heat heatinging and and heat heat diffusion diffusion progress progress simultaneously. simultaneously. When the AC excitation finishes,finishes, the heat didiffusionffusion process dominates the transient temperature

Energies 2020, 13, 2613 7 of 14 Energies 2020, 13, x FOR PEER REVIEW 7 of 14 response. In In general, general, the the heat heat flow ﬂow in each layer, considering the Joule heating and heat didiffusionﬀusion processes, areare givengiven by:by:

∂T∂TTTTλ λλ∂2 ∂∂∂T222∂2T ∂2T λ = =+++thth( ()(,,,)+ + ) + th thqxyztq(x, y, z, t ) (3)(3) ∂t ∂t ρcρc ∂x∂∂∂2xyz222∂y2 ∂z2 ρcρc where TT = T (x,, y,, z, t) is the temperature distribution, and q (x, y, z, t) is thethe internalinternal heatheat generationgeneration per unitunit volume.volume. As a result of joint degradation,degradation, the thermographythermography of transienttransient thermal response curves would be distorted compared with thethe intactintact inspectedinspected sample.sample.

4. Finite Element Simulation and Experimental Results

4.1. Simulation Models and Analysis The correspondingcorresponding 3D 3D model model of the of studiedthe studied IGBT withIGBT heat-sink, with heat-sink, in the Solid in Works the Solid environment, Works isenvironment, shown in Figure is shown7. The in physicalFigure 7. parametersThe physical and parameters thickness and of eachthickness layer of in each the ANSYSlayer in simulationthe ANSYS modulesimulation are module listed inare Table listed1. in The Table ANSYS 1. The simulation ANSYS simulation environment environment is usually is used usually to evaluate used to theevaluate electrothermal the electrothermal eﬀects and effects safe and operating safe oper temperaturesating temperatures for power for modules,power modules, providing providing critical designcritical guidelines.design guidelines.

Figure 7. 3D simulation module of inspected IGBT with heat-sink.

In thethe proposedproposed ECPTECPT method,method, the eddyeddy currentcurrent heatheat inducedinduced isis mainlymainly accumulatedaccumulated at thethe bottom of the module,module, thenthen transferredtransferred toto thethe toptop surface.surface. Therefore, because of its large thermal mass, the heat sink is assigned to be a a heat heat source, source, imposed imposed by by the the pulsed pulsed internal internal heat, heat, in in simulations. simulations. Then, the related related simulation simulation parameters parameters and and bounda boundaryry conditions conditions of of the the ANSYS ANSYS program program are are setset as asfollows: follows: initial initial ambient ambient temperature temperature is 25 is °C, 25 ◦ convecC, convectiontion coefficient coefficient (baseplate (baseplate downside downside face) face) is 5 is× 510⁻⁶10 W/mm6 W/2mm·°C,2 theC ,internal the internal heat generation heat generation is set isto set5 × to10 5⁸ W/m1083, Wand/m the3, and pulse the heat pulse flux heat is a flux square is a × − ·◦ × squarewave for wave 200 forms. 200 ms. In this simulation, the degradation change is modeledmodeled by decreasing the area of thermal grease (e(effectivelyffectively increasingincreasing thethe resistanceresistance associatedassociated withwith greasegrease drying).drying). By this means, the thermal path between the top die and heat-sink becomes worse as the thermal grease areaarea decreases.decreases. The average junction temperaturetemperature ofof diodediode DD11 ((TTjD1jD1)) is observed.observed. The transienttransient thermalthermal responseresponse ofof TTjD1jD1 can be extracted for degradation detection. Figure 88a,a, showsshows threethree varyingvarying degreesdegrees ofof thermalthermal greasegrease layerlayer area, and the related temperature distribution at 1 s with a 100%100% connectionconnection condition is shown in Figure8 8b.b.

Energies 2020, 13, x FOR PEER REVIEW 8 of 14 EnergiesEnergies 20202020,, 1313,, 2613x FOR PEER REVIEW 8 of 14

(a)( a) (b()b ) Figure 8. (a) Three varying degrees of thermal grease layer with different connection area and (b) the Figure 8. (a) Three varying degrees of thermal grease layer with didifferentﬀerent connection area andand (b)) thethe temperature distribution at 1 s with 100% connection area. temperaturetemperature distributiondistribution atat 11 ss withwith 100%100% connectionconnection area.area.

TheThe temperaturetemperature temperature mapmap map of of theof the cross-sectionalthe cross-sectional cross-sectional IGBT IG IGBT module,BT module, module, considering considering considering a 70% a a70% thermal 70% thermal thermal grease grease area,grease isarea,area, shown is isshown shown in Figure in in Figure 9Figure. Since 9. 9.Since there Since there is there an is edge isan an ed disconnection edgege disconnection disconnection between between between the heat-sinkthe the heat-sink heat-sink and and the and the bottom the bottom bottom of theofof the IGBT the IGBT IGBT module, module, module, the the thermalthe thermal thermal path path path at at this atthis this edge edge edge region region region is is the isthe the worstworst worst pathpath path acrossacross across the the cross-sectional. cross-sectional. Accordingly,Accordingly,Accordingly, thethe the highesthighest highest temperaturetemperature temperature appearsappears appears atat the atthe the edges edges edges of of theof the the bottom. bottom. bottom.

(a)( a) (b()b ) FigureFigure 9. 9.Temperature Temperature map map of of cross-sectional cross-sectional IGBT IGBT module module based based on on simulation: simulation: (a )(a 100%) 100% connection connection Figure 9. Temperature map of cross-sectional IGBT module based on simulation: (a) 100% connection andand (b )(b 70%) 70% connection connection area. area. and (b) 70% connection area.

InIn Figure Figure 10 10,, the the transient transient response response curve curve of ofTj Tforj for di differentfferent connection connection areas areas is plotted.is plotted. The The peak peak In Figure 10, the transient response curve of Tj for different connection areas is plotted. The peak junctionjunction temperature temperature at 70% at connection,70% connection,Tjp1 is 33.6Tjp1 is◦C, 33.6 which °C, is which slightly is lower slightly than atlower 100% than connection, at 100% junction temperature at 70% connection, Tjp1 is 33.6 °C, which is slightly lower than at 100% Tjp2connection,34.4 ◦C. However, Tjp2 34.4 the°C. delayHowever, time the to the delay peak time temperature to the peakTd1 temperaturewith a 70% connectionTd1 with a 70% is about connection 1.9 s connection, Tjp2 34.4 °C. However, the delay time to the peak temperature Td1 with a 70% connection longeris about than 1.9 with s longer 100% than connection with 100%Td2. Theconnection peak die Td2 temperature. The peak die is mainly temperature determined is mainly by the determined specific is about 1.9 s longer than with 100% connection Td2. The peak die temperature is mainly determined heat capacity and density of the materials. In the simulation, the thermal grease volume decreases byby the the specific specific heat heat capacity capacity and and density density of of the the materials. materials. In In the the simulation, simulation, the the thermal thermal grease grease with the connection area. So, this reducing substance part can be ignored compared with the IGBT volumevolume decreases decreases with with the the connection connection area. area. So, So, this this reducing reducing substance substance part part can can be be ignored ignored compared compared modulewith the volume. IGBT module Hence, volume. the peak Hence,Tjp1 is the slightly peak lowerTjp1 is slightly than Tjp2 lowerat the than end Tjp2 of at the the thermal end of the transient thermal with the IGBT module volume. Hence, the peak Tjp1 is slightly lower than Tjp2 at the end of the thermal pulse. Although reducing substance can be ignored during the aging process, the thermal path is transienttransient pulse. pulse. Although Although reducing reducing substance substance can can be be ignored ignored during during the the ag aginging process, process, the the thermal thermal significantly changed. pathpath is issignificantly significantly changed. changed. Consequently, the delay time to the peak temperature under pulsed heat power is a better indicator than peak temperature, to evaluate thermal network degradation.

Energies 2020, 13, x FOR PEER REVIEW 9 of 14

Energies 2020, 13, 2613 35 Peak Tjp2 9 of 14 Energies 2020, 13, x FOR PEER REVIEW 9 of 14 33 Delay Td2 C) o 31 35 Peak Tjp2 Delay Td1 Peak Tjp1

3329 Delay Td2 C) o

Temperature( 3127 Delay Td1 Peak Tjp1 25 29 0.5 11.52 Time (s)

Temperature( 27 70% connection area 100% connection area

25 Figure 10. Comparison0.5 of Tj response11.5 curves with varying thermal2 grease area. Time (s) 70% connection area 100% connection area Consequently, the delay time to the peak temperature under pulsed heat power is a better

indicator than Figurepeak temperature, 10. ComparisonComparison to of ofevaluate TTj jresponseresponse thermal curves curves network with with varying varying degradation. thermal thermal grease grease area. area. 4.2. Experimental Result and Discussion 4.2. ExperimentalConsequently, Result the anddelay Discussion time to the peak temperature under pulsed heat power is a better indicatorThe theoreticalthan peak temperature,analysis ofof the to proposed evaluate detectiondetectiothermaln network methodmethod isdegradation.is assessed by the ECPT system shown inin Figure 1111.. 4.2. Experimental Result and Discussion The theoretical analysis of the proposed detection method is assessed by the ECPT system shown in Figure 11.

Figure 11. ConfigurationConﬁguration of the experimental ECPT system.

The specifications speciﬁcations of the proposed ECPT system are shownshown inin TableTable2 2.. TheThe inductioninduction heatingheating source isis EasyheatEasyheat 224224 fromFigurefrom CheltenhamCheltenham 11. Configuration Induction Inductio of then Heating, experimentalHeating, Ltd., Ltd., ECPT Cheltenham, Cheltenham, system. UK, UK, which which provides provides a maximuma maximum excitation excitation power power of of 2.4 2.4 kW kW with with an an excitation excitation frequency frequency range range ofof 140–400140–400 kHz.kHz. TheThe AC excitationThe specifications frequencyfrequency chosen chosen of the in theinproposed the experiment experiment ECPT is system 145 is kHz.145 are AkHz. FLIRshown A SC655FLIR in Table IRSC655 camera 2. TheIR iscamera induction used for is recording usedheating for sourcethermalrecording is videos Easyheat thermal with 224videos a maximum from with Cheltenham a IR maximum resolution Inductio IR of resolution 640n Heating,480 on of a 640×480Ltd., 7.5–14.0 Cheltenham, onµm a InSb7.5–14.0 detector UK, μ whichm andInSb transmitsprovides detector × athemand maximum transmits to PC. Then,excitation them the to temperaturepower PC. Then, of 2.4 the informationkW temperature with an canexcitation information be easily frequency extracted can be range througheasily of 140–400extracted the FLIR kHz. through ResearchIR The ACthe excitationsoftwareFLIR ResearchIR provided frequency software by chosen FLIR provided Systems, in the byexperiment Inc. FLIR (Wilsonville, Systems, is 145 Inc. OR, kHz. USA). A FLIR SC655 IR camera is used for recording thermal videos with a maximum IR resolution of 640×480 on a 7.5–14.0 μm InSb detector and transmits them to PC. Then, Tablethe temperature 2.2. ECPT system information speciﬁcations.specifications. can be easily extracted through the FLIR ResearchIR software provided by FLIR Systems, Inc. ParametersParameters ValueValue Parameters Parameters Value Value Inspected IGBT module SK75GB12T4T Recording duration 60 s Inspected IGBT moduleTable SK75GB12T4T 2. ECPT system specifications. Recording duration 60 s InductionInduction heating heating source source 2.4 2.4 kW kW Infrared Infrared (IR) (IR) camera camera FLIR FLIR SC655 SC655 HeatingHeatingParameters duration duration 200Value 200 ms ms Frame FrameParameters rate rate of of IR IR camera camera Value 25 25 ExcitationExcitationInspected current current IGBT frequency modulefrequency SK75GB12T 140–400 140–400 kHz kHz4T AmbientRecording Ambient temperature temperatureduration 22 60 22 °Cs◦ C Induction heating source 2.4 kW Infrared (IR) camera FLIR SC655

The thermalHeating propagation duration within the IGBT 200 ms module, Frame with the rate 50 of s ofIR recording camera time t 25R, is plotted in Excitation current frequency 140–400 kHz Ambient temperature 22 °C Figure 12. After short term heat excitation (200 ms), the heat begins to transfer from the bottom to the

EnergiesEnergies 20202020,, 1313,, xx FORFOR PEERPEER REVIEWREVIEW 1010 ofof 1414

EnergiesTheThe2020 thermalthermal, 13, 2613 propagationpropagation withinwithin thethe IGBTIGBT module,module, withwith thethe 5050 ss ofof recordingrecording timetime ttR,R, isis plottedplotted10 of 14 inin FigureFigure 12.12. AfterAfter shortshort termterm heatheat excitationexcitation (200(200 msms),), thethe heatheat beginsbegins toto transfertransfer fromfrom thethe bottombottom toto thethe surfacesurface ofof thethe die.die. Then,Then, thethe surfacesurface temperaturtemperaturee stabilizes.stabilizes. OnceOnce thethe averageaverage temperaturetemperature ofof thethe surface of the die. Then, the surface temperature stabilizes. Once the average temperature of the observed observedobserved surfacesurface reachesreaches itsits maximummaximum temperature,temperature, thethe inspectedinspected averageaverage TTj j startsstarts toto declinedecline exponentially.surfaceexponentially. reaches its maximum temperature, the inspected average Tj starts to decline exponentially.

((aa)) ttRR == 00 ss ( (bb)) ttRR == 1010 ss ( (cc)) ttRR == 2020 ss

((dd)) ttRR == 3030 ss ( (ee)) ttRR == 4040 ss ( (ff)) ttRR == 5050 ss

FigureFigure 12. 12. ThermalThermal propagation propagation of of the the inspected inspected IGBT IGBT module module with with an an IR IR camera. camera.

InInIn orderorderorder to toto simulate simulatesimulate thermal thermalthermal resistance resistanceresistance degradation degrdegradationadation between betweenbetween die anddiedie heat-sink,andand heat-sink,heat-sink, the thermal thethe thermalthermal grease greaseareagrease between areaarea betweenbetween the heat thethe sink heatheat and sinksink IGBT andand IGBTIGBT bottom bottombottom is spread isis spreadspread as in asas Figure inin FigureFigure 13. 13.13. The TheThe thermal thermalthermal grease greasegrease area areaarea is isreducedis reducedreduced from fromfrom 100% 100%100% to toto 70% 70%70% as asas the thethe e effectiveﬀeffectiveective connection connectconnectionion area areaareashrinks shrinksshrinks from fromfrom the thethe edge edgeedge towardtoward thethe center center (drying(drying starts starts from from thethe edges).edges). The TheThe volume volumevolume of of thermal thermalthermal grease grease isis about about 0.12 0.12 cm cm33,, and and the the whole whole bottom bottom areaarea is is about about 15.0 15.0 cmcm22.. Therefore,Therefore, thethe thicknessthickness ofof ther thermalthermalmal grease grease is is approximately approximately 80 80 μµμmm when when the the greasegrease is is evenly evenly wiped wiped on on the the bottom bottom for for aa 100%100% connection.connection.

((aa)) ≈≈ 100%100% ( (bb)) ≈≈ 85%85% ( (cc)) ≈≈ 70%70% Figure 13. Configuration of thermal grease with different area. FigureFigure 13.13. ConfigurationConfiguration ofof thermalthermal greasegrease withwith differentdifferent area.area. Because of lift-off impact and an unfixed location under the coil, the test results under the Because of lift-off impact and an unfixed location under the coil, the test results under the same sameBecause conditions of lift-off are similar, impact but and not an identicalunfixed location [27]. Therefore, under the each coil, test the resulttest results is repeated under 10the times.same conditions are similar, but not identical [27]. Therefore, each test result is repeated 10 times. The Theconditions resulting are thermal similar, transient but not responseidentical curves[27]. Theref withore, three each connection test result areas is repeated in 100 s recording 10 times. timeThe resultingresulting thermalthermal transienttransient responseresponse curvescurves withwith threethree connectionconnection areasareas inin 100100 ss recordingrecording timetime areare are demonstrated in Figure 14. It is demonstrated that the measured Tj rises more slowly than the demonstrated in Figure 14. It is demonstrated that the measured Tj rises more slowly than the simulationsdemonstrated shown in Figure in Figure 14. 10 It. Theis demonstrated difference is likely that to the come measured from the thicknessTj rises more and ruggednessslowly than of the simulations shown in Figure 10. The difference is likely to come from the thickness and ruggedness practicalsimulations thermal shown grease, in Figure which 10. could The increasedifference the is thermal likely to resistance come from and the make thickness heat conduction and ruggedness worse. of the practical thermal grease, which could increase the thermal resistance and make heat conduction of theThe practical peak temperaturethermal grease, of thewhich inspected could increase area declines the thermal slightly resistance with decreasing and make connectionheat conduction area. worse. Theworse. delay time to the peak temperature also extends with area reduction. In Figure 14, the trends of an increasing time constant are plotted in 4 s sampling (100 frames). The maximum temperature increment ∆Tj and the corresponding delay time to the peak temperature Tjp with different connection area are plotted in Figure 15. According to the previous analysis, there is a corresponding relationship between the effective connection area and the joint degradation degree. Thus, it is studied that Figure 15a,b illustrates the relationships between ∆Tj and the degradation degree and between Tjp and the degradation degree, respectively.

Energies 2020, 13, x FOR PEER REVIEW 11 of 14

Energies 2020, 13, 2613 11 of 14 Energies 2020, 13, x FOR PEER REVIEW 11 of 14

Figure 14. Comparison of thermal transient response curves with varying thermal grease area in the same test area.

The peak temperature of the inspected area declines slightly with decreasing connection area. The delay time to the peak temperature also extends with area reduction. In Figure 14, the trends of an increasing time constant are plotted in 4 s sampling (100 frames). The maximum temperature increment ΔTj and the corresponding delay time to the peak temperature Tjp with different connection area are plotted in Figure 15. According to the previous analysis, there is a corresponding relationship between the effective connection area and the joint degradationFigureFigure 14.14. degree. Comparison Thus, ofit thermalis studied transient that Figure response 15a,b curvescurv illustrateses with varying the relationships thermal grease between areaarea inin theΔtheT j and the degradationsamesame testtest area.area. degree and between Tjp and the degradation degree, respectively.

The peak temperature of the inspected area declines slightly with decreasing connection area. The delay time to the peak temperature also extends with area reduction. In Figure 14, the trends of an increasing time constant are plotted in 4 s sampling (100 frames). The maximum temperature increment ΔTj and the corresponding delay time to the peak temperature Tjp with different connection area are plotted in Figure 15. According to the previous analysis, there is a corresponding relationship between the effective connection area and the joint degradation degree. Thus, it is studied that Figure 15a,b illustrates the relationships between ΔTj and the degradation degree and between Tjp and the degradation degree, respectively.

(a) (b)

Figure 15. Ten10 times10 times test test results results vary vary with with three three connection connection areas: areas: ( a(a)) peakpeak temperaturetemperature increment; ((b)) delaydelay timetime toto peakpeak temperature.temperature.

Since didifferentfferent connection areas distort the thermalthermal conduction path differently, differently, the peak temperature increment increment and and delay delay time time corresponding corresponding to different to different connection connection areas areas in Figure in Figure 15 show 15 showobvious obvious differences. differences. This means This the means peak thetemperatur peak temperaturee increment incrementand the delay and time the can delay be extracted time can beas two extracted indicators as two for indicatorsthermal degradation for thermal detection. degradation However, detection. due to However, good linearity due to and good sensitivity, linearity andthe peak sensitivity, temperature the peak delay temperature time is a delay better time indica is ator. better Of indicator.course, a calibration Of course, atest calibration between testthe betweentemperature the temperaturemeasurement(a measurement) and the sample and module the sample with module given withaging given degrees(b aging) is required. degrees is Then, required. the Then,multidimensionalFigure the multidimensional 15. Ten 10database times test fordatabase results the varystudied for with the threeIGBT studied connection module IGBT areas: modulecan be (a) peakbuilt can be temperaturefor built the forfollowing increment; the following aging agingevaluation.(b evaluation.) delay time to peak temperature.

4.3. Comparison Since different and Analysis connection areas distort the thermal conduction path differently, the peak temperatureThe detaileddetailed increment comparison comparison and delay between between time the corresponding the degradation degradation to of different sensitive of sensitive electricalconnection electrical parameters areas parameters in Figure based 15 method based show andmethodobvious the ECPT anddifferences. the method ECPT This is method presented means isthe inpresented peak Table temperatur3. The in Ta proposedblee increment3. The ECTP proposed methodand the ECTP odelayffers method atime non-contact can offers be extracted heatinga non- approach,contactas two indicatorsheating by using approach, for an thermal induction by degradation using heater. an Compared inductiodetection.n with However,heater. conventional Compared due to good heating with linearity methods,conventional and heat sensitivity, heating energy couldmethods,the peak not heat betemperature accurately energy could delay controlled not time be accurately likeis a withbetter a cont directindicarolled currenttor. like Of with (DC)course, a powerdirect a calibration current supply. (DC) In test terms power between of supply. a TSEP the detectionIntemperature terms of method, a measurement TSEP the detection measured and method, the die sample temperature the modulemeasured is anwi averagethdie given temperature ofaging the wholedegrees is an inspected averageis required. powerof the Then, device.whole the However,inspectedmultidimensional overallpower temperaturedevice. database However, for mapping the overall studied can temperat be IGBT captured uremodule bymapping an can IR camera, becan built be butcaptured for the the sample byfollowing an rate IR iscamera, loweraging thanevaluation. for the TSEP method. Therefore, the thermal propagation time of the inspected module should be determined appropriately. 4.3. ComparisonDuring field and operation, Analysis the prognostic procedure should be as simple as possible to avoid dismantlingThe detailed of the comparison converter system. between As the a result,degradation being of a non-contactsensitive electrical heating parameters and temperature based measurementmethod and the approach, ECPT method the proposed is presented ECPT systemin Table is an3. The efficient proposed degradation ECTP method detection offers method a non- for powercontact converter heating systems.approach, by using an induction heater. Compared with conventional heating methods, heat energy could not be accurately controlled like with a direct current (DC) power supply. In terms of a TSEP detection method, the measured die temperature is an average of the whole inspected power device. However, overall temperature mapping can be captured by an IR camera,

Energies 2020, 13, 2613 12 of 14

Table 3. Comparison between electrical parameters based method and ECPT method.

Parameters Electrical Parameters Based Method Proposed ECPT Method The heat energy is induced by the The heat energy is supplied by power Heat incentive eddy current in the metal. It is a source, and it is an active heat method. positive heat method. The chip and module temperature are The chip and module temperature measured by some temperature Temperature measurement can be captured by an infrared sensitive electrical parameters (TSEP) camera at one time. or thermal couples. It is necessary to calibrate the Automatic calibration in IR Calibration relationships between electrical camera system. parameters and temperature. The heat excitation and The heat excitation and temperature temperature measurement process Operation measurement process should be in are carried out by a contact with inspected objects. non-contact method.

5. Conclusions In this paper, an eddy current pulsed thermography system has been used for IGBT degradation detection. ECPT offers a non-contact heating approach by exciting the eddy current within the metals of the inspected IGBT module. An IR camera is used for overall temperature measurement. The degradation defect induced by accumulative thermal-mechanical stresses was discussed. The most common degradation characteristic is decreased effective connection area between two layers. As a result, the thermal emission path of IGBT based power converter deteriorates with joint degradation. The thermal time constant of the power converter increases with effective connection area reduction. This joint degradation can be reflected by the changes in the transient thermal response curves. A 3-D IGBT module was established in the ANSYS simulation environment. Accumulative degradation is fulfilled by decreasing the effective connection area of the thermal grease from 100% to 70%. Maximum temperature increment and delay rise time can be extracted as two effective indicators for degradation detection. The peak temperature increment of the inspected chip declines with decreasing connection area. However, the delay time to peak temperature is prolonged with decreasing connection area. The results can be used for IGBT module defect detection, and further quantitative experiments and analysis are to be carried out.

Author Contributions: Conceptualization, G.T. and W.L.; methodology, H.L.; software, X.L. and H.L.; validation, X.L., J.Z. and H.L.; formal analysis, J.Z.; investigation, Y.C.; resources, G.T.; data curation, Y.C.; writing—original draft preparation, X.L.; writing—review and editing, G.T., W.L. and H.L.; visualization, H.L.; supervision, G.T.; project administration, G.T.; funding acquisition, G.T. All authors have read and agreed to the published version of the manuscript. Funding: This work was supported in part by European Commission project HEMOW and in part by the Natural Science Foundation of China under Grant 61960206010. This work was also supported by the Fundamental Research Funds for the Central Universities. Conﬂicts of Interest: The authors declare no conﬂict of interest. Energies 2020, 13, 2613 13 of 14

References

1. Agustin, C.A.; Yu, J.; Lin, C.; Fu, X. A Modulated Model Predictive Current Controller for Interior Permanent-Magnet Synchronous Motors. Energies 2019, 12, 2885. [CrossRef] 2. Ryndzionek, R.; Sienkiewicz, Ł. Evolution of the HVDC Link Connecting Offshore Wind Farms to Onshore Power Systems. Energies 2020, 13, 1914. [CrossRef] 3. Schulze, H.-J.; Niedernostheide, F.-J.; Pfirsch, F.; Baburske, R. Limiting Factors of the Safe Operating Area for Power Devices. IEEE Trans. Electron Devices 2013, 60, 551–562. [CrossRef] 4. Ji, B.; Song, X.; Cao, W.; Pickert, V.; Hu, Y.; Mackersie, J.W.; Pierce, G. In Situ Diagnostics and Prognostics of Solder Fatigue in IGBT Modules for Electric Vehicle Drives. IEEE Trans. Power Electron. 2015, 30, 1535–1543. 5. Ji, B.; Pickert, V.; Cao, W.; Zahawi, B. In Situ Diagnostics and Prognostics of Wire Bonding Faults in IGBT Modules for Electric Vehicle Drives. IEEE Trans. Power Electron. 2013, 28, 5568–5577. [CrossRef] 6. Ciappa, M. Selected failure mechanisms of modern power modules. Microelectron. Reliab. 2002, 42, 653–667. [CrossRef] 7. Bahman, A.S.; Iannuzzo, F.; Uhrenfeldt, C.; Blaabjerg, F.; Munk-Nielsen, S. Modeling of Short-Circuit-Related Thermal Stress in Aged IGBT Modules. IEEE Trans. Ind. Appl. 2017, 53, 4788–4795. [CrossRef] 8. Oh, H.; Han, B.; McCluskey, P.; Han, C.; Youn, B.D. Physics-of-Failure, Condition Monitoring, and Prognostics of Insulated Gate Bipolar Transistor Modules: A Review. IEEE Trans. Power Electron. 2015, 30, 2413–2426. [CrossRef] 9. Katsis, D.C.; van Wyk, J.D. Void-induced thermal impedance in power semiconductor modules: Some transient temperature effects. IEEE Trans. Ind. Appl. 2003, 39, 1239–1246. [CrossRef] 10. Ma, K.; Bahman, A.S.; Beczkowski, S.; Blaabjerg, F. Complete Loss and Thermal Model of Power Semiconductors Including Device Rating Information. IEEE Trans. Power Electron. 2015, 30, 2556–2569. [CrossRef] 11. Smet, V.; Forest, F.; Huselstein, J.; RASHED, A.; Richardeau, F. Evaluation of Vce Monitoring as a Real-Time Method to Estimate Aging of Bond Wire-IGBT Modules Stressed by Power Cycling. IEEE Trans. Ind. Electron. 2013, 60, 2760–2770. [CrossRef] 12. Zhang, J.; Du, X.; Yu, Y.; Zheng, S.; Sun, P.; Tai, H. Thermal Parameter Monitoring of IGBT Module Using Junction Temperature Cooling Curves. IEEE Trans. Ind. Electron. 2019, 66, 8148–8160. [CrossRef] 13. Kong, Q.; Du, M.; Ouyang, Z.; Wei, K.; Hurley, W.G. A Method to Monitor IGBT Module Bond Wire Failure Using On-State Voltage Separation Strategy. Energies 2019, 12, 1791. [CrossRef] 14. Xiang, D.; Ran, L.; Tavner, P.; Yang, S.; Bryant, A.; Mawby, P. Condition Monitoring Power Module Solder Fatigue Using Inverter Harmonic Identification. IEEE Trans. Power Electron. 2012, 27, 235–247. [CrossRef] 15. Avenas, Y.; Dupont, L.; Khatir, Z. Temperature Measurement of Power Semiconductor Devices by Thermo-Sensitive Electrical Parameters—A Review. IEEE Trans. Power Electron. 2012, 27, 3081–3092. [CrossRef] 16. Zhang, J.; Du, M.; Jing, L.; Wei, K.; Hurley, W.G. IGBT Junction Temperature Measurements: Inclusive of Dynamic Thermal Parameters. IEEE Trans. Device Mater. Relib. 2019, 19, 333–340. [CrossRef] 17. Dupont, L.; Avenas, Y.; Jeannin, P. Comparison of Junction Temperature Evaluations in a Power IGBT Module Using an IR Camera and Three Thermosensitive Electrical Parameters. IEEE Trans. Ind. Appl. 2013, 49, 1599–1608. [CrossRef] 18. Hu, Z.; Du, M.; Wei, K.; Hurley, W.G. An Adaptive Thermal Equivalent Circuit Model for Estimating the Junction Temperature of IGBTs. IEEE Trans. Emerg. Sel. Top. Power Electron. 2019, 7, 392–403. [CrossRef] 19. Li, K.; Tian, G.Y.; Cheng, L.; Yin, A.; Cao, W.; Crichton, S. State Detection of Bond Wires in IGBT Modules Using Eddy Current Pulsed Thermography. IEEE Trans. Power Electron. 2014, 29, 5000–5009. [CrossRef] 20. Cheng, L.; Tian, G.Y. Transient Thermal Behavior of Eddy-Current Pulsed Thermography for Nondestructive Evaluation of Composites. IEEE Trans. Instrum. Meas. 2013, 62, 1215–1222. [CrossRef] 21. Bui, H.K.; Wasselynck, G.; Trichet, D.; Ramdane, B.; Berthiau, G.; Fouladgar, J. 3-D Modeling of Thermo Inductive Non Destructive Testing Method Applied to Multilayer Composite. IEEE Trans. Magn. 2013, 49, 1949–1952. [CrossRef] 22. Nnebe, I.M.; Feger, C. Drainage-Induced Dry-Out of Thermal Greases. IEEE Trans. Adv. Packag. 2008, 31, 512–518. [CrossRef] Energies 2020, 13, 2613 14 of 14

23. Shammas, N.Y.A.; Rodriguez, M.P.; Plumpton, A.T.; Newcombe, D. Finite element modelling of thermal fatigue eﬀects in IGBT modules. IEE Proc. Circ. Devices Syst. 2001, 148, 95–100. [CrossRef] 24. Batard, C.; Ginot, N.; Antonios, J. Lumped Dynamic Electrothermal Model of IGBT Module of Inverters. IEEE Trans. Compon. Pack. Manuf. Technol. 2015, 5, 355–364. [CrossRef] 25. Wintrich, A.; Nicolai, U.; Tursky, W.; Reimann, T. Application Manual Power Semiconductors, 2nd ed.; SEMIKRON International GmbH: Ilmenau, Germany, 2015. 26. Bohm, J.; Wolter, K.J. Inductive excited lock-in thermography for electronic packages and modules. In Proceedings of the 33rd International Spring Seminar Electronics Technology (ISSE), Warsaw, Poland, 12–16 May 2010; pp. 190–195. 27. Yin, A.; Gao, B.; Yun Tian, G.; Woo, W.L.; Li, K. Physical interpretation and separation of eddy current pulsed thermography. J. Appl. Phys. 2013, 113, 64101. [CrossRef] 28. Alamin, M.; Tian, G.Y.; Andrews, A.; Jackson, P. Principal Component Analysis of Pulsed Eddy Current Response from Corrosion in Mild Steel. IEEE Sens. J. 2012, 12, 2548–2553. [CrossRef] 29. May, D.; Wunderle, B.; Schacht, R.; Michel, B. Transient thermal response as failure analytical tool—A comparison of diﬀerent techniques. In Proceedings of the 14th International Thermal, Mechanical and Multi-Physics Simulation and Experiments in Microelectronics and Microsystems (EuroSimE), Wroclaw, Poland, 14–17 April 2013; pp. 1–5. 30. Riccio, M.; Irace, A.; Breglio, G. Lock-in thermography for the localization of prebreakdown leakage current on power diodes. In Proceedings of the Ph.D. Research in Microelectronics and Electronics (PRIME), Cork, Ireland, 12–17 July 2009; pp. 208–211. 31. Gershman, I.; Bernstein, J.B. Nondestructive Quantitative Analysis of Crack Propagation in Solder Joints. IEEE Trans. Compon. Pack. Manuf. Technol. 2013, 3, 1263–1270. [CrossRef]

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).