Electronics Guide

Failure Modes and Effects Analysis

Failure modes and effects analysis (FMEA) works from the bottom up. Each component or function is examined in turn, its plausible failure modes are listed, and the local and system-level effects of each mode are traced. Adding criticality ranking produces FMECA, which prioritizes modes by the combination of severity and likelihood. IEC 60812 is the current international reference for both; the older military standard that popularized criticality analysis is no longer maintained.

Traditional practice multiplies severity, occurrence, and detection ratings into a risk priority number. That arithmetic is widely criticized because the ordinal ratings are not truly numeric and because very different risk profiles can produce identical products. Newer practice ranks actions by severity first, using a lookup table rather than a product, so that a catastrophic but rare mode cannot be scored away by a low occurrence rating.

For energy harvesters the analysis must distinguish hard failures from parametric drift. A cracked cantilever is an obvious failure mode, but a harvester whose resonance has shifted five percent away from the source frequency, or whose storage capacitor has lost a third of its capacitance, may still appear functional in a bench test while failing to power its load in the field. Defining the failure criterion in measurable terms, such as minimum average power delivered at end of life, keeps the analysis honest.

Fault Trees and Top-Down Analysis

Fault tree analysis complements FMEA by working downward from an undesired top event, such as loss of monitoring coverage on a critical machine, through the logical combinations of lower-level faults that could cause it. The top-down view exposes common-cause failures and dependencies that a component-by-component review can miss, including a shared power rail, a single sealing operation, or one supplier feeding several nominally redundant units.

The two methods are best used together. FMEA ensures that no component is overlooked; fault tree analysis ensures that no combination is. Where quantitative data exist, fault trees also yield probability estimates for the top event that can be compared against a reliability requirement.

Physical Failure Analysis

When units fail, physical analysis establishes the mechanism rather than the symptom. Non-destructive steps come first: X-ray imaging reveals voids, cracks, and misalignment inside joints; scanning acoustic microscopy detects delamination and voids at interfaces by imaging reflections from disbonded layers; infrared imaging locates hot spots under bias; and curve tracing characterizes shorts, opens, and leakage at individual terminals.

Destructive analysis follows. Decapsulation exposes the die, cross-sectioning reveals joint microstructure and crack paths, and scanning electron microscopy with energy-dispersive X-ray spectroscopy identifies corrosion products, intermetallic phases, and contaminants. Fracture surface examination distinguishes fatigue, which leaves beach marks and striations, from overload fracture, which does not.

Root-cause discipline requires linking the observed physical evidence to a specific cause and then to a corrective action. A report that ends at "solder joint crack" is incomplete; the useful conclusion identifies whether the crack came from thermal cycling strain, an assembly defect, or mechanical overload during handling, because each implies a different fix.

Cyclic Stress and Strain Accumulation

Fatigue failure occurs when materials subjected to repeated loading develop microscopic damage that accumulates over time, eventually leading to crack initiation and propagation. Unlike static failure that occurs when stress exceeds material strength, fatigue failure can occur at stress levels well below the yield strength when loading is cyclic.

Energy harvesters operating at resonance experience particularly high cycle counts. A harvester resonant at 100 Hz accumulates over 3 billion cycles annually during continuous operation. Even at lower frequencies and intermittent operation, cycle counts reach hundreds of millions over typical service lives, placing demanding requirements on material fatigue resistance.

Stress amplitude, mean stress, and stress ratio all influence fatigue life. High-amplitude oscillations accelerate damage accumulation, while tensile mean stresses reduce fatigue life compared to fully reversed loading. Understanding the stress history that harvesters experience in service enables appropriate fatigue analysis and design.

Strain-based fatigue analysis becomes necessary when plastic deformation occurs at stress concentrations. Local plasticity dissipates energy and generates heat, potentially degrading piezoelectric or other functional materials. Design approaches that minimize stress concentrations and maintain elastic behavior throughout the structure improve fatigue performance.

Crack Initiation and Propagation

Fatigue cracks typically initiate at surfaces, particularly at stress concentrations including corners, notches, and surface defects. Manufacturing processes that leave surface scratches, grinding marks, or residual stresses create preferential crack initiation sites. Surface treatments including polishing, shot peening, and coatings can improve fatigue resistance by reducing stress concentrations and introducing beneficial compressive residual stresses.

Once initiated, fatigue cracks propagate through the material with each load cycle. Crack growth rate depends on the stress intensity factor range, which increases as cracks grow longer. Initially slow growth accelerates as cracks extend, eventually reaching critical size where rapid fracture occurs.

Fracture mechanics analysis predicts crack propagation rates and remaining life once cracks are detected. Paris law and its modifications relate crack growth rate to the stress intensity factor range, enabling life prediction when initial crack sizes are known. Non-destructive inspection techniques including ultrasonic testing, dye penetrant inspection, and acoustic emission monitoring detect cracks before catastrophic failure.

Damage-tolerant design accepts that cracks may exist and ensures that structures remain safe during the time required for cracks to grow to critical size. Inspection intervals are set to detect cracks well before they reach dangerous lengths. This approach is particularly relevant for harvesters in critical applications where failure consequences are severe.

High-Cycle Fatigue in Resonant Harvesters

Resonant energy harvesters operate in the high-cycle fatigue regime, where cycle counts exceed millions before failure. High-cycle fatigue is dominated by elastic behavior, with damage accumulating through dislocation motion and microstructural changes rather than gross plastic deformation.

S-N curves characterize high-cycle fatigue behavior, plotting stress amplitude against cycles to failure. Many materials exhibit a fatigue limit or endurance limit below which fatigue life becomes essentially infinite. Designing harvesters to operate below this limit ensures long service life, though this may require accepting lower power output to reduce stress levels.

Piezoelectric ceramics present special fatigue considerations. Domain switching under cyclic electric fields causes mechanical damage analogous to mechanical fatigue. Coupled electromechanical loading accelerates degradation compared to pure mechanical loading. Fatigue-resistant piezoelectric compositions and operating well below coercive field strengths improve reliability.

Thin-film piezoelectric materials on cantilever harvesters experience both substrate fatigue and film degradation. Interface stresses between film and substrate can initiate delamination. Matching thermal expansion coefficients and ensuring good adhesion through surface preparation and deposition process control prevent interface failures.

Low-Cycle Fatigue Considerations

Some energy harvesting applications involve low-frequency, high-amplitude loading that causes plastic deformation with each cycle. Structural monitoring harvesters on bridges or buildings may experience large deflections during seismic events or extreme wind loading. Low-cycle fatigue life is typically measured in thousands rather than millions of cycles.

Coffin-Manson relationships relate plastic strain amplitude to cycles to failure in the low-cycle regime. Design against low-cycle fatigue requires limiting plastic strain accumulation, often through displacement stops or overload protection mechanisms that prevent excessive deflection.

Material selection for low-cycle fatigue resistance differs from high-cycle requirements. Ductile materials that accommodate plastic deformation without cracking perform better in low-cycle applications. Superelastic shape memory alloys offer exceptional low-cycle fatigue resistance through reversible martensitic transformation.

Temperature Effects on Materials

Temperature affects energy harvester performance and reliability through multiple mechanisms. Thermal expansion causes dimensional changes and internal stresses, particularly in assemblies combining materials with different expansion coefficients. Elevated temperatures accelerate chemical reactions including oxidation and diffusion-driven degradation.

Piezoelectric materials lose polarization at elevated temperatures, with complete depolarization occurring at the Curie temperature. Long-term exposure to temperatures well below the Curie point causes gradual depolarization through thermal activation of domain switching. Selecting materials with Curie temperatures far above operating temperatures and accounting for aging effects in design ensures reliable performance.

Thermoelectric materials may undergo phase changes, grain growth, or sublimation at elevated temperatures. These changes alter electrical and thermal properties, reducing conversion efficiency. Interface reactions between thermoelectric elements and electrodes can form resistive layers that increase electrical losses.

Polymer materials in flexible harvesters, encapsulants, and adhesives have limited temperature tolerance. Glass transition temperature marks the onset of dramatic property changes in amorphous polymers. Crystalline polymer melting points set upper temperature limits. Selecting polymers with adequate thermal stability for the application environment prevents premature failure.

Atmospheric Contaminants

Industrial and outdoor environments contain contaminants that can degrade energy harvester surfaces and materials. Sulfur compounds cause tarnishing of silver and copper. Chlorides accelerate corrosion of many metals. Particulates can abrade surfaces, block light, or interfere with mechanical motion.

Photovoltaic harvesters suffer soiling losses as dust, pollen, and pollution deposits on surfaces reduce light transmission. Self-cleaning coatings based on hydrophobic or photocatalytic materials reduce soiling accumulation. Regular cleaning may be necessary in heavily polluted environments.

The contaminant load itself varies sharply with site. Marine air carries chloride aerosol, industrial air carries hydrogen sulfide and sulfur dioxide that attack silver contacts and copper interconnects, and agricultural sites carry ammonia and organic dusts. Characterizing the deployment atmosphere is therefore a design input in its own right; the corrosion mechanisms these species drive, and the coatings and material choices that resist them, are covered under corrosion resistance below.

Piezoelectric Aging

Piezoelectric ceramics exhibit aging, a gradual reduction in piezoelectric properties over time. Aging results from domain wall motion that relaxes the polarization state toward equilibrium. The aging rate is highest immediately after poling and decreases logarithmically with time.

Manufacturers quantify aging as a percent change in a property per decade of time, so the change between one day and ten days after poling equals the change between ten days and one hundred days. Reported figures for the piezoelectric charge constant of commercial lead zirconate titanate commonly amount to a few percent per decade. A harvester characterized shortly after poling therefore delivers noticeably less output a year later, which is why data sheets state the elapsed time at which properties were measured.

Hard, acceptor-doped compositions generally age faster than soft, donor-doped ones, contrary to a common assumption. Acceptor doping leaves oxygen vacancies that pair with the dopant to form defect dipoles; as those dipoles gradually align with the local polarization they clamp domain walls and depress the dielectric and piezoelectric constants. Donor doping suppresses the oxygen vacancy population, so soft compositions age more slowly even though their mobile domain walls make them easier to depole under high stress or reverse field. Material selection therefore trades three separate properties, not one: initial coefficient, aging rate, and depoling resistance.

Mechanical and electrical stresses accelerate piezoelectric aging through stress-induced domain switching. Operating harvesters at high stress or field levels increases aging rates. Limiting operating levels and accounting for expected aging in initial design specifications ensures adequate end-of-life performance.

Thermal excursions can partially restore aged piezoelectric properties through thermal deaging, but may also cause permanent degradation if temperatures approach the Curie point. Understanding the thermal history of deployed harvesters helps predict actual aging behavior compared to laboratory predictions.

Polymer Degradation

Polymers degrade through chain scission, crosslinking, and oxidation reactions that alter mechanical, electrical, and optical properties. UV radiation, elevated temperature, oxygen, and mechanical stress all accelerate polymer degradation.

Piezoelectric polymers including PVDF and its copolymers can lose polarization through chain relaxation and crystalline phase changes. The beta phase that provides piezoelectric properties may convert to non-piezoelectric alpha phase over time or under thermal stress. Copolymers with higher beta-phase stability offer improved long-term reliability.

Encapsulant polymers must maintain adhesion, flexibility, and protective properties throughout service life. Hardening, cracking, and delamination can expose underlying components to environmental degradation. Selecting encapsulants with proven long-term stability and testing under accelerated aging conditions validates suitability.

UV stabilizers, antioxidants, and other additives improve polymer stability but may migrate out of thin films over time. Barrier layers that prevent additive loss help maintain long-term stability. Formulations optimized for long-life applications differ from those for general-purpose use.

Semiconductor Degradation

Semiconductor materials in photovoltaic and thermoelectric harvesters experience degradation mechanisms including defect generation, impurity migration, and interface degradation. Understanding these mechanisms guides material selection and process optimization.

Light-induced degradation affects silicon and some compound semiconductor solar cells. The Staebler-Wronski effect in amorphous silicon creates metastable defects under illumination that reduce efficiency. Thermal annealing can reverse the degradation, but performance oscillates seasonally in field operation. Advanced amorphous silicon formulations and tandem structures reduce light-induced degradation.

Perovskite solar cells face stability challenges from moisture, oxygen, heat, and light exposure. Ion migration under electric fields produces hysteresis and partial recovery in the dark, so a single measurement can misstate the true state of a cell. Because the qualification sequences written for crystalline silicon modules do not capture these behaviors, the research community reports perovskite stability against the ISOS protocols, a consensus framework of shelf-storage, thermal, light-soaking, and outdoor procedures first developed for organic photovoltaics and later extended to perovskites. Composition tuning, interface engineering, and improved encapsulation continue to extend perovskite lifetimes, but reported stability figures are comparable only when the test protocol is stated.

Thermoelectric material degradation includes sublimation of volatile components, grain boundary migration, and interdiffusion at interfaces. Diffusion barriers between thermoelectric elements and contacts prevent formation of resistive intermetallic phases. Operating temperature limits must account for diffusion kinetics over the intended service life.

Interconnection Failures

Electrical interconnections including wire bonds, solder joints, and conductive adhesives are common failure sites. Thermal cycling, mechanical stress, and corrosion can cause opens, intermittent connections, or increased resistance.

Failure symptoms differ in how they present. A clean open is easy to diagnose, but an intermittent connection that closes when warm and opens when cold, or a joint whose resistance has risen from milliohms to ohms, degrades output without tripping any fault detection. Because harvesters deliver microwatts to milliwatts into a high-impedance conversion stage, a series resistance increase that would be invisible in a mains-powered circuit can consume a large fraction of the available power. Bond-level mechanisms are treated in detail under interconnection reliability below.

Flip-chip connections using solder or conductive adhesive bumps experience fatigue from thermal cycling. The mismatch in thermal expansion between chip and substrate causes shear stress in the bumps. Underfill encapsulant distributes stress and dramatically improves fatigue life. Joint geometry and material selection also influence fatigue resistance.

Printed conductors in flexible and printed harvesters may crack under mechanical deformation. Conductor geometry optimization, strain-relieving patterns, and crack-stopping features improve flex durability. Metal-polymer composite conductors and liquid metal alternatives offer inherent flexibility.

Dielectric Breakdown

High voltages in piezoelectric harvesters and power conversion circuits can cause dielectric breakdown in insulating materials. Breakdown creates conductive paths through insulators, resulting in short circuits and device failure.

Piezoelectric materials operated near their coercive field strength risk dielectric breakdown, particularly at defects or inclusions where fields concentrate. Breakdown strength decreases with temperature and aging. Design margins must account for field enhancement at electrodes edges and manufacturing variability.

Polymer dielectrics in capacitors and insulating layers have voltage limits that depend on thickness, temperature, and time under stress. Partial discharge at defects causes progressive damage that eventually leads to breakdown. Quality control during manufacturing and appropriate design margins prevent dielectric failures.

Surface tracking and creepage failures occur when contamination or moisture creates conductive paths along insulator surfaces. Adequate creepage distances, conformal coatings, and surface treatments that resist contamination prevent surface breakdown.

Electromigration

High current densities in interconnects cause electromigration, the transport of metal atoms by momentum transfer from current-carrying electrons. Over time, electromigration creates voids at cathode ends and hillocks at anode ends, eventually causing open circuits.

Current density limits for reliable operation depend on conductor material, geometry, and temperature. Aluminum interconnects are particularly susceptible; copper offers significantly better electromigration resistance. Wide, thick conductors operating at low current density avoid electromigration concerns.

Energy harvesters typically operate at low currents where electromigration is not limiting. However, power management circuits with higher current densities and elevated operating temperatures must consider electromigration in design. Current-carrying capacity ratings account for electromigration reliability requirements.

Brittle Fracture in Ceramics

Piezoelectric ceramics and other brittle materials fail suddenly without warning when stress exceeds strength. Unlike ductile materials that deform plastically before failure, brittle materials provide no opportunity for load redistribution or visible damage indication before fracture.

Ceramic strength varies statistically, requiring design based on probability of failure rather than a single strength value. Weibull statistics characterize brittle material strength distributions. Design stress levels must account for the weakest specimens in the population, not average strength.

Surface flaws from machining, handling, or thermal shock concentrate stress and initiate fracture. Surface treatments that remove or heal flaws improve strength. Compressive surface layers from ion exchange or case hardening resist crack opening and improve apparent strength.

Proof testing applies loads exceeding service conditions to screen out weak specimens. Survivors have demonstrated adequate strength for the application. The proof test level balances rejection rate against confidence in survivor strength.

Impact and Shock Failure

Energy harvesters experience mechanical shock during handling, transportation, installation, and operation. Impact loads can fracture brittle components, break bonds, or cause permanent deformation that affects resonance frequency or alignment.

Shock testing characterizes device robustness against impact. Standard profiles including half-sine, trapezoidal, and sawtooth pulses simulate different impact scenarios. Peak acceleration and duration determine severity. Shock testing reveals design weaknesses before field failures occur.

Shock protection through mechanical stops, damping materials, and compliant mounting reduces transmitted shock levels. Packaging design must consider drop scenarios during shipping and handling. Fragile components benefit from protective features that absorb impact energy.

Repetitive shock differs from single impacts, with cumulative damage accumulating over many events. Harvesters on vibrating machinery or vehicles may experience thousands of shocks over their service life. Fatigue-like damage accumulation under repeated shock requires different analysis approaches than single-event shock.

Wear Mechanisms

Sliding, rotating, or impacting surfaces experience wear that removes material and changes geometry. While most energy harvesters avoid sliding contacts, some designs include mechanical rectifiers, variable capacitors with sliding electrodes, or bearings that experience wear.

Adhesive wear occurs when surfaces bond at contact points and material transfers between surfaces. Lubricants or low-friction coatings prevent adhesive wear. Material combinations that resist adhesion, such as dissimilar metals or polymer-metal pairs, reduce wear rates.

Abrasive wear removes material through cutting or plowing by hard particles or asperities. Surface hardening, filtration to exclude abrasive particles, and smooth surface finishes reduce abrasive wear. Hard coatings protect against abrasion in harsh environments.

Fretting wear results from small-amplitude oscillatory motion between contacting surfaces. Harvesters may experience fretting at mounting interfaces or electrical contacts. Fretting generates wear debris that accelerates wear and can cause electrical contact problems. Tight joints, adhesives, or compliant interfaces prevent fretting motion.

Creep and Stress Relaxation

Time-dependent deformation under sustained load causes dimensional changes and stress redistribution in energy harvesters. Creep and stress relaxation are significant for polymer and soft metal components, particularly at elevated temperatures.

Creep causes gradual dimensional change under constant load. In harvesters, creep can misalign components, change resonance frequencies, or create interference. Material selection considering creep resistance and operating temperature ensures dimensional stability.

Stress relaxation reduces internal stress under constant strain. Preloaded fasteners, spring contacts, and interference fits experience stress relaxation that may loosen joints or reduce contact force. Periodic retorquing, self-locking fasteners, or designs tolerant of reduced preload address stress relaxation concerns.

Viscoelastic polymers exhibit both creep and stress relaxation, with behavior depending on time, temperature, and load history. Time-temperature superposition enables prediction of long-term behavior from short-term testing at elevated temperature. Understanding viscoelastic behavior guides polymer selection and design.

Thermal Cycling Damage

Repeated temperature changes cause expansion and contraction that stress interfaces between materials with different thermal expansion coefficients. Solder joints, adhesive bonds, and thin-film interfaces are particularly vulnerable to thermal cycling damage.

Coffin-Manson type relationships predict thermal cycling fatigue life based on strain range, which depends on temperature excursion and expansion mismatch. Reducing temperature swings, minimizing expansion mismatches, or using compliant interfaces extends thermal cycling life.

Power cycling, where devices self-heat during operation and cool when idle, causes thermal cycling at the die level even when ambient temperature is constant. Junction temperature rise during operation depends on power dissipation and thermal resistance. Thermal design that limits temperature rise improves power cycling reliability.

Underfill and die attach materials must accommodate thermal cycling strain without cracking or delaminating. Material properties including modulus, expansion coefficient, and toughness determine thermal cycling capability. Ramp rate can also affect damage, with faster temperature changes causing higher instantaneous stresses.

Thermal Runaway

Positive feedback between temperature and power dissipation can cause thermal runaway, where self-heating leads to progressively higher temperatures until failure occurs. Thermal runaway is a concern for components with negative temperature coefficient of resistance or positive temperature coefficient of power dissipation.

Power electronics in energy harvester systems can experience thermal runaway if cooling is inadequate. MOSFET on-resistance increases with temperature, increasing conduction losses and causing further heating. Thermal design must ensure stable operating points under worst-case conditions.

Thermoelectric coolers driven beyond their maximum temperature differential can enter thermal runaway as Joule heating overwhelms Peltier cooling. Current limiting and thermal protection prevent destructive thermal runaway in thermoelectric devices.

Runaway is a design margin problem before it is a circuit problem: the stable operating point must survive worst-case ambient temperature, worst-case thermal resistance, and end-of-life device parameters, since each of these shifts the balance between generated and removed heat. Where margin alone cannot guarantee stability, the sensing, thresholding, and shutdown circuitry that arrests runaway is covered in Safety Systems and Protection.

Hot Spot Formation

Non-uniform current distribution creates localized hot spots that experience accelerated degradation compared to the average temperature. Solar cells with localized shading or defects, thermoelectric modules with current crowding at contacts, and circuits with poor thermal spreading can develop hot spots.

Bypass diodes in photovoltaic strings prevent hot spot formation by routing current around shaded or damaged cells. Without bypass diodes, reverse-biased cells dissipate power from the entire string, potentially causing thermal damage or fire.

Thermal imaging identifies hot spots that might not be apparent from average temperature measurements. Infrared cameras reveal temperature distributions across devices and modules. Hot spots indicate design issues, manufacturing defects, or degradation that require attention.

Design for uniform temperature distribution includes symmetric current paths, adequate metallization spreading, and thermal management that addresses the hottest locations. Simulation tools predict temperature distributions under various operating conditions, enabling thermal optimization before fabrication.

Moisture Permeation Mechanisms

Moisture enters packages through three primary mechanisms: permeation through encapsulants and seals, leakage through defects in seals, and outgassing from internal materials. Each mechanism requires different control approaches.

Polymer encapsulants are permeable to water vapor, with permeation rate depending on material composition, thickness, and temperature. Water vapor transmission rate (WVTR) characterizes permeation properties. Multi-layer barriers with alternating organic and inorganic layers achieve very low WVTR by creating tortuous diffusion paths. Organic photovoltaic cells and organic piezoelectric polymers set the hardest requirement, because both moisture and oxygen attack them; multi-layer barrier films approach the permeation performance of glass while remaining flexible.

Seal defects including cracks, voids, and incomplete bonding create fast paths for moisture ingress. Quality control during sealing operations and leak testing detect defective seals. Redundant seals provide backup protection if primary seals fail.

Internal moisture sources include absorbed water in polymers and adsorbed water on surfaces. Baking assemblies before final sealing removes absorbed moisture. Dry packaging and desiccants maintain low humidity during shipping and storage. Getters inside sealed packages absorb moisture that enters over time.

Packaging classes differ in how completely they address these paths. Hermetic packaging provides the most effective moisture protection but adds cost and complexity, and a truly hermetic seal prevents any moisture ingress over the product lifetime. Semi-hermetic packages slow ingress sufficiently for many applications. Non-hermetic packages rely on conformal coatings or potting compounds for moisture resistance.

Condensation cycling deserves particular attention in harvester deployments. Outdoor and unheated enclosures cross the dew point twice a day, so an enclosure that merely slows vapor ingress eventually accumulates liquid water on internal surfaces. Sealed enclosures also pump: heating raises internal pressure and cooling draws air, and any entrained moisture, back through the weakest seal. Vents fitted with hydrophobic membranes equalize pressure while blocking liquid water, which is often more effective than attempting a perfect seal on a low-cost housing.

Corrosion from Moisture

Moisture enables electrochemical corrosion that degrades metal conductors, contacts, and structural elements. Corrosion rate increases dramatically above critical relative humidity levels, typically around 60-70% depending on contamination.

Galvanic corrosion occurs between dissimilar metals in electrical contact when moisture is present. The less noble metal corrodes preferentially, potentially causing open circuits or mechanical failure. Avoiding galvanic couples, using corrosion-resistant materials, or isolating dissimilar metals prevents galvanic corrosion.

Ionic contamination accelerates corrosion by providing mobile charge carriers. Flux residues, fingerprints, and airborne contaminants introduce ionic species. Cleaning processes remove contamination before packaging. Conformal coatings provide additional protection in humid environments.

Corrosion-resistant materials and coatings provide inherent protection. Gold plating prevents corrosion of contact surfaces. Stainless steels and aluminum form protective oxide layers. Organic coatings and platings protect base metals from corrosive environments.

Moisture-Induced Electrical Failures

Beyond corrosion, moisture causes direct electrical failures through leakage currents, electrochemical migration, and dielectric degradation. These failures can be sudden or gradual depending on the mechanism.

Surface leakage currents increase dramatically when moisture forms conductive paths across insulators. Contamination that was benign when dry becomes conductive when wet. Conformal coatings and adequate creepage distances prevent surface leakage.

Electrochemical migration forms conductive dendrites between biased conductors in the presence of moisture and ionic contamination. Silver and copper are particularly susceptible. Dendrite growth can cause sudden short circuits after operating reliably for extended periods. Clean manufacturing, appropriate conductor spacing, and moisture barriers prevent electrochemical migration.

Moisture absorption in polymer dielectrics reduces insulation resistance and breakdown voltage. Polar water molecules increase dielectric losses and can cause localized heating. Selecting low-moisture-absorption dielectrics and providing moisture barriers maintains electrical properties.

Polymer Photodegradation

UV radiation breaks chemical bonds in polymers, initiating chain reactions that cause discoloration, embrittlement, and loss of mechanical properties. The energy of UV photons exceeds the bond energy of many polymer bonds, enabling direct photolysis.

Photo-oxidation combines UV damage with oxidation, accelerating degradation in oxygen-containing environments. Radical species generated by UV absorption react with oxygen to form peroxides that further degrade the polymer. The combination of UV and oxygen is more damaging than either alone.

UV stabilizers absorb UV radiation or scavenge free radicals to slow photodegradation. UV absorbers convert photon energy to heat without bond breaking. Hindered amine light stabilizers (HALS) trap radical species before they can cause damage. Stabilizer selection depends on polymer type and required service life.

Carbon black and other pigments provide UV protection by absorbing radiation before it reaches the polymer matrix. Black encapsulants and housings protect internal components from UV exposure. Where transparency is required, UV-absorbing coatings on outer surfaces block harmful wavelengths.

Organic Semiconductor Degradation

Organic photovoltaic materials and organic piezoelectric polymers are particularly sensitive to UV damage. The conjugated structures that provide electronic functionality also absorb UV strongly, making photodegradation inherent to their operation.

Organic solar cells degrade under the same light they harvest. UV filtering removes the most damaging wavelengths while passing visible light for power generation. The trade-off between UV filtering and light harvesting requires careful optimization for each application.

Oxygen and moisture accelerate UV degradation of organic semiconductors. Hermetic encapsulation excluding both oxygen and moisture dramatically improves stability. Operating in inert atmosphere or vacuum avoids photo-oxidation entirely.

Material design can improve intrinsic photostability. Side chain engineering, crosslinking, and morphology optimization all influence photodegradation rates. Continued research into stable organic semiconductors aims to achieve photovoltaic and piezoelectric devices with commercial durability.

UV-Resistant Design Strategies

Designing for UV resistance combines material selection, protective measures, and operational strategies to achieve required service life under UV exposure.

Material selection prioritizes inherently UV-stable polymers and compounds where possible. Fluoropolymers, silicones, and certain acrylics offer good UV resistance without stabilizers. Inorganic materials including glass, ceramics, and metals are largely immune to UV damage.

UV-blocking encapsulation protects sensitive internal components from UV exposure. Glass covers provide excellent UV protection for solar cells while maintaining optical transmission. UV-blocking polymer films offer lower cost and flexibility with adequate protection for many applications.

Accelerated UV testing using concentrated UV sources provides rapid assessment of material and design durability. Standards including ASTM G154 and IEC 61215 define UV testing protocols. Correlation between accelerated testing and real-world exposure enables lifetime prediction.

Types of Corrosion

Different corrosion mechanisms require different prevention strategies. Understanding which types of corrosion threaten a specific application guides material selection and protection approaches.

Uniform corrosion removes material evenly across surfaces. While predictable, uniform corrosion can consume protective coatings and eventually penetrate structural elements. Corrosion allowances in design account for expected material loss.

Pitting corrosion creates localized deep attacks that can penetrate thin sections rapidly. Pitting is particularly dangerous because small surface pits hide deep penetration. Stainless steels and aluminum are susceptible to pitting in chloride-containing environments.

Crevice corrosion occurs in gaps and confined spaces where oxygen depletion creates differential aeration cells. Tight joints, under gaskets, and beneath deposits are vulnerable to crevice corrosion. Design that eliminates crevices or ensures good access for protective environments prevents crevice corrosion.

Stress corrosion cracking combines tensile stress with specific corrosive environments to cause brittle fracture at stress levels far below normal strength. Certain alloy-environment combinations are particularly susceptible. Avoiding susceptible combinations and controlling residual stress prevents stress corrosion cracking.

Corrosion Protection Methods

Multiple approaches prevent or slow corrosion, often used in combination for critical applications. Protection method selection depends on the corrosion threats, design constraints, and cost considerations.

Barrier coatings isolate metal surfaces from corrosive environments. Organic coatings including paints and powder coatings provide economical protection for many applications. Metallic coatings including zinc, nickel, and chromium provide harder, more durable protection. Ceramic coatings offer exceptional resistance in extreme environments.

Cathodic protection supplies electrons to prevent oxidation of protected metals. Sacrificial anodes of zinc or magnesium corrode preferentially, protecting connected steel structures. Impressed current systems use power supplies to drive protective currents. Cathodic protection is common for large structures but rarely practical for small energy harvesters.

Corrosion inhibitors added to environments or coatings slow corrosion reactions. Inhibitors may work by forming protective films, neutralizing corrosive species, or interfering with electrochemical reactions. Volatile corrosion inhibitors protect enclosed spaces by releasing inhibiting vapors.

Material selection can eliminate corrosion concerns by choosing inherently resistant alloys. Stainless steels, titanium, and nickel alloys resist many corrosive environments. Higher material cost may be offset by elimination of coatings and extended service life.

Design for Corrosion Resistance

Design details significantly affect corrosion performance. Good corrosion design practices prevent many corrosion problems without adding cost.

Drainage prevents water accumulation that accelerates corrosion. Sloped surfaces, drain holes, and designs that avoid pockets and crevices reduce water retention. Orientation during operation and storage affects drainage effectiveness.

Dissimilar metal separation prevents galvanic corrosion. Insulating bushings, coatings, and gaskets break electrical contact between incompatible metals. When contact cannot be avoided, using metals close together in the galvanic series minimizes corrosion rate.

Access for coating application and inspection ensures that all surfaces receive adequate protection and that developing corrosion can be detected. Complex geometries with hidden surfaces are difficult to protect and inspect. Simple, open designs facilitate corrosion control.

Adhesion Fundamentals

Adhesion between layers depends on mechanical interlocking, chemical bonding, and physical forces including van der Waals attraction. Strong adhesion requires compatible materials, clean surfaces, and appropriate processing conditions.

Surface preparation critically affects adhesion. Cleaning removes contaminants that interfere with bonding. Surface treatments including plasma activation, chemical etching, and primer application promote adhesion. Surface roughening increases mechanical interlocking and bond area.

Thin-film adhesion depends on deposition conditions including substrate temperature, deposition rate, and process pressure. Interface engineering through seed layers, adhesion-promoting interlayers, and graded compositions improves thin-film adhesion.

Adhesion testing quantifies bond strength for process development and quality control. Peel tests, pull tests, and scratch tests measure adhesion through different loading modes. Standard test methods enable comparison between materials and processes.

Thermal Stress Effects on Delamination

Thermal expansion mismatch between layers creates interfacial shear stress that can initiate and propagate delamination. Thermal cycling amplifies the effect through repeated stress reversals.

Minimizing expansion mismatch reduces thermal stress. Material selection considering thermal expansion compatibility prevents large mismatches. Intermediate layers with graded expansion properties provide transition between dissimilar materials.

Compliant interlayers accommodate expansion differences through deformation rather than interfacial stress. Soft adhesives, elastomeric layers, and patterned interfaces with reduced stiffness allow relative motion without delamination.

Stress-relief features including slots, saw cuts, and compliant regions reduce stress concentration at layer boundaries. Finite element analysis identifies high-stress regions for targeted stress relief. Iterative design optimization balances stress reduction against functional requirements.

Moisture-Induced Delamination

Moisture absorption weakens interfaces and can cause delamination through several mechanisms. Hygroscopic expansion creates stress. Water molecules displace adhesive bonds at interfaces. Dissolved contaminants concentrate at interfaces and attack bonding.

Moisture sensitivity levels classify electronic packages by their tolerance to moisture absorption before assembly processes. Baking before solder reflow removes absorbed moisture and prevents moisture-induced failures. Dry packaging maintains low moisture content during storage and shipping.

Interface design for moisture resistance includes hydrophobic treatments, moisture-resistant adhesives, and edge sealing that prevents moisture ingress along interfaces. Interfacial moisture permeation is often faster than bulk permeation, making edges particularly vulnerable.

Testing for moisture-induced delamination uses pressure cooker testing, HAST (highly accelerated stress testing), and humidity storage followed by reflow simulation. Acoustic microscopy detects delamination non-destructively for process monitoring and failure analysis.

Wire Bond Reliability

Wire bonding remains the most common die-level interconnection method. Wire bond reliability depends on bond quality, wire material, loop geometry, and environmental exposure.

Gold ball bonds on aluminum pads form intermetallic compounds that grow with time and temperature. The purple AuAl₂ phase, long known as purple plague, is frequently blamed, though it is itself reasonably conductive and mechanically sound. Bond weakening is better attributed to the growth of gold-rich phases with differing diffusion rates and to the Kirkendall voiding that results, which opens a plane of voids at the bond interface. Halogen contamination accelerates the process markedly. Controlled bonding parameters, clean assembly, and limits on sustained junction temperature keep intermetallic growth within acceptable bounds.

Aluminum wedge bonds avoid gold-aluminum intermetallic issues but are more susceptible to corrosion. Aluminum wire is also more prone to fatigue than gold wire. Application requirements determine optimal wire bond type.

Wire bond loop geometry affects mechanical reliability. Low loops minimize package height but increase stress at bond heels. High loops accommodate more movement but risk wire sweep during molding. Loop design balances mechanical requirements against process constraints.

Solder Joint Reliability

Solder joints connect components to substrates and interconnect layers in multilayer structures. Solder joint reliability depends on joint geometry, solder composition, and thermal-mechanical loading.

Thermal cycling causes solder fatigue as joints strain to accommodate expansion mismatch between components and substrates. Joint life depends on strain range, which increases with component size and temperature excursion. Underfill, appropriate component sizing, and thermal management improve solder joint life.

Lead-free solders, now required by many regulations, have different reliability characteristics than traditional tin-lead solders. Most lead-free solders are harder and less ductile than tin-lead, affecting fatigue behavior. Some lead-free solders are susceptible to tin whisker growth that can cause short circuits.

Solder joint inspection uses X-ray imaging to detect voids, cracks, and insufficient solder. Automated optical inspection checks external fillet appearance. Cross-sectioning provides detailed joint analysis but destroys the sample. Inspection standards define acceptance criteria for different reliability requirements.

Conductive Adhesive Connections

Conductive adhesives offer lower processing temperatures than solder and eliminate concerns about intermetallic formation. However, conductive adhesives have their own reliability considerations.

Isotropic conductive adhesives (ICA) use high filler loadings to achieve conductivity in all directions. Conductivity depends on particle-to-particle contact that can be disrupted by thermal cycling or moisture absorption. Silver-filled epoxies are the most common ICAs.

Contact resistance of ICA joints may increase over time through oxide formation on filler particles and polymer relaxation. Stable contact requires maintaining mechanical pressure between particles. Aging behavior must be considered in design specifications.

Anisotropic conductive films (ACF) conduct only in the z-direction, enabling fine-pitch connections without shorting between adjacent pads. ACF reliability depends on maintaining adequate pressure on captured particles. Viscoelastic relaxation of the adhesive can reduce contact force over time.

Capacitor Wear-Out

Aluminum electrolytic capacitors wear out by losing electrolyte through the seal, a diffusion-driven process that raises equivalent series resistance and lowers capacitance until the part no longer performs its function. The widely used rule of thumb doubles useful life for each 10 degree Celsius reduction in core temperature. Because catalog ratings are typically a few thousand hours at the maximum rated temperature, an electrolytic capacitor is a poor choice wherever decade-scale unattended life is required.

Ceramic and film capacitors have no comparable wear-out mechanism and are preferred for long-life harvester buffers. Class II ceramics bring their own caveats: capacitance falls substantially with applied direct voltage and drifts downward with time after the ferroelectric dielectric is heated past its Curie point during soldering. Sizing a cold-start reservoir from the nominal catalog value, rather than the value at working voltage and end of life, is a recurring design error.

Rising equivalent series resistance has a particular consequence for harvesters. Cold-start circuits must accumulate enough charge to bring a converter out of its dormant state; when storage impedance climbs, a system that started reliably for years can eventually fail to start at all while every other component remains intact.

Supercapacitor Degradation

Electrochemical double-layer capacitors degrade gradually rather than failing abruptly. Capacitance fades and equivalent series resistance rises under the combined influence of cell voltage and temperature, driven by electrolyte decomposition and by changes at the electrode surface. End of life is therefore defined by thresholds, such as a specified percentage of capacitance lost or a specified multiple of the initial resistance, rather than by an open or short circuit.

Voltage derating is the most effective lever available to the designer. Operating a cell below its rated voltage extends life substantially, and the same is true of limiting peak temperature. Series stacks require balancing, whether by resistive bleed networks or active circuits, because an unbalanced cell sits above its share of the stack voltage and ages faster than its neighbors, which accelerates the imbalance.

Hybrid and lithium-ion capacitors offer higher energy density than symmetric double-layer devices but reintroduce voltage and temperature limits closer to those of batteries, including a minimum voltage below which the cell is damaged. Protection against deep discharge matters for harvesters that may sit unpowered through long periods without ambient energy.

Battery Aging under Harvested Charging

Rechargeable cells in harvesting systems experience a duty cycle unlike that of consumer electronics: shallow, intermittent charging at low current, long periods at high state of charge, and no opportunity for scheduled replacement. Calendar aging consequently dominates over cycle aging. Storage at high state of charge and elevated temperature accelerates capacity fade, so a charge controller that holds the cell at a reduced float voltage often yields longer service life than one that maximizes stored energy.

Low-temperature charging is a specific hazard for outdoor deployments. Charging a lithium-ion cell below roughly the freezing point of water promotes lithium plating on the anode, which permanently consumes lithium inventory and can create internal short circuits. A harvester exposed to winter conditions must therefore measure cell temperature and refuse to charge when it is too cold, accepting lost harvest in exchange for cell integrity.

Where cycle count matters more than energy density, thin-film solid-state cells and lithium titanate chemistries tolerate far more cycles and a wider temperature range at the cost of capacity. Some designs avoid the problem entirely by pairing a supercapacitor for daily cycling with a primary lithium cell held in reserve, so that the rechargeable element carrying the cycling load has no calendar-life constraint.

Designing for End-of-Life Storage Capacity

Storage should be sized against end-of-life capacitance or capacity, not the beginning-of-life value. A buffer that meets the load profile exactly when new will fail to meet it after predictable fade, producing field failures that the original analysis never anticipated. Adding margin at design time is far cheaper than replacing storage elements distributed across thousands of installed nodes.

Monitoring closes the loop. Periodically measuring the storage element's terminal behavior, such as the voltage decay after a known charge injection, yields a trend that indicates capacity fade long before the node loses function. That trend feeds directly into the prognostic methods discussed below.

Acceleration Factor Models

Acceleration factor models relate test conditions to field conditions, enabling lifetime extrapolation. Different failure mechanisms follow different acceleration relationships.

Arrhenius acceleration applies to thermally-activated failure mechanisms. The exponential temperature dependence allows significant acceleration through elevated temperature testing. Activation energy, determined experimentally, characterizes the temperature sensitivity of specific mechanisms.

Coffin-Manson acceleration applies to fatigue failures from cyclic loading. Cycles to failure scale inversely with the inelastic strain range raised to a power, so a test that doubles the strain range compresses life by roughly the square of that ratio. Reported fatigue exponents for solder joints cluster near two, with a value close to 1.9 widely used for eutectic tin-lead and comparable values reported for tin-silver-copper alloys. The Norris-Landzberg modification adds terms for cycling frequency and peak temperature, which matters because accelerated tests cycle far faster than field conditions and therefore allow less creep relaxation per cycle than the field does.

Power law relationships accelerate voltage, humidity, and other stress-dependent mechanisms. The exponent in power law acceleration varies by mechanism and must be determined experimentally for accurate lifetime prediction.

Combined acceleration using multiple simultaneous stresses produces faster testing but complicates analysis. Interactions between stresses can create failure modes not seen under single stresses. Eyring models extend Arrhenius acceleration to include non-thermal stresses.

Standard Reliability Tests

Industry standards define reliability tests for consistent evaluation across products and manufacturers. Standards specify test conditions, durations, and acceptance criteria.

Temperature cycling per JEDEC JESD22-A104 alternates between temperature extremes with controlled ramp rates and dwell times. Standard profiles range from 0 to 100 degrees Celsius for consumer products to -65 to 150 degrees Celsius for military applications. Cycle counts from 500 to several thousand cycles depending on reliability requirements.

High-temperature operating life (HTOL) tests operate devices at elevated temperature under bias to accelerate thermally-activated mechanisms. JEDEC JESD22-A108 defines standard conditions. Test duration of 1000 hours is common, with longer durations for high-reliability applications.

Highly accelerated stress testing (HAST) combines elevated temperature and humidity with optional electrical bias. Because JEDEC JESD22-A110 places the chamber above the boiling point of water, the humidity is maintained with pressurized saturated steam. The most severe common condition is 130 degrees Celsius at 85 percent relative humidity for roughly 96 hours; a milder 110 degree Celsius condition at the same humidity runs for several hundred hours. HAST compresses into days the moisture stress that steady-state 85 degree Celsius, 85 percent relative humidity testing per JESD22-A101 applies over a thousand hours.

Mechanical shock and vibration testing per MIL-STD-810 or IEC 60068 characterizes robustness against mechanical stresses. Drop testing simulates handling and shipping events. Random vibration testing evaluates performance under continuous vibration exposure.

Highly Accelerated Life Testing

Highly accelerated life testing (HALT) uses extreme conditions to find design and manufacturing weaknesses quickly. Unlike qualification testing that verifies meeting requirements, HALT aims to identify failure modes for improvement.

HALT protocols step through increasingly severe conditions until failures occur. Temperature stepping from cold to hot extremes identifies thermal weaknesses. Vibration stepping with six-degree-of-freedom random vibration reveals mechanical vulnerabilities. Combined thermal-vibration stresses find interaction failures.

HALT findings drive design improvements that extend margin beyond minimum requirements. Each failure mode identified in HALT is an opportunity to improve robustness. Iterative HALT testing after improvements verifies effectiveness.

Highly accelerated stress screening (HASS) applies HALT-like conditions to production units to screen out infant mortality failures. HASS stresses are set below HALT operating limits to avoid damaging good units. Production screening catches manufacturing defects before field failure.

Health Monitoring Techniques

Health monitoring measures parameters that indicate system condition. Trending these parameters reveals degradation before functional failure occurs.

Electrical monitoring tracks power output, conversion efficiency, and impedance changes. Degradation often manifests as gradual performance loss before catastrophic failure. Power output trending with environmental compensation enables comparison over time.

Mechanical monitoring measures vibration signatures, acoustic emissions, and resonance frequencies. Changes in harvester resonance frequency indicate structural damage or material property changes. Acoustic emission detects crack initiation and growth.

Environmental monitoring tracks temperature, humidity, and other conditions that affect degradation rate. Correlating environmental exposure with health indicators improves remaining life prediction. Cumulative damage models integrate stress history over time.

Failure Prediction Algorithms

Algorithms process health monitoring data to predict remaining useful life. Physics-based and data-driven approaches offer different capabilities for different applications.

Physics-based prognostics use failure mechanism models to predict life from measured degradation indicators. Crack growth models predict remaining cycles from measured crack length. Capacity fade models predict battery end of life from discharge capacity trending. Physics models require understanding of failure mechanisms but generalize well from limited data.

Data-driven prognostics use machine learning to identify patterns in historical data that precede failure. Training on run-to-failure data from similar systems enables prediction without explicit mechanism understanding. Data-driven approaches require substantial training data but can capture complex multi-factor degradation.

Hybrid approaches combine physics and data-driven methods. Physics models provide structure while data fitting adapts to specific system characteristics. Hybrid prognostics offer robustness across varying conditions with reduced data requirements.

Remaining Useful Life Estimation

Remaining useful life (RUL) estimation provides the actionable output of prognostics systems. RUL enables planning maintenance, ordering replacement parts, and managing inventory.

Point estimates of RUL provide a single value for remaining life. While simple to use, point estimates do not convey uncertainty. Decisions based on point estimates may be overly aggressive or conservative depending on actual uncertainty.

Probabilistic RUL estimates provide probability distributions for remaining life. The distribution conveys both expected life and uncertainty. Decision-makers can set thresholds based on acceptable risk of failure before maintenance.

Confidence bounds on RUL quantify prediction uncertainty. Wide bounds indicate unreliable predictions that should not drive critical decisions. Narrowing bounds as failure approaches enables confident action timing.

Handbook Prediction Methods

Handbook methods provide component failure rate estimates based on technology type, quality level, and application environment. MIL-HDBK-217, Telcordia SR-332, and FIDES are the commonly used prediction standards, but they differ sharply in currency. The last revision of MIL-HDBK-217F dates from the mid-1990s and the handbook has not tracked component technology since, a limitation that has drawn sustained criticism; Telcordia SR-332 and FIDES have continued to receive new issues. The 217Plus methodology, developed from the same military research programs that produced the handbook, addresses part of the gap by adding process and environmental factors to the component models.

Handbook predictions multiply base failure rates by factors for operating conditions. Temperature, voltage stress, and environmental severity modify base rates. Quality factors account for manufacturing and screening differences.

Handbook methods provide quick estimates useful for early design phases and comparative analysis. Limitations include reliance on historical data that may not represent current technologies and lack of insight into failure mechanisms.

Parts count and parts stress prediction approaches trade accuracy for effort. Parts count uses generic conditions for rapid estimation. Parts stress analysis using actual operating conditions provides more accurate results when detailed design information is available.

Physics-of-Failure Prediction

Physics-of-failure (PoF) prediction models specific failure mechanisms using fundamental understanding of degradation processes. PoF provides mechanism-specific insights that enable design improvement.

PoF models require identifying relevant failure mechanisms for each component and load condition. Mechanism models then predict time to failure based on operating conditions and material properties. Multiple mechanisms compete, with the shortest predicted life determining overall reliability.

Material and geometry inputs to PoF models require detailed design information. Interface materials, joint geometries, and operating temperatures must be known or bounded. Sensitivity analysis identifies critical parameters for design attention.

PoF predictions enable design optimization for reliability. Varying design parameters in models identifies opportunities to improve life without trial-and-error testing. Virtual qualification using simulation reduces development time and cost.

System Reliability Analysis

System reliability analysis combines component reliabilities to predict overall system performance. Series, parallel, and complex configurations require different analysis approaches.

Series system reliability is the product of component reliabilities when any component failure causes system failure. The weakest component dominates system reliability. Design efforts should focus on improving the least reliable components.

Parallel redundancy improves reliability by providing backup components that take over if primary components fail. Active redundancy has all components operating simultaneously. Standby redundancy activates backups only when needed, potentially achieving higher reliability if standby components degrade more slowly.

Reliability block diagrams graphically represent system configurations for analysis. Fault trees identify combinations of failures that cause system failure. Both tools support systematic reliability analysis of complex systems.

Lifetime Prediction and Warranty Analysis

Lifetime prediction supports warranty period setting, maintenance planning, and spare parts stocking. Accurate prediction requires appropriate models and quality input data.

Weibull analysis fits failure data to distributions that characterize infant mortality, random failures, and wear-out. The shape parameter indicates the dominant failure mode. Life estimates at various percentiles support different planning needs.

Warranty analysis estimates returns and costs from lifetime predictions and usage patterns. Warranty period setting balances customer satisfaction against warranty cost. Extended warranty pricing requires accurate lifetime prediction.

Field data analysis validates and improves predictions. Comparing predicted versus actual failure rates identifies model deficiencies. Updating models with field experience improves future predictions. Continuous improvement of prediction capability supports better design decisions.