Every test report, every chamber spec sheet, every standard document assumes you already know the vocabulary. You're expected to know that a "dwell" is not a pause but a defined soak. That "gRMS" is not a weight. That "HALT" is not a standard. That "uniformity" and "stability" measure entirely different things. That the "DUT" and the "product" are the same physical object, but never called the same thing in the same sentence.
This glossary defines the terms the way a senior test engineer would explain them to a new team member on their first week: precisely, with the nuance that matters in practice, and with a direct note wherever the term is routinely confused with something it isn't. We've expanded it from 50 to 75 terms, adding the HALT and HASS vocabulary that most general-purpose glossaries skip, plus the terms that appear without explanation in test reports and leave junior engineers guessing.
Bookmark it. You'll use it.
A
Acceleration factor (AF)
The ratio of test duration to equivalent field exposure time. An acceleration factor of 10 means one hour in the chamber represents ten hours in the field. Calculating a valid acceleration factor requires a physics-based model — Arrhenius for thermally-driven degradation, Coffin-Manson for fatigue — and knowledge of both the test conditions and the actual field conditions. A number cited without the underlying model is not an acceleration factor. It's a guess with a label.
Acceptance criteria
The defined pass/fail threshold for a test. Not specified by the chamber or the test method — specified by the engineer, the customer, or the standard governing the product. A chamber can run a technically perfect test against acceptance criteria that are meaningless. The acceptance criteria are the engineering judgement in the test programme. Everything else is execution.
Accelerometer
A transducer that measures acceleration, typically based on piezoelectric technology, providing a voltage output proportional to the acceleration at its location. In HALT and HASS testing, multiple accelerometers are attached to the UUT to monitor its vibration response. The distinction matters: the table input is what the system is commanded to deliver; the UUT response is what the product actually experiences. They are rarely the same number, especially above 2 kHz.
Air control thermocouple
A thermocouple providing air temperature feedback to the chamber thermal control system. May function as a safety limit controller, a primary control sensor, or both, depending on configuration. What the air control thermocouple reads is not what the product experiences. A product with thermal mass lags behind the air. A product generating heat may exceed it. The distinction between air temperature and product temperature is the most common source of misinterpreted test results.
Alternating low-level vibration
Also called modulated vibration or tickle vibration. A technique used in HALT vibration step stressing where, at the end of each dwell, the vibration level drops briefly to 5 gRMS and a complete functional test is run — before returning to the next step level. The purpose: some failures are precipitated at high gRMS but are only electrically detectable at low vibration levels. Cracked solder joints, for instance, may close at 30 gRMS but remain open at 5 gRMS. Alternating low-level vibration catches them.
Ambient conditions
The temperature and humidity of the room in which the chamber operates. Ambient conditions affect chamber performance, particularly at temperature extremes. A chamber rated to −70°C at 23°C ambient may only reach −60°C in a summer lab at 35°C. Spec sheet figures are always referenced to a standard ambient — typically 23°C — and always measured empty. Your lab is neither 23°C in August nor empty.
Arrhenius equation
A mathematical model relating chemical reaction rate to temperature. The foundational model for calculating acceleration factors for thermally-activated failure mechanisms — oxidation, dielectric degradation, polymer ageing. The equation contains an activation energy term (Ea), specific to the failure mechanism, that must be determined empirically or from published literature. Using the wrong activation energy produces an acceleration factor that has no connection to physical reality. In semiconductor reliability, activation energies between 0.3 and 1.0 eV are commonly cited depending on the mechanism; applying one mechanism's Ea to another mechanism's failure produces confident-looking numbers that mean nothing.
B
Broadband vibration
Vibration that contains energy over a wide frequency range, typically 10 Hz to 5 kHz or higher. HALT chambers generate broadband random vibration using pneumatic repetitive shock actuators. The broadband nature means energy is distributed across many frequencies simultaneously, exciting resonances across the product's entire frequency response — not just at a single tuned frequency. A sine sweep finds resonances sequentially. Broadband vibration finds them all at once.
Burn-in
A production screen in which powered product is held at elevated temperature — typically 70–125°C — for a defined period, with the goal of precipitating early-life failures before shipment. Effective for thermally-activated failure mechanisms. Poorly matched to mechanical failure modes (solder joint fatigue, connector seating, cold welds) that require cycling or vibration to precipitate. HASS has largely replaced burn-in in programmes where the dominant latent defect is mechanical rather than thermally-activated.
C
Calibration
The process of comparing a chamber's sensor readings against a traceable reference standard and documenting the difference. Calibration establishes whether the chamber is reading accurately — it does not adjust the chamber. Adjustment is a separate step, and not always performed. A calibrated chamber whose sensors read 2°C high will run every test 2°C hotter than programmed unless that offset is corrected. The calibration certificate tells you what the error is. It doesn't fix it.
Chamber conditioning
Running a chamber through a defined temperature or humidity cycle before a test begins, to stabilise the internal environment and remove residual moisture or contamination from previous tests. Skipping conditioning is a common source of test-to-test variability, particularly in humidity tests where moisture absorbed into chamber walls during a previous high-RH test affects initial conditions of the next test. It shows up as repeatability issues that take weeks to diagnose.
Coffin-Manson relationship
A fatigue model relating plastic strain amplitude per cycle to cycles to failure. The foundational model behind temperature cycling test acceleration factor calculations. In solder joint reliability, the strain per cycle is driven by temperature range, CTE mismatch, and joint geometry. A wider temperature range in the test produces higher strain per cycle, faster failure, and a larger acceleration factor relative to a narrower field temperature range — which is why test conditions are often more severe than field conditions by design.
Corrective action (CA)
A change implemented in design or process to eliminate a product weakness or flaw identified during testing. The final step of the HALT process — and the step that produces the actual reliability improvement. A HALT test that generates failures but no corrective actions has produced a list of problems, not a more reliable product. Corrective actions may include design changes, component substitutions, process modifications, or tightened incoming inspection. Each corrective action should be verified by retesting.
CTE (Coefficient of Thermal Expansion)
The fractional change in a material's dimension per degree of temperature change, in ppm/°C. CTE mismatch between joined materials is the mechanical driver of fatigue failure in solder joints, wire bonds, and adhesive interfaces under thermal cycling. Common values: copper ~17 ppm/°C, FR-4 PCB laminate ~14–18 ppm/°C in-plane and ~50–70 ppm/°C through-thickness (z-axis), alumina ceramic ~6–7 ppm/°C, silicon ~2.6 ppm/°C. The larger the mismatch, the higher the strain per cycle.
D
Destruct limit (DL)
In HALT testing, the stress level at which damage to the product is permanent and non-recoverable. A hard failure defines the destruct limit. The destruct limit is the upper boundary of the HASS screen corridor — no production screen should approach it. Confusing the destruct limit with the operating limit produces HASS screens that destroy good product. The distinction is reversibility: operating limit failures recover, destruct limit failures don't.
Dew point
The temperature at which air, at a given moisture content, becomes saturated and condensation begins. Relevant whenever a product surface is colder than the surrounding air's dew point. A product that has been cold-soaked at −40°C, then removed from the chamber into 23°C ambient air, will condense moisture on every exposed surface. Whether that condensation is part of the test intent or an accidental post-test condition changes the interpretation of any failure that follows.
DUT (Device Under Test)
The product, component, or assembly being tested. DUT and product are the same object — the term DUT is used in test engineering to distinguish the item being stressed from the surrounding hardware (chamber, fixture, cables, instrumentation). Every test report should specify the DUT precisely: part number, revision, serial number, assembly state. Two DUTs that differ in conformal coating status, potting, or revision are not the same test article, regardless of shared product name.
Dwell time
The time a product spends at a temperature extreme before the next ramp begins. Dwell must be long enough for the product — not just the chamber air — to reach thermal equilibrium at the setpoint. Ten minutes is adequate for a small IC. It is inadequate for a 3 kg aluminium assembly. Insufficient dwell means the product never reaches the specified temperature, which makes the test less severe than intended and invalidates the acceleration factor. Standards that specify dwell-to-stabilisation rather than dwell-by-time are more physically meaningful.
E
Elevation (altitude) testing
Exposure of a product to reduced air pressure, simulating conditions at altitude. At 35,000 feet (~10,700 m), atmospheric pressure is approximately 26 kPa — about one quarter of sea level. Relevant for: cooling systems (convection degrades at low pressure), sealed enclosures (pressure differential can cause deformation or outgassing), and electrical systems (arcing voltage thresholds drop in thin air — the Paschen curve tells you how much). Governed by IEC 60068-2-13 and MIL-STD-810 Method 500.
F
Failure mode
The physical mechanism by which a product fails — not the symptom. "Open circuit" is a symptom. "Solder joint fatigue crack at the heel fillet of an 0402 MLCC at position C47, propagating through 60% of the joint cross-section" is a failure mode. Test programmes are designed around failure modes. A test that finds a symptom without identifying the mechanism has produced an observation, not an engineering result.
Fast Fourier Transform (FFT)
An algorithm that converts a time-domain vibration signal into its component frequencies, providing magnitude and frequency information about each frequency component. In HALT and HASS testing, FFT analysis of accelerometer data reveals the frequency content of the vibration input and the UUT's response. A product with a mechanical resonance at 450 Hz will show amplification of the 450 Hz component in the UUT response relative to the table input. FFT is how you find that resonance.
Fixture
The hardware used to hold, connect, and support a DUT inside a test chamber. Fixture design is consistently underestimated in test programme planning. A fixture with high thermal mass slows product ramp rates below the chamber's rated rate. A fixture that blocks airflow creates temperature gradients around the DUT. A fixture with metal-to-metal contact between the DUT and chamber structure creates a thermal path that bypasses the air. Poor fixture design is responsible for more unexplained lab-to-lab result differences than any other single factor.
Functional test
A test of the UUT that measures functionality and critical parameters to determine whether the product meets specifications or has degraded. In HALT and HASS, functional tests are performed throughout the stress sequence — not just at the start and end. A product that fails functionally at −55°C but passes at ambient cannot be confirmed as a failure unless a functional test is run at −55°C. Missing intermediate functional tests is the most common way HALT testing fails to find its own results.
G
gRMS (Root Mean Square acceleration)
The standard measure of vibration intensity, equal to the root mean square of the acceleration signal across a specified frequency band, normalised to the acceleration due to gravity. In HALT chambers, gRMS integrates the power spectral density across the full broadband frequency range. Real-world field vibration environments: 0.5–5 gRMS for most products. HALT chambers: 20–60 gRMS. The difference is intentional. HALT is not simulating the field. It is finding the product's limits beyond it.
H
HALT (Highly Accelerated Life Testing)
A reliability discovery methodology that applies escalating combined stresses — rapid thermal cycling and six-degree-of-freedom vibration — to find design weaknesses before production. Not a standard. Not a compliance test. A methodology. The entire value of HALT is in the failure analysis and corrective action following each failure event. A HALT test that generates failures but no corrective actions has produced a list of problems, not a more reliable product. A HALT test with no failures has produced no useful information at all.
Hard failure
A non-recoverable failure. When the stress that caused the failure is reduced or removed, the product does not recover function. Hard failures define the destruct limit. Contrast with soft failure, which recovers when stress is removed. The distinction determines whether the failure defines the operating limit or the destruct limit — and therefore whether the HASS corridor has been defined correctly.
HASS (Highly Accelerated Stress Screening)
A production screen that uses stresses derived from HALT operating limits to precipitate latent manufacturing defects before shipment. Requires prior HALT data to establish valid screen levels. Requires proof-of-screen testing — subjecting known-good units to the full HASS profile and verifying no degradation — before the screen can be used in production. The HASS corridor, between operating limit and destruct limit, is where every HASS screen must operate. A HASS screen running above the destruct limit is consuming product life.
HASS corridor
The stress range between the product's operating limit and its destruct limit, established through HALT. The HASS production screen must operate within this corridor — severe enough to precipitate latent defects, but not so severe as to damage good product. The corridor width varies by product design. A product with a wide margin between operating limit and destruct limit has a wide, forgiving HASS corridor. A product with a narrow margin requires a precisely calibrated screen.
HASS development
The process of determining and defining the appropriate HASS profile for a production screen, based on HALT operating limit data. Includes stress level selection, cycle count, and the verification process (proof of screen) that confirms the screen does not damage conforming product. HASS development cannot be performed without prior HALT data. Attempting to define a HASS screen without HALT data is guesswork with a production line attached to it.
Humidity (Relative Humidity, RH)
The ratio of actual water vapour content of air to the maximum it could hold at that temperature, expressed as a percentage. At 100% RH, further cooling causes condensation. RH is temperature-dependent — the same mass of water vapour produces different RH values at different temperatures. Chamber humidity controllers compensate for this relationship continuously. Humidity specs always reference a temperature range because the physics of water vapour set hard limits on achievable RH outside that range. A chamber rated "10–98% RH" without a temperature range qualification is giving you incomplete information.
I
IEC 60068
The International Electrotechnical Commission's family of environmental testing standards for electrical and electronic equipment. The parent document (IEC 60068-1) defines general principles. Individual test methods are numbered parts: IEC 60068-2-1 (cold), IEC 60068-2-2 (dry heat), IEC 60068-2-14 (thermal shock and cycling), IEC 60068-2-78 (damp heat steady state). The most widely referenced environmental testing framework globally. Reading a datasheet or test report that cites "IEC 60068" without a part number tells you nothing — which is often the intention.
IP rating (Ingress Protection)
A classification system defined in IEC 60529 specifying protection against solid particles and liquids. Two digits: the first indicates protection against solids (0–6), the second against liquids (0–9K). IP67 means complete dust protection and temporary submersion in up to 1 metre of water for 30 minutes. The test conditions — nozzle size, flow rate, distance, duration — are precisely defined. A product that "meets IP67 in the lab" using a garden hose has not been tested to IP67.
L
Latent defect
A manufacturing defect that does not cause immediate failure under inspection or functional test, but which reduces reliability and will eventually cause field failure under operational stress. The target of HASS screening. A cold solder joint that passes a room-temperature functional test but fails after 300 hours at 85°C is a latent defect. The difference between a product that fails at 3 months and one that fails at 5 years is often latent defects in the early population.
LDL (Lower Destruct Limit)
In HALT, the lowest temperature at which the product suffers permanent, non-recoverable damage. The lower boundary below which no HASS screen must go. Determined during cold thermal step stressing by identifying the temperature at which damage is observed that does not recover when temperature returns to ambient.
LOL (Lower Operating Limit)
In HALT, the lowest temperature at which the product malfunctions — but recovers functionality when returned to ambient. The gap between the LOL and the expected field minimum temperature is the cold-side operating margin. A product with an LOL of −70°C used in an environment that reaches −40°C has a 30°C margin on the cold side. Finding that margin is the first goal of cold step stressing in HALT.
M
MIL-STD-810
The US Department of Defense's environmental engineering standard. Structured around environments — altitude, temperature, humidity, vibration, shock — rather than prescriptive test conditions. Unique among major standards in requiring test conditions to be tailored to measured field environments, not defaulted to standard table values. Currently at revision H (2019). Widely adopted outside defence as a ruggedisation benchmark, often without implementation of the tailoring requirement — which is specifically the part that makes the standard physically meaningful.
MLCC (Multilayer Ceramic Capacitor)
The most common passive component in electronics and the component most sensitive to thermal shock. The ceramic dielectric (CTE ~7–10 ppm/°C) is significantly mismatched to the PCB laminate and copper pad it's soldered to. Under rapid thermal transitions, the CTE mismatch generates tensile stress in the ceramic body that can cause fracture — a failure mode distinct from the solder joint fatigue that thermal cycling targets. MLCC fracture is why thermal shock and temperature cycling are separate tests: they find different failures.
N
NTF / NFF (No Trouble Found / No Fault Found)
A product returned from the field as failed, which cannot be reproduced in a bench test. The most expensive category of warranty return, because it costs the full cost of disassembly, analysis, and reassembly without generating any useful failure data. NTF failures are often latent defects that are stress-dependent — they appear under field vibration or temperature excursion, but disappear at ambient on the bench. HALT-style alternating low-level vibration is specifically designed to make NTF failures detectable. "No fault found" usually means "we didn't stress it the right way."
O
Operating limit (OL)
In HALT, the stress level at which the product malfunctions but recovers when stress is reduced. A soft failure defines the operating limit. The upper and lower operating limits (UOL and LOL) define the product's functional boundary under the combined thermal and vibration stresses of HALT. The gap between the operating limit and the destruct limit is the design margin the HASS screen exploits.
Operating window
The range between the lower and upper operating limits established through HALT thermal and vibration step stressing. The wider the operating window relative to the expected field environment, the higher the product's reliability margin. HALT's purpose is to find the edges of the operating window and push them outward through design improvement — not just to observe where they currently are.
P
Parametrically marginal unit
A unit that has been determined to be marginal based on functional testing results — not failed outright, but performing at the edge of its specification. Parametrically marginal units are the early warning of a manufacturing process shifting. In HASS testing, they often represent latent defects that would cause field failure within the first few thousand operating hours. Tracking parametrically marginal units across production lots reveals process drift before it produces outright failures.
Passthrough (cable passthrough)
A sealed port in the chamber wall through which electrical cables, thermocouples, or fluid lines enter without breaking the thermal or humidity seal. Passthrough design affects chamber performance. An oversized or unsealed passthrough allows chamber air to leak and ambient air to enter, destabilising temperature and humidity near the port. For humidity tests, passthrough sealing is especially critical — moisture escaping through a poorly sealed passthrough shows up as humidity setpoint drift that the control system cannot correct.
PID controller
A control algorithm — Proportional, Integral, Derivative — used to regulate chamber temperature and humidity. Reads sensor values, computes error relative to setpoint, adjusts heating, cooling, and humidity outputs to close the error. PID tuning determines how accurately and stably the chamber holds setpoint. A poorly tuned PID overshoots and oscillates, producing test conditions that look stable on a dashboard but aren't. A well-tuned PID can hold within ±0.3°C of setpoint continuously. The tuning is specific to the thermal load inside the chamber — adding or removing a large DUT can require re-tuning.
Picket fencing
A characteristic PSD shape produced by a vibration system whose actuators all strike at the same frequency but at slightly different times. The resulting spectrum shows high vibration levels at the strike frequency and its harmonics, with little energy between harmonics — resembling a picket fence. Picket fencing is undesirable in HALT vibration because it concentrates energy at specific frequencies rather than distributing it broadly. Well-designed HALT systems use randomised strike timing to produce flat broadband PSD profiles.
Power Spectral Density (PSD)
A measurement of vibration amplitude and frequency content, expressed in gRMS²/Hz. The shape of the PSD graph shows relative energy levels across the frequency range, allowing analysis of vibration fixture effectiveness and the UUT's response characteristics. A flat PSD means energy is evenly distributed across the frequency range. A peaked PSD reveals resonances — in the table, the fixture, or the product. PSD is to broadband vibration what a frequency spectrum is to acoustics: the full picture, not a single number.
Product control thermocouple
A thermocouple attached to the product that provides product temperature — not air temperature — to the chamber control system. When the control system uses product temperature rather than air temperature for closed-loop control, it ensures the product reaches the specified condition regardless of air temperature lag or lead. Critical for accurate dwell validation and for large-thermal-mass products where air temperature and product temperature diverge significantly during transitions.
Proof of screen
A verification test performed before deploying a HASS screen in production, in which known-good units are subjected to the complete proposed HASS profile — repeatedly — to confirm the screen does not consume meaningful product life. A screen that damages good product is worse than no screen at all. Proof of screen is the step that confirms the HASS corridor has been correctly identified and the screen operates within it.
PTH (Plated Through-Hole)
A via or component hole in a PCB whose barrel has been electroplated with copper to create electrical continuity between layers. PTH barrels are vulnerable to fatigue cracking under thermal cycling — z-axis CTE mismatch between copper and laminate puts the barrel in tension on every cycle. Cross-sectioning PTH barrels after thermal cycling tests is the standard method for detecting barrel cracking before it propagates to an open circuit. A barrel that has cracked 30% through its cross-section will not show up on a continuity check — but it will fail in the field.
R
Ramp rate
The rate at which temperature changes, in °C per minute. Specified ramp rates on chamber data sheets are measured in an empty chamber under standard ambient conditions. Loading the chamber with product reduces the achievable ramp rate — how much depends on the DUT's thermal mass. Test profiles that specify ramp rate must account for product thermal mass to confirm the specified rate is achievable. For thermal shock applications, ramp rate in a single-zone chamber is the limiting parameter — two-zone systems achieve transitions that no ramp rate can match.
Repetitive Shock (RS) vibration
The vibration mechanism used in HALT and HASS chambers, originating from repeated pneumatic actuator strikes on a table supported by springs. RS vibration produces six-degree-of-freedom (6DoF) motion — simultaneous excitation in X, Y, Z, roll, pitch, and yaw — which replicates a broader range of real-world vibration environments than single-axis electrodynamic shakers. The broadband, multi-axis nature of RS vibration is why HALT chambers find failure modes that sine sweep and single-axis random vibration tests miss.
Return on Investment (ROI) — HALT
In the context of HALT, the ratio of cost savings from field failure prevention to the total cost of HALT testing (equipment, infrastructure, personnel, time). The challenge is that field failures avoided don't appear in any cost ledger — they are counterfactual. ROI calculations for HALT require warranty cost data from similar previous products and engineering estimates of the failure rate reduction attributable to design changes made from HALT results. The calculation is imprecise. The value is not.
Root Cause Analysis (RCA)
Identifying the true cause of a weakness — fully understanding what failed and why. Not the symptom. Not the first proximate cause. The underlying mechanism and the conditions that allowed it to exist. RCA may require failure analysis tools including scanning electron microscopy, cross-sectioning, energy-dispersive X-ray spectroscopy, or dye-and-pry analysis. An RCA that concludes "solder joint failure" has described the anatomy, not the cause. The cause is why that joint was insufficient for that stress in that design at that location.
S
Salt spray (salt fog) test
A corrosion test exposing product to a continuous fog of 5% sodium chloride solution at 35°C, defined by ASTM B117. Used to evaluate corrosion resistance of coatings, platings, and base materials. ASTM B117 defines the test method only — not pass criteria. Hours to first corrosion, percentage of corroded surface area, or coating adhesion post-exposure are all possible criteria, specified separately. A salt spray test duration without specified pass criteria is a data point, not a test.
Seeded defects
Defects intentionally inserted into test units to verify the effectiveness of a HASS screen. Seeded defects replicate manufacturing process deviations — cold solder joints, nicked component leads, improper component insertions. If the HASS screen consistently detects seeded defects that represent plausible manufacturing failures, the screen is considered effective. If it misses them, the screen levels are insufficient. Seeded defect testing is the empirical verification of what a HASS screen actually catches.
Setpoint
The target value of a controlled variable — temperature, humidity — that the chamber controller is programmed to achieve and hold. The setpoint is the command. The condition at the DUT is what matters. These diverge when the chamber has poor uniformity, when the product has high thermal mass, when the DUT generates heat, or when fixture placement puts the DUT far from the control sensor. The setpoint tells you what you asked for. A thermocouple on the DUT tells you what you got.
Six Degree of Freedom vibration (6DoF)
Vibration with simultaneous acceleration in three linear axes (X, Y, Z) and three rotational axes (roll, pitch, yaw). HALT chambers generate 6DoF vibration through multiple pneumatic actuators striking the table from different positions and angles. The result is a vibration environment that excites every structural mode of the product simultaneously — more representative of complex real-world environments than single-axis testing, and more effective at finding vibration-sensitive failure modes regardless of their orientation.
Soft failure
A recoverable failure mode. When the stress that caused the failure is reduced or removed, the product returns to normal function. Soft failures define the operating limit in HALT — the boundary of the product's functional envelope. Soft failures are not benign: a product that soft-fails at −55°C will eventually soft-fail in any field environment that approaches that temperature, and soft failures that are not corrected become hard failures under repeated stress.
Stability
The ability of a chamber to maintain a setpoint over time at a fixed sensor location. Measured as variation at a single point over a defined period at steady state. Distinguished from uniformity, which measures variation across locations simultaneously. A chamber can be stable — consistent over time at one point — but not uniform — with large gradients between points. Most calibration certificates report stability. Most TUS reports report uniformity. Neither tells you what the other measures.
Step stressing
A stress testing process in which stress level is incrementally increased — in fixed steps, with dwells and functional tests at each level — to identify product weaknesses and limits. The systematic approach that makes HALT a repeatable methodology rather than a destruction test. Step stressing reveals where failures occur, at what stress level, and whether they are soft or hard failures — providing the data needed to set both operating limits and HASS screen levels.
T
Test margin
The gap between the stress level at which a product is qualified and the stress level at which it fails. A product qualified to −40°C that fails at −65°C has 25°C of cold-side test margin. Narrow margins are not failures — they are information. A product with a 5°C margin between qualification condition and failure point is one manufacturing variation away from field problems. HALT is the methodology designed to measure test margins and extend them through design improvement before production begins.
Thermal equilibrium
The state in which the DUT has reached the same temperature as the surrounding environment, with no net heat transfer occurring. Thermal equilibrium at the DUT is necessary before dwell time begins for the test to be valid. Achieving it takes longer for products with high thermal mass, dense packaging, or internal heat generation. The only way to verify thermal equilibrium is a thermocouple on or inside the DUT. The chamber air sensor cannot tell you when your product is ready.
Thermal mass
The heat energy required to raise an object's temperature by one degree, equal to mass × specific heat capacity. High thermal mass means slow response to temperature changes. A large aluminium chassis adds thermal mass that slows effective ramp rate, delays thermal equilibrium during dwell, and can sustain the DUT at a different temperature than the chamber air during transitions. Fixture thermal mass compounds this effect. A test programme that doesn't account for thermal mass produces results that are difficult to reproduce.
Thermal ramp rate
The rate of temperature change in a system. As a thermal system specification, it is the fastest change rate the chamber can achieve — measured in the air, at a centrally located point, under empty chamber conditions. Higher ramp rates in HALT produce more thermal stress on the UUT and a more effective test. The thermal ramp rate applied to the UUT is always lower than the chamber's rated air ramp rate, because thermal energy must transfer from air to product through convection, which takes time.
Thermal shock
A test applying near-instantaneous transition between temperature extremes, targeting brittle fracture and gradient-induced failure modes. Distinguished from temperature cycling by transition time: thermal shock transfers between zones in under 30 seconds; temperature cycling ramps at 3–20°C/min. The two tests target different failure mechanisms and are not interchangeable. A temperature cycling chamber cannot perform thermal shock regardless of its ramp rate capability. Governed by IEC 60068-2-14 Test Method Na.
Triaxial accelerometer
Three accelerometers in a single package with three separate outputs, monitoring X, Y, and Z axes simultaneously from a single mounting location. Used in HALT and HASS testing to characterise vibration in all three linear axes at critical UUT locations simultaneously. A single-axis accelerometer positioned at the same location provides one third of the information. In 6DoF vibration environments, missing two axes is missing most of the picture.
TUS (Temperature Uniformity Survey)
A measurement of temperature variation across multiple points within a chamber workspace, conducted simultaneously under defined operating conditions. Required before a chamber is used for controlled testing and at defined intervals thereafter. A TUS that reveals a ±5°C gradient across the workspace means any result is referenced to the chamber's nominal setpoint, not to a known condition at the DUT. Products in different positions experience different actual conditions — which is why TUS data belongs in every qualification test report.
U
UDL (Upper Destruct Limit)
In HALT, the highest temperature at which the product suffers permanent, non-recoverable damage. The upper boundary above which no HASS screen may go. Determined during hot thermal step stressing by identifying the temperature at which damage is observed that does not recover when temperature returns to ambient.
Uniformity
The variation in temperature — or humidity — across different spatial locations within the chamber workspace at a given moment. A chamber with ±2°C uniformity has a 4°C spread between its hottest and coldest zones at steady state. That gradient is invisible to the controller, which reads from a single sensor. Products in different parts of the workspace experience different actual conditions — which matters for test reproducibility, for pharmaceutical stability studies, and for any application where the specified condition must be met at the product, not the control sensor.
UOL (Upper Operating Limit)
In HALT, the highest temperature at which the product malfunctions but recovers. The gap between the UOL and the expected field maximum temperature is the hot-side operating margin. A product with a UOL of 95°C in an application where it never exceeds 70°C has a 25°C hot-side margin. Finding that margin, then extending it through corrective action, is the purpose of hot thermal step stressing in HALT.
UUT (Unit Under Test)
Functionally identical to DUT (Device Under Test). UUT is more common in HALT and HASS contexts; DUT is more common in semiconductor and general electronics testing. Neither term implies anything different about what is being tested — only about who is doing the testing and which community they learned the vocabulary from.
V
VDL (Vibration Destruct Limit)
In HALT vibration step stressing, the gRMS level at which the product suffers permanent, non-recoverable damage. The upper boundary for HASS vibration screens. If HASS vibration exceeds the VDL, production units accumulate damage. The VDL, like all destruct limits, is product-specific and must be determined empirically through step stressing — not estimated from datasheet specifications.
VOL (Vibration Operating Limit)
In HALT vibration step stressing, the gRMS level at which the product malfunctions but recovers when vibration is removed. The gap between VOL and field vibration levels is the vibration operating margin. Extending the VOL upward through design improvements is the purpose of vibration step stressing.
VPD (Vapour Pressure Deficit)
The difference between air's water vapour capacity at a given temperature and its actual water vapour content, in kPa. The primary driver of plant transpiration in growth chambers — not relative humidity. At high VPD (dry relative to temperature), plants transpire rapidly; insufficient root water supply causes wilting. At low VPD, stomata close, slowing transpiration and nutrient uptake. Optimal VPD for most crops: 0.8–1.2 kPa. A growth chamber specification that lists only RH without temperature provides insufficient information to calculate VPD.
W
Walk-in chamber
A test chamber large enough for personnel to enter, for testing large assemblies, complete systems, or multiple units simultaneously. Unique challenges: door seal integrity affects uniformity near the entry; product placement affects airflow patterns across the workspace; personnel safety during hot or cold soak must be addressed in the test protocol. Floor load requirements — typically 400–1,000 kg/m² for a loaded chamber — are frequently underestimated in facility planning, and discovered after the chamber has been installed.
Z
Z-axis CTE
The coefficient of thermal expansion of a PCB laminate measured through the laminate thickness — perpendicular to the board plane. Z-axis CTE for standard FR-4 is 50–70 ppm/°C, three to five times higher than in-plane CTE. This is why plated through-hole barrel cracking occurs under thermal cycling: the laminate expands far more through its thickness than copper does, putting the barrel in tension on every cycle. High-reliability designs use low z-axis CTE laminates — polyimide-based materials — for applications with demanding thermal cycling requirements.
The terms that trip people up most
Stability vs. uniformity. Stability is variation over time at one point. Uniformity is variation across space at one time. A chamber can fail one and pass the other. Most calibration certificates report stability. Most TUS reports report uniformity. Neither tells you what the other measures. Both matter. They measure different problems.
Operating limit vs. destruct limit. Both are HALT terms. Both involve failure. The difference is reversibility. Operating limit: malfunction that recovers. Destruct limit: damage that doesn't. HASS screen levels must stay below the destruct limit. Confusing the two results in screens that damage conforming product during production — which is the opposite of what a screen is supposed to do.
Ramp rate vs. transition time. Ramp rate is °C per minute — the relevant parameter for temperature cycling. Transition time is total seconds between hot and cold zones — the relevant parameter for thermal shock. The two describe different tests targeting different failure mechanisms. A chamber with a fast ramp rate still cannot achieve the transition time of a two-zone thermal shock system. The fastest possible single-zone ramp is not thermal shock.
HALT vs. HASS. HALT is development. HASS is production screening. HALT finds limits and generates corrective actions. HASS verifies that production units conform to a design already improved through HALT. Running HASS without prior HALT data is running a screen with no verified corridor — you do not know whether you are screening within the operating-to-destruct margin or outside it.
DUT temperature vs. air temperature. The chamber air sensor controls the setpoint. The DUT temperature is what matters for the test. They are the same only when the DUT has negligible thermal mass, no self-heating, and perfect airflow contact on all surfaces. That is rarely the case. A qualification test report that does not include DUT temperature data has documented the chamber's condition, not the product's condition.