Sealing performance testing data processing transforms raw pressure, flow, and tracer gas signals into quantified leak rates and defensible pass or fail decisions. This article explains the complete data workflow, from temperature correction and leak rate calculation to statistical analysis and uncertainty evaluation, as applied by accredited third-party laboratories and manufacturers working to ISO, IEC, ASTM, EN, GB, UL, and JIS standards. Mastering these processing rules is essential for reproducible tightness qualification of valves, gaskets, dynamic seals, and sealed packages.
Technical Overview
Sealing performance measurement evaluates the ability of a component, joint, or package to resist the escape or ingress of fluids. However, leak testing instruments never report tightness directly: they produce time series of pressure, differential pressure, mass flow, or tracer gas concentration. Sealing performance testing data processing is the discipline that converts these raw signals into standardized leak rates, tightness classifications, and conformity statements that engineers, regulators, and customers can trust.
The processing chain typically covers five stages: data acquisition with calibrated sensors, signal conditioning and thermal correction, conversion to standardized leak rate units, statistical evaluation of repeated measurements, and conformity assessment against specification limits. Frameworks such as EN 1779 for leak testing method selection, ISO 20486 for reference leak calibration, ISO 5208 and ISO 15848 for valve tightness and fugitive emissions, ASTM F2338 for vacuum decay package testing, GB/T 13927 for industrial valve pressure testing, and UL 157 for gasket and seal qualification all depend on clearly defined data handling rules. Errors introduced during processing, such as uncorrected temperature drift, an inaccurate test volume, or missing background subtraction, remain among the most common causes of false failures in production leak testing.
Testing Samples
Industrial Valves and Pressure Retaining Components
Globe, ball, gate, check, and safety valves are tested for shell and seat tightness according to ISO 5208 or GB/T 13927, and for fugitive emissions at valve stems according to ISO 15848-1 and ISO 15848-2. Samples range from small instrument valves to pipeline valves above DN 600, with internal test volumes from a few cubic centimeters to several liters, which directly affects the sensitivity of the calculated leak rate.
Gaskets and Static Seals
Sheet gaskets, spiral wound gaskets, O-rings, and bolted flange assemblies are compressed in standardized rigs and monitored for leakage under sustained load, pressure, and temperature. Material and assembly qualification often follows UL 157 or relevant ASTM gasket practices, with processed results expressed as a pressure decay slope or a standardized leak rate.
Dynamic Seals
Rotary shaft seals, reciprocating rod seals, and pneumatic cylinder seals generate leakage data that must be corrected for shaft speed, stroke frequency, and pressure cycling. Data processing separates true dynamic leakage from break-in effects, which typically decay over the first test cycles and should be excluded from the reported steady-state leak rate.
Sealed Packages and Enclosures
Medical device packaging validated to ISO 11607, pharmaceutical containers, automotive lamp housings, battery packs, and IP-rated electronics enclosures tested to IEC 60529 are typical samples for vacuum decay testing per ASTM F2338, bubble emission testing per ASTM F2096, or dye penetration testing per ASTM F1929.
Testing Items
After processing, the raw measurement data are converted into the following reportable quantities:
- Leak rate in standardized units such as Pa·m³/s, mbar·L/s, or standard cubic centimeters per second (scc/s).
- Pressure decay rate in kPa/min or Pa/s for direct comparison with process control limits.
- Gross leak or seal integrity classification, reported as pass or fail with a stated detection limit.
- Fugitive emission leak rate in mg/(s·m) referenced to seal perimeter, used to assign ISO 15848-1 tightness classes such as AH, BH, and CH.
- Permeation rate through polymeric materials, separated from mechanical leakage by time-domain analysis of the pressure curve.
- Water ingress and airflow leakage results used to verify IP code ratings per IEC 60529.
Testing Methods
The applicable data processing rules depend strongly on the measurement principle, so the method must always be documented together with the result. The principal methods are:
- Pressure decay method: the part is pressurized, isolated, and the internal pressure drop is recorded over time; the leak rate is derived from the slope of the corrected pressure curve multiplied by the effective test volume.
- Differential pressure method: a calibrated reference volume is pressurized in parallel with the test part, and the pressure difference between the two is analyzed, improving resolution by one to two orders of magnitude.
- Vacuum decay method per ASTM F2338: packages are placed inside an evacuation chamber and the chamber pressure rise is processed to detect channel defects down to the micrometer scale.
- Helium mass spectrometry: tracer gas signals are converted into leak rates using calibration factors traceable to certified reference leaks per ISO 20486, with background subtraction applied before any result is reported.
- Bubble emission method per ASTM F2096: parts are submerged and internally pressurized; leakage is quantified from bubble formation frequency or retained as a qualitative detection limit.
- Mass flow method: a flow sensor in a closed loop compensates gas loss; the stabilized mass flow equals the leak rate directly and requires mainly zero-offset and temperature correction.
Data Processing Workflow
Step 1: Data Acquisition and Validity Checks
All transducers must be calibrated and traceable, with sampling rates high enough to resolve the expected pressure change. After pressurization, a stabilization period is mandatory because the filling process heats the gas adiabatically; the initial portion of the curve, dominated by thermal relaxation rather than leakage, is excluded from the evaluation window or modeled explicitly.
Step 2: Temperature and Signal Corrections
Pressure decay data must be corrected for thermal drift. Under the ideal gas assumption, a temperature change of ΔT produces an apparent pressure change of approximately (P/T) × ΔT, which is subtracted from the measured pressure drop. A practical approach fits the decay curve with an exponential relaxation model and extracts the terminal linear slope, which represents the true leakage while removing filling transients. Zero drift of the sensor and ambient pressure variations are compensated in the same step.
Step 3: Leak Rate Calculation
For the isothermal case, the leak rate is calculated as Q = V × (ΔP / Δt), where V is the effective test volume including fixture piping, ΔP is the corrected pressure drop, and Δt is the evaluation time. Dividing Q by atmospheric pressure yields the standard volume leak rate in scc/s. Any uncertainty in the test volume propagates proportionally into the leak rate, so volume determination by calibration expansion is part of the processing.
Step 4: Gas Species Conversion
Helium leak rates are often converted to air equivalents for specification comparison. In the molecular flow regime, Graham's law gives Q(air) = Q(He) × sqrt(M_He / M_air), which equals approximately 0.37 × Q(He); in the viscous regime the conversion factor approaches unity. The applicable flow regime must therefore be recorded with the data.
Step 5: Statistical Evaluation
Repeated measurements on identical samples are reduced to the arithmetic mean, standard deviation, and relative standard deviation. Outliers are screened using the Grubbs test at a defined significance level before averaging, and 95 percent confidence intervals are attached to the mean value. For production capability, the lower one-sided performance index against the maximum allowable leak rate is calculated from the distribution of processed results.
Step 6: Uncertainty Evaluation and Decision Rules
Following the ISO/IEC Guide 98-3 (GUM) methodology, an uncertainty budget is built from transducer calibration, volume determination, temperature measurement, timing resolution, curve fit residuals, and background instability. The combined standard uncertainty is obtained by root-sum-square summation of the individual contributions, and the expanded uncertainty is reported with a coverage factor of k = 2, corresponding to approximately 95 percent confidence. Conformity decisions then apply the laboratory's stated decision rule, for example a simple acceptance or a guard-banded acceptance that keeps the risk of false pass below a defined level, consistent with ISO/IEC 17025 requirements.
Applications
Processed sealing performance data support critical engineering and regulatory decisions across many industries:
- Automotive: leak testing of fuel systems, brake circuits, cooling loops, EV battery packs, and lamp housings, where temperature-corrected leak rates are compared against plant control limits.
- Oil and gas: valve shell and seat tightness certification to ISO 5208 and GB/T 13927, plus fugitive emissions qualification to ISO 15848 for stem seals and body joints.
- Medical and pharmaceutical: sterile barrier integrity and package seal validation to ISO 11607 using vacuum decay and bubble emission data with statistically justified detection limits.
- Refrigeration and HVAC: tightness verification of coils, compressors, and assembled systems, including helium leak rate conversion to refrigerant equivalents.
- Electronics and lighting: ingress protection verification to IEC 60529, combining pressure-based leakage data with water ingress test outcomes.
- Aerospace and energy: qualification of fittings, valves, and welded assemblies where full uncertainty budgets and traceable processing records are contractually required.
FAQ
Why must pressure decay data be corrected for temperature?
A temperature change of a single kelvin in a small test volume can produce an apparent pressure signal larger than the true leak. Without thermal correction, parts can fail or pass incorrectly, so the corrected pressure drop, not the raw reading, must always be used in the leak rate calculation.
How is a helium leak rate converted to an air equivalent value?
In the molecular flow regime, multiply the helium leak rate by the square root of the molecular mass ratio, giving an air equivalent of roughly 0.37 times the helium value. In viscous flow the factor approaches one, so the flow regime must be confirmed before conversion.
How many repeat measurements are required?
Standards such as ISO 15848-2 define the required number of test cycles for valve qualification, while general laboratory practice uses at least three to five replicates to estimate mean leak rate, standard deviation, and confidence interval. More replicates are needed when the leak rate is close to the specification limit.
What is the difference between a leak rate and a tightness class?
A leak rate is a quantitative result in units such as Pa·m³/s or mg/(s·m), while a tightness class is a graded classification, such as the ISO 15848-1 classes AH, BH, and CH, assigned by comparing the processed leak rate against class thresholds under defined test conditions.
Can a part be rejected even when the measured leak rate is below the limit?
Yes. Under guard-banded decision rules, the measured value plus the expanded uncertainty must remain below the limit. If uncertainty overlaps the specification limit, the result is reported as conditional or rejected to control the risk of a false acceptance.
Which records should accompany processed leak test data?
A compliant report includes the test method and standard, calibration traceability of sensors and reference leaks, test pressure and medium, effective test volume, temperature correction procedure, raw and corrected curves, calculation parameters, statistical summary, uncertainty budget, and the decision rule applied for the conformity statement.