The Two Dominant Methods in Industrial Leak Testing
Most industrial production leak testing falls into two categories: direct pressure decay and differential pressure. Choosing the wrong method for an application is the most common cause of chronic false rejects, inadequate sensitivity, or unnecessary test cycle length.
Direct Pressure Decay Testing
In direct pressure testing, the test cavity is pressurized to a target level, isolated from the supply, and held for a defined stabilization period. A pressure sensor measures the absolute pressure in the cavity during a measurement window. Any leak causes the pressure to drop; the tester compares the measured decay against a pass/fail threshold.
Characteristics:
- Resolution: Typically 0.1–1 kPa, limited by sensor noise floor and thermal drift
- Test pressure range: Wide — vacuum to over 1200 kPa in a single instrument (e.g., Figurtech LY Series)
- Cycle time: Fast — as short as 15–25 seconds for automotive and electronics applications
- Best for: Components with large internal volumes, high-pressure housings, IP67/IP68 waterproof testing, EV battery cooling circuits
- Limitation: Thermal effects limit sensitivity. A 1°C temperature rise in a 100 mL cavity produces approximately 0.33 kPa pressure increase — which can mask or mimic a small leak
Differential Pressure Testing
In differential pressure testing, the test cavity is pressurized simultaneously with a hermetically sealed reference chamber of identical volume. A high-sensitivity differential pressure sensor measures the pressure difference between the two chambers. Because both sides experience the same ambient temperature and barometric changes, common-mode disturbances cancel out in the differential measurement.
Characteristics:
- Resolution: 0.01 kPa or better — more than 10× higher sensitivity than direct pressure methods
- Test pressure range: Typically vacuum to 700 kPa (Figurtech LD Series)
- Cycle time: 14–120 seconds depending on test volume and required sensitivity
- Best for: Small-volume precision components, medical devices, surgical instruments, sealed bearings, low-pressure applications (10–100 kPa)
- Key advantage: Ambient temperature fluctuation, barometric pressure changes, and HVAC cycling affect both DUT and reference identically — producing zero differential output. This enables reliable testing in non-climate-controlled production environments
Side-by-Side Comparison
| Parameter | Direct Pressure Decay | Differential Pressure |
|---|---|---|
| Typical sensitivity | 0.1–1 kPa | 0.01 kPa |
| Max test pressure | 1200+ kPa | 700 kPa |
| Thermal drift rejection | Partial (on-board compensation) | Inherent (bridge architecture) |
| Cycle time | 15–30 s typical | 14–120 s depending on volume |
| Minimum detectable leak | ~3–10 SCCM at 40 mL | ~0.3 SCCM at 40 mL |
| Best application | High-pressure, large volume, IP ratings | Low-pressure, small volume, medical |
| Cost | Lower | Higher (precision bridge sensor) |
Mass Flow Testing: A Third Method
For components with inherent bore geometry variation (such as capillary tubes, flow-control valves, or membranes), neither pressure method provides a reliable go/no-go criterion. Mass flow testing measures the actual volumetric flow rate through the DUT under steady-state pressurization. The accept window is defined by the correct flow characteristic of a properly manufactured part, not an arbitrary pressure threshold. The Figurtech LF Series supports flow ranges of 0–10 L/min and test pressures up to 500 kPa.
Selection Guide
- Use direct pressure if: test pressure > 300 kPa, volume > 200 mL, or cycle time < 20 seconds is required
- Use differential pressure if: leak limit < 1 kPa decay, test pressure < 300 kPa, or ambient temperature variation is a concern
- Use mass flow if: the part has inherent dimensional variation that makes pressure-based testing unreliable
Figurtech application engineers can review your specific leak specification, component volume, and cycle time requirement to recommend the appropriate instrument platform. Contact us for a technical evaluation.