Detecting a Slow Leak in a Helium-Traced Vacuum Chamber
A vacuum chamber with a design leak-rate limit of 1×10⁻⁷ atm·cc/s is tested by the helium spray (hood) method against a mass spectrometer leak detector. The chamber is pumped to 5×10⁻⁶ torr and the detector is calibrated with a certified calibrated leak. Background helium rises steadily to 4×10⁻⁸ atm·cc/s after the chamber is isolated. The technician sprays helium over a suspect weld flange and sees the signal climb from 4×10⁻⁸ to 9×10⁻⁸ atm·cc/s over about 30 seconds and then hold. The technician concludes the flange leaks at 9×10⁻⁸ atm·cc/s, which is close to the limit, and recommends reworking the flange. The system's response time on this chamber is estimated at 20 seconds for helium to reach the detector from the far side of the chamber.
1. Is the technician's interpretation of a 9×10⁻⁸ atm·cc/s leak at the flange defensible? What should be done before reworking the flange?
Read the worked answer
Not yet defensible. The observed signal is a rise from a background of 4×10⁻⁸ to 9×10⁻⁸ atm·cc/s - a net change of 5×10⁻⁸, not a steady 9×10⁻⁸ leak. Because the system response time (~20 s) is comparable to the observation window (30 s), the signal may still be equilibrating. Also, the background itself is rising, which contaminates the measurement.
Before reworking: (1) verify the calibrated leak and detector linearity in this range; (2) wait until the signal is stable for several response times (e.g., 3×20 s = 60 s) after spraying; (3) stop spraying and confirm the signal decays back to background - a real leak at the flange will decay, while helium from elsewhere or system outgassing may not; (4) re-check the background with helium spray off; (5) if the net rate is still near the limit, isolate the flange with a hood or bag and measure the accumulation rate. Only after a repeatable, decaying, net signal is confirmed should rework be considered.
2. What is the significance of the rising background helium, and how does it affect the leak-rate determination?
Read the worked answer
A rising background indicates helium is accumulating in the chamber or detector from a source other than the spray - e.g., residual helium from prior tests, permeation, or a leak in the detector plumbing. It biases every reading upward and can mask or mimic a leak. The true leak rate attributable to the flange is the difference between the sprayed and unsprayed steady-state signals, corrected for the response time. If background is not stable, the measurement is not valid; the chamber should be purged or the detector zeroed and the test repeated.