Leak Detection Expert

Today’s question: How does a leak rate change when changing the fill gas pressure?

In general, you have to differentiate between two different types of gas flow inside the leak channel:

1. Laminar or viscous flow predominates at leak rates above 10^-5 mbarl/s. In this case, the pressure impacts the leak rate with the second power (leak rate depends on increasing pressure to the second power).
   
   
   
   p₁, p₃ = fill gas pressure in bar, absolute
   
   p₂, p₄ = pressure on detection side (~0 bar for vacuum leak testing, ~1 bar for sniffer leak testing)
    
   
2. Molecular flow occurs at leak rates below 10^-7 mbarl/s. Under these conditions, the leak rate changes linear with changing pressure.
   
   
    
   
3. Leak rates in the range of 10^-5 mbarl/s to 10^-7 mbarl/s cannot be calculated exactly. The true value will be somewhere between the two specific cases.
    

Examples:

• Vacuum leak test
  Leak rate at original fill gas pressure (5 bar › 0 bar) = 1·10^-6 mbarl/s
  
  New fill gas pressure 10 bar › 1 bar
  
  Case 1 (laminar flow)
  Leak rate = 1·10^-6 mbarl/s · (10²-0²) / (5²-0²) = 4·10^-6 mbarl/s
  
  Case 2 (molecular flow)
  Leak rate = 1·10^-6 mbarl/s · (10-0) / (5-0) = 2·10^-6 mbarl/s
  
  
• Sniffer leak test
  Leak rate at original fill gas pressure (5 bar › 1 bar) = 1·10^-6 mbarl/s
  
  New fill gas pressure 10 bar › 1 bar
  
  Case 1 (laminar flow)
  Leak rate = 1·10^-6 mbarl/s · (10²-1²) / (5²-1²) = 4.1·10^-6 mbarl/s
  
  Case 2 (molecular flow)
  Leak rate = 1·10^-6 mbarl/s · (10 – 1) / (5 –1) = 2.3·10^-6 mbarl/s

To make sure no failed parts pass a leak test, the rejection leak rate (i.e. trigger setting) should be calculated assuming laminar conditions (case 1). Thus, leak rates in this transition range will be measured slightly higher than they really are, but never too low.