The Pioneer of Modern Leak Testing

Determining Leak Rate by Specification NOT by Sight

Why using actual test pressure to set leak test specification is key to accurate reject limits.

Man inspecting part that failed leak test.To clearly explain why determining leak rate by specification and not by sight is important, let’s begin with an example:

For several hours, an operator has leak tested production parts and upon completion of testing, he has sorted the “acceptable” parts into one bin, and the “rejected” parts into another bin.

The rejected parts are then taken to a water dunk tank for visual bubble monitoring. As air is reintroduced into the part, one can determine the leak’s location.

As expected, the majority of “rejected” parts display a visual stream of bubbles once submerged in water. However, a collection of rejected parts do not display any visual bubble formation to the operator.  So, are these parts truly “rejected,” or was the tester mis-calibrated?

Historically, parts that fall into the above described scenario are classified as “can’t find.”

Porosity Too Small for The Eye To See

A leak tester monitors the accumulative pressure change which occurs in a product over a given time, thus  determining the leak rate. What happens with “can’t find” parts is possibly more leakage locations but with porosity too small for the eye to see.

For example, a single point of leakage 10 cc/min. at 10 psig, would be very easy to see underwater. But, 20 points of leakage, each 0.5 cc/min. would not be visible, yet the product would be rejected by the electronic tester due to accumulative pressure loss and tester sensitivity and resolution.

Setting New Leak Test Specifications Using Actual Test Pressure

To properly set testing specifications, one should use the actual test pressure (for pressures under 200 psig) as the normal pressure encountered because this pressure is the typical operating pressure of the part under service conditions. However, this pressure may be lowered for practical reasons, such as available air supply pressure, safety, and test fixture design. 

An Example of the Effect of Test Pressure on Allowable Leak Rates

Let’s assume that you set a 20 psig test pressure based on the application pressure and the maximum air allowable leak rate was 20 cc/min.

Further, let’s assume there is a justified reason for reducing the pressure to 10 psig, perhaps due to the large opening that seals the part and requires high clamping force. Since this is lower than the actual application pressure, you must adjust your maximum air leakage rate correspondingly about 10 cc/min. While this is just an approximation, it works linearly with gauge pressure, in practice, on test pressures under 100 psig. When the test pressures get considerably higher than that, the reduction in the maximum allowable air leakage rate does not reduce linearly.

You can verify this by testing leakage at various pressures on actual parts. The leakage from these parts must be within the maximum allowable leak rate range.

Maximum Allowable Air Leak Rate

Engineers discussing correlation testingThe real test of any specification is to use actual parts with pre­measured air leakage rates from a conventional manufacturing process and then put those parts in a service test condition. By testing various samples with pre-measured leak rates, the sample tests determine where the maximum allowable air leakage is detrimental to the parts operation, e.g., external leakage or other causes. Keep in mind, there is a point that you must consider from which air does leak and liquid does not.

The maximum allowable air leakage rate varies between aluminum, cast iron and plastic parts. Aluminum tends to be more porous and can stand more maximum allowable air leakage rates since it’s normally a multi­-path leakage. However, this isn’t normally true in cast iron parts. Cast aluminum, lost foam aluminum, and cast iron all test differently so be sure to adjust accordingly.

Don’t Forget: Historical Specifications

You must realize that with the advent of new parts and materials, the change of specifications is worth investigating. However, you must respect the historical testing specifications, particularly for automatic leak testing in your evaluation. Sometimes a specification on a new part is a copy from the old part, and the old part specification was copied from an older part. And the specification on the older part was copied from an unrelated part. The modification of historical specifications is not improper and in many cases, they are too conservative and have been unnecessarily costly in reworking rejected parts.

So if the specifications say “No Leak” and they are discussing air leakage, remember this is not a valid or realistic specification. “No Leak” liquid is a valid specification but must be realistically clarified for air pressure decay, water immersion, or gas detection methods.

1 cc Leak Rate Bubble Demonstration:

Ask For Help


Product Literature

The complete library of literature for all our products is now online.

Access Literature Club

White Paper

Download the new FDA 21 CFR Part 11 for Medical Device Leak Testing.

Download Now

Contact Us


Uson L.P. - Corporate

8640 N. Eldridge Parkway
Houston, TX 77041 USA

Tel: +1-281-671-2000
Fax: +1-281-671-2001


Uson UK – Europe

Western Way
Bury St. Edmunds
Suffolk, IP33 3SP United Kingdom

Tel: +44-1284-760606
Fax: +44-1284-763049


Uson – China

Room 601, 567 Lan Gao Road
Shanghai, 200333, China

Main: 021-583 65859, 021-583 59187
Fax: 021-583 59185, 021-337 73559