Effective Leak Testing: Are you testing the part or the seal?

Effective Leak Testing: Are you testing the part or the seal?

By Robert Plumridge, Leak Test Specialist, Sciemetric Instruments

Leak testing is a crucial quality assurance process – leaky parts that slip through the cracks lead to problems down the line or warranty claims. But operators often lack the expertise and the tools to ensure a consistently reliable seal between the part and the test station. This can skew test results, allow faulty parts to pass and fail perfectly good ones, and erode confidence in the leak test system even if it isn’t the source of the problem.

What can operators and quality engineers do?

The two most common types of leak test are pressure decay and mass flow. With pressure decay, the part is filled with a volume of air to a specific pressure, then isolated from the air source. The pressure within the part is then monitored to track decreases that indicate the presence and rate of a leak. With mass flow, the part remains connected to the air source to maintain the target pressure within the part. A flow meter tracks how much air must be continuously pumped into the part to maintain the target pressure as air escapes through a leak or leaks.

The best practices described in this article, to ensure a reliable seal between part and station, apply regardless of which type of leak test is best for a given scenario.

Based on our experience with customers in industries such as automotive, off-highway and medical, here are the top considerations to ensure operators are in fact testing the part and not the seal.

1) Picking the best seal for the job

There are two types of seals used during leak testing: Seals that double as connection points for the leak tester, and seals that plug additional openings in the part.

The best seals have a mechanical component that physically locks them to the part and ensures proper orientation. This can come in the form of:

  • A mechanical or pneumatically (air) activated seal that is physically deformed to provide the mechanical lock
  • A bolted connection
  • An automated seal on the end of a ram or robot that is locked into position during testing.

The best seal designs, weather mechanical or pneumatic, provide feedback to indicate that proper seating pressure and orientation was obtained after installation. An electronic sensor internal to the seal, proximity switches for automated rams and simple travel stops on manually actuated seals provide tactile feedback that ensures the seal is properly inserted.

If pneumatic seals are used, best practice is to also measure and control the pilot air to the seals during the test to ensure required pressure was maintained for the duration of the test. It’s also important to ensure there isn’t too much pressure that will accelerate how quickly the seal wears out.

Fig. 1: In this waveform of a flow rate leak test cycle, the green line indicates the expected process signature. The blue line indicates pressure deviations due to the movement of the part at the 50-second mark that affected its seal to the test station. This deviation could be mistaken for an actual leak that could lead to a false fail for the part.

2) The ideal safety pressure ratio

The force holding the seal must be three times the force acting against the seal from the test pressure within the part. This prevents false leaks through the opening. A 3-1 ratio ensures the seal is “crushed” or compressed enough to be air tight, without being too tight and contributing to seal creep (the gradual wear and weakening of the seal, or movement of the seal during test).

This is determined by Force = Pressure X Interior Area.

For example, a one-inch diameter circular hole being sealed by a face seal mounted on a pneumatic ram with a test pressure of 5 PSI would give a total force of Area (Pi x Diameter) X Pressure, or 1 X Pi X 5, or 3.14 X 5, or 15.7 pound force (lbf). This means the pressure applied to the seal must be 15.7 X 3(safety factor) to assure no false leak. Or at least 48 lbf. 

This guideline is used to size the cylinder that will hold the seal that will seal the hole:

  • A ¼-inch diameter bore cylinder X 60 PSI shop air pressure X Pi (3.14) yields 47.1 lbf (marginal too low)
  • A 3/8-inch diameter bore cylinder X 60 PSI shop air pressure X Pi (3.14) yields 70.65 lbf (just right)
  • A ½-inch diameter bore cylinder X 60 PSI shop pressure X Pi (3.14) yields 94.2 lbf (OK, but may cause seal wear)

3) Choosing the sealing surface

Seals operate best on clean machined surfaces. Typically, the part geometry will dictate the seal type. In the case of a machined pipe, a seal forced to expand against the interior of the part will provide the most reliable air tight seal. The next best option is an exterior seal, and then a face seal.

4) Not too hard, not too soft, just right

The durometer, or hardness, of the seal must be selected for the environment and part type involved, the test pressure and how many test cycles the seal must endure in a production shift. Too soft, and the seal material will wear too quickly. Too hard and it will not allow the operator to make a quick and reliable seal. The key is to strike the right balance to ensure the longest possible life, depending on the manufacturing environment and the contact surface.

High durometer seals will have a longer life span and experience less creep during a test, but also require more uniform surfaces. As a general rule, the hardest seal that provides a consistent leak free connection with the part is advisable to maximize the seal’s life cycle and minimize maintenance activities. Companies that manufacture sealing materials and fixtures can provide guidance on the best product for each unique situation.

5) Part connections

In the case of a seal that is also acting as a filling point, care must be taken to select the right hose and connections to link the leak tester to the part without effecting the seal’s performance. It’s best to match the diameter of the fill port seal with the supply line from the leak tester. This maximizes the speed at which the part can be filled. Hoses must be of a material that will not deform under pressure. The flexibility of the hose and the rigging must be carefully selected so that the hose does not put undue stress on the seal, creating inconsistencies in the clamping force.

Fig 2: In this waveform image, the inconsistency in the flow rate of the leak test, as indicated by the volatility in pressure readings (blue line in the area circled in red), again indicates part movement that is affecting the seal between the part and the test station.

6) Mitigating the human element

During a recent head-to-head trial at a major automaker’s plant, two leak test stations with two different testing systems were set up. We wanted to do a true, objective panel-to-panel comparison. The same operator took five parts through each station six times. They manually clamped and unclamped the parts from each station for each test cycle, just as they would do during a regular production shift.

The trial yielded surprising results – documented air pressure variances that suggested some parts were leaking enough to be considered a fail.

But then we looked at how the operator clamped and unclamped the parts. We ran the tests again without unclamping and saw a substantial increase in the consistency of the results when we applied process signature analysis to compare the data. The problem hadn’t been leaky parts but leaky seals. The cause was operator error.

Overlaying the waveforms from the base test and the unclamped retests showed that the pressure decay curves were inconsistent for the re-clamped tests. Not only were the slopes of the waveforms different, but there were abrupt changes in the pressure that indicate a physical part movement – in this case, the seal was “burping” and moving position (see Figures 1 & 2). The fixture which held the part had lost alignment, and allowed the part to move if the operator did not take great care in positioning the part and engaging the seal.

It isn’t always an error. Sometimes it’s operator preference – some prefer a looser fit than others. In instances where the test uses a softer seal, this can cause problems. In the above example, the clamping force on the seal had been reduced to make the seal easier to engage. This further aggravated the issues we experienced.

Investing in operator training is of course important to address this. But another step is to use air-actuated connectors with positive stops. These allow for fast, safe and repeatable connections that all but eliminate the human element.

In summary, a reliable test starts with a reliable seal

Confidence in your leak test begins by reducing or eliminating all the external factors that can undermine accuracy and repeatability. Quality engineers and operators must understand and compensate for variations in the air supply, the thermal effects of compressed air, the impact of ambient air and part temperature variations, and so on. Identifying and addressing these challenges is much easier if the best possible combination of seal and fitting is used for that crucial connection between the part and the test station. It all rests on having the right combination of equipment, station setup and operator training.