Designing a sealing plug for insertion into an unused terminal in a connector would hardly seem to be an engineering challenge. Why not just measure the air pressure that will build up in the connector when the connector mating process occurs on the assembly line, and make sure the plug retention is sufficiently robust?
Not so simple, it turns out. It took a project by USCAR (United States Council for Automotive Research LLC) to come up with a solution to the overall manufacturing/testing issue.
Is it really critical if a plug eventually pops out because the design was inadequate? The answer is that moisture and other contaminants get into the connector and cause corrosion. With today's electrical interrelationships through data buses, trouble in one connection can affect more than one circuit, and no carmaker wants even one corrosion-prone connector.
The root issue, determined from specially developed test procedures, was the inability to determine the peak pressure in the connector, as air in the existing instrumentation was reducing the reading from the connector. Further, the data flow through the instrumentation was too slow; as a result, key data was being missed, explained Dr. Donald Price, Ford representative in the organization's EWCAP (Electrical Wiring Component Applications Partnership). This group develops common, optimized designs, test specifications, and design guidelines for the industry.
Speaking in April at the SAE 2013 World Congress in Detroit, he reported on new instrumentation for much more accurate results.
Internal pressure test a must
The default test pressure for a connector is 7.0 psi (48 kPa), but if a plug is used, the connector is to be tested against a specification based on the internal pressure, per SAE/USCAR-2 Section 5.6.6. The design could be over-engineered—that is, super-tight—but it could make proper insertion more difficult, and no one wants the likely assembly problems. A further specification (actually a rule-of-thumb recommendation) has been to have a 0.3-mm (0.01-in) interference (at worst-case manufacturing tolerances) between connector and plug. But this has been found to be no more than a rough guide. Larger interferences must be used on larger seals, and the increases needed are not linear, Dr. Price said.
The ideal answer is to design the plug just to be what it needs to be, plus an appropriate safety margin, so it stays in place. That called for better information—i.e., an accurate measurement of the internal pressure reached within the connector. The objectives were to get pressure sampling rates of under 10 ms, accuracy of course, and low inertia—the latter deemed a major issue with mechanical pressure measuring systems, according to Dr. Price.
The alternatives to the mechanical pressure sensor in use that were considered: using CAD and calculating the peak from the volume and the ideal gas law, a fluid displacement model, and a MEMS (micro-electromechanical system) pressure sensor with electronic calibration. MEMS pressure sensing is miniaturized sensing, and is used in many systems and devices from airbag crash sensing to inkjet cartridges.
MEMS sensor selected
The MEMS sensor approach was selected because of the availability of a small size with an air opening that's the same as an automotive wire (2.5 mm/0.1 in). This contrasts with the mechanical sensor, which requires reducers to get down to the size of the electrical connector. That's a reason for the amount of air in the testing system that led to the inaccuracies in the readings that were being taken.
The mechanical sensing system had been found good for measuring stable pressure, but the air pressure in a connector goes through a rise and then fall (as there is air leakage between the wire(s) and connector). Testing had shown a pressure drop of 20% in the first 1000 ms after the connector was mated, so a sampling rate of 1.0/s was too slow and indicated why the mechanical sensor accuracy was inadequate.
However, there was no available MEMS device that would work without major modifications to each electrical connector. So EWCAP realized it would have to develop one, and it did.
A control with a bridge-type resistive circuit was designed, and simplified with a protocol for electronic calibration created in Labview. The sensor and circuitry was packaged in a 40 x 25-mm (1.6 x 1.0-in) box, creating a small handheld device that could be used even in less-accessible locations.
EWCAP did not exclude consideration of the theoretical methods and compared the MEMS sensor readings with those derived from CAD calculations on three different connectors. It found the CAD methods produced much more conservative numbers (i.e., all much higher than the MEMS measurements), in one case 2.75 times the measured number.