A study of USAF aircraft fuel/seal leakage.

The role of the fuel seal o-ring on board military aircraft is another of the unseen-if-successful-but-highly-visible-if-not applications encountered so often in the sealing industry. Traditional o-rings generally work quite well as aircraft fuel seals, but the varieties of different fuels and flight conditions do present a spectrum of shifting demands.

This article will discuss in brief and general terms the application of the o-ring as an aircraft fuel seal. Some recent work by the United States Air Force, Wright Laboratories Materials Directorate, Systems Support Division, and the University of Dayton Research Institute has been directed to this subject and future efforts are under consideration.

Aircraft fuels must be used on an as-available basis in all locations and under all climatic conditions. The temperature and pressure ranges do not vary so much extensively as suddenly. But it is the sudden change from one fuel media to another, referred to as "switch loading," that is thought to be especially troublesome as different environments of chemistry and physics are rapidly presented to the seal.

For many years either the nitrile or the fluorosilicone elastomer has been used for aircraft fuel seals. The demands of the seal in principle do not appear strenuous. The temperature range cannot be given specifically because of the variability of climate. aircraft and on-hoard seal location, but -54 degrees C to 72 degrees C (-65 degrees F to 160 degrees F) is a reasonable overall estimate. Maximum fuel line pressures are not more than about 865 kPa (say, 125 psi), and usual operating pressures are about one-half of that value. At times the fuel lines and tanks am dried ouL and the fuels are necessarily of low viscosity throughout the temperature range. These latter conditions can provide an unusual stress on the o-ring. But it is the fuels themselves, i.e. their varying compositions, that may present the most demanding conditions. By tar the most common fuel in use, stateside, is JP-4, and, as we will see, there is JP-4 and there is JP-4, and then there is JP-8.

The U.S. Air Force can justly be taken as a world-wide facility. Outside the U.S., particularly in Europe, the most common fuel is JP-8. The two fuels, JP-4 and JP-8, present contrasting media to the seals, the change from one to the other ("switch loading") may be especially critical in going from JP-4 to JP-8. Within the next five years the Air Force bases in the continental United States plan to change from JP-4 to JP-8. This will be done on a region by region basis. and the opportunity for switch loading should be greatly decreased. In the meantime, the problem remains and, in fact, may never disappear completely.

The composition of JP-4 appears especially critical. JP-4 is described as a "wide cut, gasoline type" fuel, whereas JP-8 is more like kerosene (much less flammable than gasoline). Their flammability and volatility differ significantly. Several relevant properties of these fuels and two common hydraulic fluids are presented in table 1. The flash point of JP-8 is controlled by specification to a minimum of 38 degrees C. There is no specification value for JP-4, its flash point is below room temperature. The compositional variation in the fuels has an even greater effect on the elastomers. Both fuels can allowably contain up to 23%.-27% aromatics with a maximum of 5% olefins. But up to 15% of the aromatic content of JP-4 can be of low molecular weight compounds - benzene, toluene, the xylenes and ethyl henzene. JP-8 will typically contain only 1% of these compounds. These low molecular weight aromatics promote swelling of the rubber and as a consequence the voltune swell of the seal in JP-4 is ordinarily about twice that in JP-8. i.e. when the allowable aromatics are in the high range. Thus if an airplane, running on JP-4, is suddenly refueled with JP-8 or a "different" JP-4 (an unavoidable situation in all cases), the elastomeric seal may tend to relax, and fuel leakage can result. This is the most accepted theory at present to explain aircraft fuel leakage at the o-ring.

The source of the JP-4 is behind this variability. Alaskan crude produces a higher content of the lower molecular weight aromatics than does the crude from the lower 48 states. It may be needless to point out but there are stores of JP-8 stateside and quantities of JP-4 overseas, and this entire fuels picture from refinery to fuel tank is well controlled by the appropriate specifications.

The specification control of the elastomers and the fuel systems, however. is not as direct. Nitrile rubber is the oldest of the synthetics. It has enjoyed many years of service, has a wide experience/data base, and is quite economical. Fluorosilicone is one of the most versatile elastomers in the marketplace, it has the widest temperature application range of any synthetic elastomer except silicone, but it tar surpasses silicone, as well as nitrile, in its resistance toward organic liquids. For the past several years it has been the first choice for aircralt fuel lines. It is also some 30 times more expensive than nitrile. A third elastomer being investigated in this work is of the phosphonitrilic family.

The performance specification lot fuel seals is MIL-P-5315, and nitrile has been the choice for years to meet its reqnirements. Fluoro-silicone rubber, on the other, hand must meet the requirements of the fluorosilicone elastomer specification, MIL-R-25988, and o-rings qualifying to this specification "are recommended for use in packing glands as specified in MILG-5514." Those experienced in the sealing of hydraulic fluids will recognize this latter specification as the all controlling document for hydraulic fluid systems. There does not appear to be a like document for fuel systems although most designers do adhere to ARP 1232. This is the Aerospace Recommended Practice document, "Gland design, elastomeric o-ring seals, static radial," available from SAE. A comparison of specification demands between MIL-R-25988 and MIL-G-5514 is not meaningful, they are simply too different. Further there is no criticism implied here toward this practice, overall material specifications are commonplace.

A particular elastomer is certified for seal use throughout each aircraft and both of these elastomers, from a physical property/fuel compatibility point of view, offer the possibility of successful sealing.

The fuels' themselves present yet another dimension to this situation. To put it simply, they are dangerous. Many controls and much care are necessary for their storage and handling. Oftentimes other liquids are used as substitute test fluids. In the case of JP-4 and JP-8 standard test fluids - Type I (for JP-8) and Type III (for JP-4) - are used. Type I is isooctane and Type III is Type I with 30% toluene. These liquids are used, in lieu of fuel, to qualify the fuel line couplers.

A typical fighter may contain 150 fuel line couplers ranging in size from 1.9 cm to 15.2 cm (3/4"-6") diameter.

Two o-rings are installed in each coupler. Qualification of these couplers (under MIL-C-22263) is intensive and includes operation with fluid circulation at temperatures to 94 degrees C and pressures to 1.7 MPa (250 psi). Vibration and intentional misalignments are also included.

To whatever extent these test fluids replicate the behavior of the JP fuels the resultant data are accurate. It should perhaps be noted that there is no switch loading requirement for coupler qualification. The fuel line couplers are basically of two types, one having a variable cavity for the o-ring seal, and the other a fixed cavity. Our work to date has focused on the variable cavity type coupler, although the fixed cavity has now begun to gain wider acceptance and it will become the focus of any future effort.

Our test program comprises an eight day exposure cycle utilizing both fuels, a temperature range from -54 degrees C to 70 degrees C, and a fuel line pressure, to date, of 139 kPa (20 psi). A dry out period is also included. The fuels are switched midway through the test and are non-circulating. This program would be the basis for future work but the pressure will be increased to 416 kPa (60 psi). Thus, the continued effort will have the dual goal of investigating the effects of higher pressure and the comparative performance of the fixed cavity coupler.

There is, however, another, more philosophical, point to address. The sealing of fuels is not approached in a manner like, say, the sealing of hydraulic fluids. The absence of an all controlling document like the MIL-G-5514 specification for hydraulics has been noted, there are two different specifications involved in the qualification of the seals, and adherence to ARP 1232 is voluntary. The seal gland volume in the variable cavity fuel line coupler is estimated at about 120% of the o-ting volume (private communication, coupler manutacturer). This is less than would be recommended for ordinary sealing, it is described as a "design compromise," but is also subject to some variation depending upon field conditions. The o-ring, therefore, in this application, becomes more of a packing than an elastomer expanding against a surface to create a seal. The property of elastomeric volume swell, normally used to enhance if not effect sealing, can actually be working against the process as the swell forces may help "set" the seal in the tighter gland. Our experience indicates that in the field the tendency for the o-ring to "take a set" in this application is all too real. One can see the reasoning behind the increasing adoption of the fixed cavity couplers, but there remains this almost psychological approach to the sealing of fuels.

No doubt behind this philosophy is the fact that fuels are fuels. Their flammability and viscosity demand tight, very tight, sealing, and by and large the current process works. The question we might raise, "is there a better way?" can be offered as an invitation for suggestions. Those experienced, without necessarily something to sell, or those with something to sell and data, might offer their views. Constructive input would be welcomed.

It will be noted in conclusion that there are actually more aircraft fires due to hydraulic fluids than from the fuels. The fluids too are flammable and under considerably more pressure than fuels. Fuel fires are more devastating but less frequent. Again, to a significant degree the present system does work, and an effort like this is only seeking to improve the situation. In fact one of the intents of this article is to solicit information from those experienced in the field as to the direction any future research should take. To what extent is aircraft fuel leakage a problem? We would appreciate comment from personnel experienced in either commercial or military aircraft. Is it the seals or the sealant? At the fuel line couplers or elsewhere? What can you offer in experience, specifics, recommendations?

Those who wish to make an input here are requested to contact Alan Fletcher (WL/MLSE, Wright-Patterson AFB, OH, 45433-6533. Telephone: [phone]). FAX: [phone].