Appendix 5 actual vacuum sample bottle. This is done to “update” the sample line with a fresh ullage gas sample before opening the sample bottle to acquire the gas sample.
(b) Each sample location will have an associated series of vacuum sample bottles, depending upon the number of samples required and the individual test setup.
Typically these bottles are analyzed after completion of the test for oxygen content and whatever additional gas analysis is desired, by a laboratory that specializes in this kind of gas sample. It is critical to have a simple, well documented, plan for gas sampling to ensure that test personnel acquire the required gas samples at the appropriate time and allow for adequate documentation. As is always the case with human/test interaction, a series of procedures and checks are required to minimize the possibility of human error.
FIGURE 5-2
Vacuum Bottle Sample System Installed on an NTSB Test Aircraft
(c) The primary safety concerns with a vacuum bottle sample system are similar to those for a continuous sample system. All sample lines penetrating the pressurized bulkhead should be leak checked and shrouded to preclude the possibility of ullage gas or fuel entering the cabin. Additionally, each sample bottle series should be housed in a shroud which is ventilated overboard to the cabin exterior. The advantage of analyzing the gas samples after the flight test is that no additional safety features are required to protect the aircraft fuel tank from the potential ignition source of a sensor or associated gas sample regulation equipment.
A5-7
Appendix 5
(2) Single Point Calibration and Sampling.
(a) It is possible to use an analyzer and sample train to obtain discrete oxygen concentration measurements during a flight test. This can be accomplished by rapidly calibrating an analyzer sensor with a calibration sample gas, and then exposing the sensor to the desired ullage gas sample. This method would probably only be effective for measuring the oxygen concentration at times and tank locations with a relatively stable pressure, temperature, and oxygen concentration within the fuel tank ullage.
(b) All of the safety considerations associated with single point calibration and sample methods are identical to those of continuous sampling methods. For this reason alone, it may not be desirable to apply this methodology to a flight test scenario.
Most of the cost and complexity of a gas sample system installation for a flight test environment is attributable to the safety features (sample line shrouds, box ventilation system, and system safety analysis). To apply these methods and not install the additional equipment to continuously sample, at a relatively small cost and complexity, does not seem cost effective. Again, every installation and application needs to be analyzed for certification requirements, data ranges, and program risk, within the confines of the regulations, to ensure a successful, cost effective flight test.
3. Selections of Sample Locations. The primary consideration when selecting the location and number of gas sample ports is the physical geometry of the fuel tanks, as well as the compartmentalization of the fuel tank. The two types of fuel tanks considered are single bay tanks and multiple bay tanks.
a. Single Bay Fuel Tanks.
(1) Single bay fuel tanks have a single volume, broken up only by structural members, which do not impede the flow of gases inside the tank. It is possible to select a single gas sample location that is representative of the entire ullage average oxygen concentration, but without prior measurements during similar inerting testing or extensive computation calculations, it is difficult to ensure a single measurement location is representative of the entire ullage. For fuel tanks where air enters the fuel tank during descent, a continuous measurement of ullage oxygen concentration during a specified aircraft descent profile is essential to determining the level of protection afforded the tank by the inerting system and is the best measure of the sizing of inerting system for the specified fuel tank. The critical sampling location would typically be located near where outside air enters the tank. Therefore if a single sampling location is proposed, substantiation that the selected sampling location will provide oxygen measurements that demonstrate the tank is inert must be provided.
(2) Measurements too close to the airplane vent (air deposit) or the NEA deposit nozzle can provide a biased measure of the average ullage oxygen concentration.
It is possible that during the mixing of NEA and vent air into the tank, brief pockets of A5-8
Appendix 5 high or low oxygen concentration ullage gas may exist within the ullage. Without compartmentalized bay walls, these pockets will be transient and difficult to identify.
Mixing can be problematic in the vertical direction (between the top and bottom of the tank), particularly when considering low flow rates of air or NEA entering a tall thin fuel tank. If a fuel tank has associated radiant heat on the bottom (heated CWT), this problematic mixing will be small in nature and short lived.
(3) At the end of every test the aircraft fuel tank ullage will have a measured average ullage oxygen concentration after the aircraft has come to a complete stop and the inerting system has been shut off. By continued examination of this average over a period of 20-60 minutes during the operation of the aircraft air cycle machines (bottom heat), good vertical mixing can be validated. Any change in the average ullage oxygen concentration after the test is complete, with no additional air or NEA is entering the tank, can only be caused by changing oxygen concentration readings due to the mixing of the resulting ullage constituency. Again, without prior knowledge, the only way to ensure the inerting system is appropriate for the aircraft fuel tank is by the continuous measure of oxygen concentration at several strategic locations within the ullage during an appropriate descent flight test.
b. Multiple Bay Fuel Tanks.
(1) Multiple bay tanks are fuel tanks broken into several smaller volumes by solid or nearly solid web structures. These structures can be somewhat limited in their partitioning with large openings and extensive gaps between the web and the tank structure as is the case with many wing fuel tanks. These web structures can also be very solid in nature, with the only large openings being doors removed during maintenance and small web gaps for equalizing fuel and ullage gas, making a very compartmentalized tank.
(2) Fuel tanks with relatively open partitions may only require a few gas sample locations across many bays to allow for the accurate measurement of the overall fuel tank ullage average. Previous research highlighting inert gas distribution behavior during ground inerting may be beneficial to determining the number and location of gas sample ports. However this is not a guarantee of good air redistribution given the different nature of the NEA deposit and the vent opening. In the end, several more oxygen concentration sample locations may be necessary to allow for validation of inert gas distribution and ullage gas mixing during certain flight tests. These additional sample locations can be utilized when the mixing of the tank needs to be examined or validated using oxygen measurement channels redundant at that time.
(3) In the case of very compartmentalized fuel tanks, inert gas distribution may be problematic, depending on the sizing of the inerting system and the NEA deposit scheme selected. This generally requires at least one gas sample in each compartment, with long thin compartments perhaps requiring two gas sample locations. Bays that have a large vertical height, while still remaining relatively thin, may require an additional sample port at the bottom of the bay. Again, any differences in the oxygen concentration A5-9
Appendix 5 of the top and bottom of an ullage compartment will be transient in nature, but the knowledge of this behavior may allow for an explanation of an unusual measurement obtained and could allow for the validation of fuel tank inerting models being utilized.
4. References.
1. Advanced Instruments, “GPR-29% O2 Analyzer Owner’s Manual.”
2. Doebelin, E., Measurement Systems Application and Design, McGraw-Hill, 1990, pages 729-730.
3. Rosemount Analytical, “Model 755R Oxygen Analyzer Instruction Manual,” 748213-S, April 2002.
4. Krihak, M., Murtagh, M., and Shahriari., M. R., “Fiber Optic Oxygen Sensors Based on the Sol-Gel-Coatings Technique”, SPIE Chemical, Biological and Environmental Fiber Sensors VIII Conference Proceedings, Volume 2836, Pages 105-115.
5. Burns, Michael and Cavage, William M., “A Description and Analysis of the FAA Onboard Oxygen Analysis System,” FAA report DOT/FAA/AR-TN03/52, June 2003.
A5-10