Testing Gas-Turbine Aircraft Engines
Gas-turbine aircraft engines, such as those used in nearly all modern aircraft (fighters, attack, helicopters, and transports), must be tested by the U.S. Department of Defense (DoD) to verify their performance and safety specifications (safe operating temperatures, rotor speeds, and vibration levels). These performance tests are very expensive and therefore must be planned carefully and executed with expertise and patience.
Figure 1 shows an example of the instruments and apparatus at the J2 Engine Test Facility that are typically used in an engine test, indicating the complexity and expertise required to conduct such a test properly. At this facility, the Arnold Engineering Development Center’s test teams successfully completed testing as part of the U.S. Air Force Research Laboratory’s Adaptive Technology Development program.
Performance testing a gas-turbine aircraft engine is complicated because they come in various configurations, from turbines with single spools to turbines with three spools. It is important in every case to determine the type of engine—from a pure jet, to a fan jet, to a prop jet. In addition, the wide range of testing missions includes standard production, sea-level acceptance testing, and heavily instrumented, altitude-developmental testing. New designs require several different test cells. The test data in virtually every case must be corrected for the differences between the observed and the specified, referenced conditions. The techniques used are based upon the rules of fluid-dynamic similarity and mass and energy conservation.
The prime objective is to determine the performance of thrust and power-producing gas-turbine aircraft engines at ambient test conditions and correct these results to specified standard operating conditions.
PERFORMANCE TESTING GOALS
This article describes testing gas-turbine aircraft engines in steady state, including turbojet, turbofan, turboshaft, and turboprop engines encompassing altitude test conditions and sea-level test conditions. The test results include a myriad of issues to investigate or prove, such as the following:
- Thrust, Power, and Efficiency
- Operating Lines and Stall Margins
- Auxiliary Power Extraction
- Fuel Flow
- Specific Fuel Consumption
- Engine Airflow
- Bleed Airflow
- Vibration Levels
- Pressures and Temperatures
- Rotor Speeds
- Engine Pressure Ratio
Prior to the test, testers, operators, and evaluators should agree in writing on the object, scope, and plan. If possible, the test should be run under the specified conditions, such as thrust and/or power output, pressures, and temperatures, or as close to the specified conditions as possible to avoid applying excessive corrections afterward. Acceptable ranges for atmospheric conditions and appropriate corrections should be determined before the test, as well as appropriate correction methods, models, and formulae. Accurate steady-state engine tests typically result in uncertainties less than ±1.0%, in general, ±5 °F (±2.8 °C) for temperatures and ±0.5% for pressures. With modern data-acquisition systems, direct instrument readings are usually unnecessary. The data can be stored digitally and sampled at intervals. Where necessary, direct observations of instrument readings should be recorded at frequent intervals during testing. A digital data acquisition system capable of steady-state and transient recording is typically used during acceptance tests.
Inputs, Outputs, and Methods of Measurement Under Test
Fundamentally, to measure the power of the air-breathing aircraft engine, we must determine the mass flows of oxygen and fuel being consumed and then the power delivered by the engine, either as thrust or shaft power. Other interesting variables measured during the test include the high-pressure turbine inlet temperature of the gas turbine, the fuel-to-air ratio for the combustor, and the brake-specific fuel consumption, which is the rate of fuel consumed per unit of power. These performance measurements are based on fundamental physical and chemical equations, some of which are modified by empirical factors determined from separate tests (e.g., effects of the unique geometry of the test cell). The primary variables measured and/or computed from the results of the test are those required for input to the equations of physics and thermodynamics so the thrust, power, and efficiency can be determined.
Core Air Flow
As a preface, air is a mixture of gases. Only about 21% of air is oxygen used for combustion of the fuel, while 78% is nitrogen, water vapor (humidity), and a list of more than eight trace amounts of rare, inert gases. There are several measured air flows of interest in testing gas turbines—the mass of air consumed by the engine to produce the thrust or power (core flow), exhaust gas flow, and the amount of bleed air extracted from the compressor section, which is normally specified as a constraining condition. The difference between these flows is the amount of air available for combustion to produce the thrust or power.
To measure the power of the air-breathing aircraft engine, we must determine the mass flows of oxygen and fuel being consumed.
In a test environment, core airflow entering the engine itself is derived from a combination of test data and analysis. Direct measurement is impractical, but there are several techniques which combine the fundamental equations of flow and thermodynamics corrected by semi-empirical, legacy engineering equations to deduce that flow. These several engine-core airflow techniques are also useful in turboshaft engines, even though direct measurement of inlet airflow is accomplished with an inlet bell mouth, orifice, or Venturi (one method to validate those semi-empirical equations).
Fuel flow can be measured with a calibrated orifice or turbine meter, typically in a pipe under 2 inches in diameter. To determine heat (energy) input while operating on liquid fuel, three parametric factors must be known: (1) fuel density at test temperature (with volume measuring flow meters), (2) fuel volumetric flow, and (3) fuel heating value. The total heat input is the product of these factors.
Temperature is measured with a variety of probes to assess aerodynamic performance, cavity conditions, or material temperature (in order of accuracy, they are resistance temperature detectors, thermocouples, and pyrometers). As previously noted, there are manifold loci on the engine, thus creating interest and concern about the strength and life of material components affected by excess temperatures.
Water vapor contained in the air influences the engine and its performance. Although the consequences are complex, they fall into two major categories—engine inlet condensation and changes in gas properties. While the relative humidity is directly related to the extent of condensation on the inlet, the absolute humidity entering the inlet is the main parameter of interest. This is because the absolute or specific humidity affects the gas properties of the engine cycle (incoming air and products of combustion) and, hence, the performance. Therefore, it should be considered when requiring accurate measurements. To minimize those effects, limits on the humidity in the test cell during testing should be imposed.
Since absolute humidity does not change as the air entering from outside is static ambient, absolute ambient humidity outside the test cell can be sampled and the measurement used. This is valid so long as test conditions preclude condensation ahead of the inlet. Humidity transducers or a psychrometer may be used to measure ambient humidity.
Another operational problem is actual condensation in an engine inlet, which depends on a series of factors—relative humidity, air temperature, air pressure, inlet Mach number, and dwell or idling time. For given humidity conditions, the probability for condensation is higher in long inlet ducts and lower in bell mouth intakes.
The goal of vibration testing is to assure that the engine is free from destructive vibrations at all engine speeds, thrusts, power levels, or torque during steady and transient operations throughout the complete operating envelope of the engine. There are always engine-vibration limits that must be verified during the engine’s production, acceptance testing, design assurance, and diagnostic testing. Most of these are purely mechanical and accomplished before the engine enters the cell. However, the test cell subjects the engine to realistic aeroelastic loadings, a prime concern.
The vibration equipment may consist of on-line measuring equipment (transducers to test cell readouts) and off-line analytic equipment (spectrum analyzers). Several of the vibration sensors most often used are as follows:
a. The most common type of transducer used in aircraft engine vibration measurement is the accelerometer. Provisions for determining amplitude and frequency in three mutually perpendicular planes at appropriate locations are part of the test’s design. Accelerometers are easily mounted on the casing of the gas turbine. Since they are mounted on the casing, they pick up the vibration problems transmitted from other components. Accelerometers are more reliable than velocity sensors for higher temperatures. The accelerometer is best suited for measurements at high frequencies, such as blade passing, gear meshing, blade flutter, dry frictional whirl, surge, and gear-teeth wear.
b. Displacement probes measure shaft movement at the probe’s location. They cannot be used very successfully to measure shaft bending away from the probe’s location. The noncontacting eddy-current sensor is most effective for monitoring and measuring vibrations near rotational and subrotational speeds and is capable of measuring vibration frequencies of more than 2 kHz.
c. Velocity pickups are often used for their flat response of amplitude as a function of frequency as a go/no-go device. Average velocity amplitude is often used as an acceptance criterion because it is sensitive to many important vibration sources associated with gas-turbine aircraft engines.
When any engine exceeds the vibration limits as stated in the manufacturer’s specification, the test is stopped until the source of the vibration is determined and eliminated.
SEA-LEVEL TEST CELL
The primary function of the engine test cell is to provide a controlled environment for testing that is compatible with the engine and not hinder its operation. It is therefore necessary to conduct tests in a facility that can provide accurate and consistent measurements of performance. All test facilities have unique characteristics that will affect the testing environment and influence the data obtained. This is particularly true of indoor test cells operating at ambient conditions on the surface. Figure 2 shows one of the typical configurations for sea-level testing of gas-turbine engines .
In addition to areas denoted on the figure, there is a test-control room for the instrumentation system, data acquisition and reduction equipment, a measured fuel supply, and an auxiliary power and control system.
Configuration Fundamentals of Engine Test Cells
Test Cell Inlet. The test cell inlet improves the incoming airflow to reduce the effects of external atmospheric wind speed, direction, and extreme temperatures. This system can include flow straighteners, heaters, screens, and noise suppressors. These components create a pressure loss which must be recorded in the test report and analysis. All spaces inside the test cells are designed to produce a uniform velocity profile approaching the engine—much like the engine would experience when flying in clear air; however, this is not simple or easy.
Engine Test Section. The engine test section is the area immediately approaching the engine under test. Generally, this area will be a sufficient cross section so that the air velocity approaching the engine inlet will not exceed approximately 15 m/s (50 ft/s). In this section of a well-designed test cell, the airflow tends to have a uniform pressure distribution. Construction of the test section’s design may incorporate tapered or concave corners at the transition where the air flows into the augmenter. Interior walls and ceilings should be smooth and free from protrusions. Vortices, turbulence, and nonuniform temperatures and pressures in the area surrounding the engine can drastically affect engine performance and the test’s repeatability. Therefore, all test cells are designed to provide stable testing conditions and minimize turbulent flow by minimizing pressure losses, temperature changes, and pressure variations.
Engine Mounts. The engine mounts support the engine during testing and permit the engine’s thrust to be accurately measured. Engine thrust is usually produced at the engine’s centerline and transmitted through the mounts to a thrust frame. The thrust frame then pushes or pulls on a load cell, enabling measurement of the reaction. The most common method of engine mounting is overhead suspension. However, at some engine test facilities, the engine is mounted on a pedestal supported by the test cell floor. The overhead mount more closely simulates the mounting in many aircraft and easily accommodates cleaning the engine test section and accessing bottom-mounted engine accessories. The engine mount should be designed to prevent transverse motion, fishtailing, or any type of lateral instability and ensure that the engine’s axial alignment is maintained during testing. With turbofans, poor lateral stability caused by the mount’s flexibility can result in severe engine oscillations during testing.
Test Cell Exhaust System. In the test cell’s exhaust system, the augmenter removes the engine’s exhaust gases while inducing the flow of secondary air for cooling, providing some noise abatement. Mixing exhaust gases with the cooling secondary airflow that goes through the augmenter is then directed through an exhaust stack prior to exiting the facility. The following exhaust system features may influence the engine’s performance and must be carefully considered:
- The augmenter’s configuration (e.g., convergent or divergent)
- The augmenter’s tube length and diameter
- Exhaust inlet tube diameter
- Axial distance between engine exhaust and augmenter inlet
- Area ratio of the engine exhaust to the augmenter
- Stack cooling (air or water)
A good test cell will not allow recirculation of engine exhaust gases from exhaust stack into the cell’s inlet and must prevent the re-ingestion of exhaust gases into the engine inlet under most environmental conditions.
TURBOSHAFT AND TURBOPROP ENGINE TEST CELLS
The second class of gas-turbine aircraft engines tested is where power is delivered via a drive shaft. Therefore, determining shaft output power for turboprop and turboshaft engines is of prime interest and constitutes the main difference from the thrust-producing engines. Figure 3 shows a typical configuration of a test cell for turboprop engines .
The product of torque times speed yields the shaft power of the engine. There are two basic methods for measuring torque: (1) measuring the reaction torque of the absorption device or (2) directly measuring the shaft torque. Dynamometers typically provide controlled torque loading to turboshaft and turboprop engines during testing, as seen in Figure 3. There are several types of dynamometers commonly used for measuring the power, torque, and speed of an engine—a water brake (essentially, a very inefficient water pump), a fan dynamometer (functions like the water brake but uses air as the working fluid), and an electromagnetic absorber (a very inefficient electric generator). These essentially just waste the energy produced by the engine and produce heat with minimal flow.
A good test cell will not allow recirculation of engine exhaust gases from the exhaust stack into the cell’s inlet.
Typically, torque is set by the dynamometer’s control system while the engine’s control maintains the required speed. Typical reaction configurations include a frictionless trunnion support with a load cell or a torsion ring firmly attached to earth. The installation is designed to minimize or eliminate forces from hoses, wires, instrumentation, etc., which can bias the measurement and add to the uncertainty. Usually, shaft-torque measurement is accomplished by directly measuring the shaft torque. This is commonly done by measuring the shaft’s strain with a strain gage or by measuring its angular twist with a phase meter.
It is occasionally necessary to test the turboshaft engine on a propeller stand with its intended propeller. If the engine shaft or propeller is equipped with a torque sensor, it can be used to measure shaft power. This sensor must be calibrated using a torque arm and calibrated weights or in a dynamometer test stand prior to propeller-stand testing.
ALTITUDE TEST CELL
The third class of test cell is required to measure the engine’s performance as specified at altitudes significantly above mean sea level. To validate such performance, especially for accepting a newly designed gas-turbine engine, it must be tested on the ground in an environment simulating the required altitudes.
An altitude test cell is a vacuum pressure vessel in which a gas-turbine aircraft engine is tested at simulated, high-altitude flight conditions. The chamber is connected to a sophisticated industrial plant of air-supply compressors, temperature conditioning equipment, and exhaust compressors. Altitude test cells may also have inlet air heaters, coolers, and driers or dehumidifiers to condition the incoming air. Altitude is set by "pumping down" the chamber to the lower atmospheric static pressures for the specified altitude. The flight Mach number is set by supplying air at the proper total pressure and total temperature to the engine inlet for the specified Mach number at that altitude. A typical altitude test cell is shown in Figure 4.
The types of testing commonly conducted at simulated altitudes in an altitude test cell are for engine development, qualification, and certification .
To measure the inlet air flow to an altitude test cell, the preferred current practice is to use a manifold of sonic flow nozzles upstream of the inlet bell mouth leading to the engine inlet or use the instrumented inlet bell mouth itself to measure the flow. These two methods are interrelated because the calibrated sonic nozzles are used to calibrate the bell mouth. The total pressure of the inlet flow is controlled. Inlet flow is measured by varying the number of nozzles through which flow is allowed and controlling the pressure upstream of these nozzles.
The inlet bell mouth is quite like a large standardized flow nozzle installed in a large pipe. However, each such bell mouth is unique, and its piping configuration and installation to the engine is likewise unique. Consequently, the calibration curve for each bell mouth depends on the peculiarities of its configuration and remains valid only so long as its installed configuration remains unchanged. Since these devices are not in strict accordance with the geometric specifications and tolerances of standardized nozzles and venturis, the generic calibration curves published in Chapter 5 of ASME – Performance Test Code (PTC) 19.5  will not apply. However, once calibrated by the rules specified therein, it becomes a primary flow device. Using a pitot rake or other velocity-sensing instrumentation upstream of the first stage of the engine may also be used to measure the inlet flow.
UNCERTAINTY OF THE TEST RESULTS
Uncertainty analysis plays a very important role in testing gas-turbine engines—from designing the test to interpreting the results—because it defines the quality of the test and if the engine meets the desired performance. The smaller the overall uncertainties, the more accurate the test results. The very nature of the test will be a function of the engine’s thermodynamic cycle and the computer model employed to calculate the engine’s performance. The best engineering practice is to perform pre-test uncertainty analyses using the known or published values for the sensors intended to be used when applied to nominal, or historic, operating conditions. Then, improvements in the measuring system can be designed into the test plan. During and after the testing program, the observed uncertainties of the measured variables can be examined to see if they meet those predicted. These analyses often take much longer than the reduction of the data and computations of the engine’s performance. Several codes and specifications available defining these processes are recommended [3, 4].
Testing very expensive aircraft engines in very expensive facilities must be planned carefully and executed with expertise and patience. For those not experienced in such detailed, expert testing, the public availability of rules for such testing, codified by balanced committees of volunteer experts, is an outstanding reference for buyers, contracting officers, and young engineers. Nearly all the original equipment manufacturers and DoD agencies have written their own such test procedures. During negotiations, it is helpful and cost-efficient to have an American National Standard handy for a second opinion.
A well-constituted, standards-development committee includes engineers representing the manufacturers, government, DoD services, and consultants. Such documents should be consulted, as they are designed to be adopted for use or guidance.
- The American Society of Mechanical Engineers (ASME) Codes and Standards. Gas Turbine Aircraft Engines. ASME PTC 55, 2013.
- The American Society of Mechanical Engineers (ASME) Codes and Standards. Flow Measurement. ASME PTC 19.5, 2004.
- International Organization for Standardization (ISO). Guide to the Expression of Uncertainty in Measurement. ISO/IEC Guide 98, 1993.
- The American Society of Mechanical Engineers (ASME) Codes and Standards. Test Uncertainty. ASME PTC 19.1, 2005.