Actuator Characterization Rig

Rig Description

In response to the NASA goal of developing lower emission engines, the Combustion group focused on the development of leaner-burning combustors. Unfortunately, instabilities were observed with these designs that were determined to be, in part, due to the in-phase coupling of the pressure waves generated by the actual combustion process with the geometrically-inherent standing acoustic waves of the combustor. Initially these instabilities were dealt with in a passive fashion with some success.

To augment the passive solution to the problem, the concept of Active Combustion Control was proposed. Its approach was to attempt active control of the instability through the modulation of the supplied fuel. It was thought that by controlled modulation of the supplied fuel it would be possible to influence the phasing of the generated combustion process pressure waves to sufficiently cause an attenuation of the instability.

Critical to the control premise of Active Combustion Control is the existence of an actuator that can effect the fuel modulation. The instability was observed to occur throughout a bandwidth of 1000 Hz. Actuators capable of this bandwidth with flat frequency response are currently not commonly available off-the-shelf commodities. Therefore, it becomes necessary to solicit specifications for such a device and invite potential actuator designers to supply a design that meets the specs. The follow-on to this process is to subject the candidate design to a battery of tests to determine its suitability for Active Combustion Control.

Test cell CE7A in Building 5 of the NASA Glenn Research Center serves as the home for the Actuator Characterization Rig. The rig permits the assessment of a candidate actuator with respect to its ability to deliver an AC-commanded flow for a given set mean flow. It is able to assess the frequency response characteristics over a bandwidth of interest, and it can determine the amplitude (or authority) of that response. These assessments are made while subjecting the candidate actuator to “cold flow” combustion conditions (i.e., the candidate actuator discharges its output into a back-pressured gaseous environment comparable to what would be seen in a combustor with no flame).

The rig is compromised of two racks. The first rack contains the elements that permit the physical testing of the candidate actuator. The second rack houses the system modeling and data acquisition elements of the experimental setup. For safety permit reasons, the rig is run with city water as the working fluid as opposed to fuel. It is delivered to the rig at approximately 40 psig and is boosted to a maximum working pressure of 600 psig. The fluid is not recirculated.

The major components of the first rack include a positive displacement piston pump (2000 psi max @ 2 gpm), a collection of flow conditioning devices, the candidate actuator, and a device referred to as the Simulated Combustor. The Simulated Combustor is a pressure vessel designed to provide the cold-flow combustion exit conditions for the candidate actuator. The top end cap of the vessel is designed with an interchangeable precision orifice that is meant to simulate a fuel injector of a given flow number. So in addition to providing the candidate actuator with simulated back-pressured exit conditions, it provides simulated injection conditions as well. A compressed air cylinder (separate from the 2 racks) provides the pressurized exit conditions in the Simulated Combustor.

The modular design of the “experiment” rack allows the candidate actuator to be directly connected to the top end cap of the Simulated Combustor, and it also permits the inclusion of various feed line lengths between the discharge of the actuator and the “injection point” in the top end cap. This latter experimental setup allows an examination of actuator performance degradation.

The second rack contains a PC and a DSpace computer. The PC serves a multi-purpose whereas the DSpace computer is used strictly to sample data from the experiment in real time and pass it along to the PC for storage. Initialy the PC is used to run Matlab/Simulink for generating a model of the experimental instrumentation. When this is completed, a Matlab utility is used to generate a C-code representation of the model which gets loaded into the DSpace computer. The GUI for the DSpace system, Control Desk, is then used to describe how the data for the experiment is to be sampled and where it is to be stored. It also permits the creation of an animated data display that runs while the experiment is in progress.

The instrumentation used for the data collection includes Endevco units (used for transducer excitation and filtering), Druck differential and Entran absolute pressure transducers, and a Blancett flow meter.

Past and Current Research

Glenn has procured 2 different actuator candidate concepts for its Active Combustion Control program. The 2 candidate concepts were from Georgia Tech and Fluid Jet Associates, respectively. Succumbing to program deadlines, a determination was made that assessed the Georgia Tech concept as being “more mature” and being a more complete actuation system. For this reason, a focused effort was placed on its performance evaluation via the characterization rig.

The Georgia Tech actuator was thoroughly evaluated for a condition where it would be flowing fuel at a flow number of 110. The testing showed it as a suitable candidate for use with Active Combustion Control control algorithms. The actuator was used to supply fuel to a single nozzle combustion rig that was designed to exhibit the instabilities observed in a “real world” aircraft engine. The Georgia Tech actuator was successful in delivering fuel to this rig (as commanded) and attenuation of the combustion instability was observed. This success allowed the achievement of a Congressional Level 1 Milestone.

Since the completion of that work, a request was made to use the same actuator on the NASA Glenn combustion rig located in CE5. The purpose for this testing was to observe the effect of pulsing pilot fuel. Pilot fuel flow in this case occurs for a flow number of around 8. Therefore the Georgia Tech actuator was evaluated in the characterization rig for the different flow number condition. There was enough experimental evidence to indicate that it could be used for the pilot flow research. The results observed in CE5, however, indicated very limited authority on the Georgia Tech actuator’s part.

Further work is being performed on the Georgia Tech actuator within the characterization rig to determine how to better “operate” it for low flow numbers. It is anticipated that it will once again be used in CE5 for research being done in conjunction with GE. Additional characterization studies are still envisioned for the 2nd (less mature) actuation concept that was procured from Fluid Jet Associates.

Contact

Facilities Index

responsible official: sanjay garg
site maintenance: edmond wong
nasa privacy statement
last updated: 2.29.08