Back to the Moon, then on to Mars

Using modal data from LMS Test.Lab, researchers at the Marshall Space Flight Center created hybrid models with LMS Virtual.Lab to determine critical loading on giant rocket nozzles of NASA's nextgeneration Ares launch vehicle. As part of its responsibility for many of the modal tests within the U.S. space agency, Marshall's Structural Dynamics Test Branch also will use LMS Test. Lab for ground vibration tests of the new vehicles. In these and other projects, NASA engineers and researchers rely on LMS technologies in designing and testing advanced spacecraft that will travel the solar system in the coming decades.

If all goes according to plan, NASA astronauts will return to the moon as early as 2015 and establish a lunar base from which six-month flights to Mars would be launched by 2030. Explorers will fly to these and other extraterrestrial destinations using a next-generation Ares launch vehicle powered by the J-2X propulsion system with the world's largest rocket nozzles. System-response models for calculating critical side-wall loads on these huge nozzles have been developed in LMS Virtual.Lab based on modal data on nozzle replicas from LMS Test.Lab. Tests were conducted by the Marshall Space Flight Center (MSFC) Structural Dynamics Test Branch, which uses LMS Test.Lab to perform modal testing for the U.S. space program in a wide range of projects - including an upcoming ground vibration test (GVT) of the complete Ares vehicle.

Earthshaking Thrust. With a core propulsion stage of five RS-68 engines - the largest liquid hydrogen/liquid oxygen rocket motor in existence - the massive Ares V cargo ship will stand 360 feet tall and lift over 286,000 pounds to low Earth orbit: almost five times the shuttle payload weight. Its sister craft, the Ares I, will serve as the crew launch vehicle carrying astronauts on space missions.

The upper stages of both vehicles will be driven by a highly efficient and powerful J-2X main liquid engine designed to deliver required levels of thrust while conserving fuel on long space flights. Over 15 feet long and 10 feet in diameter, the nozzle for the J-2X will be the world's largest rocket nozzle ever made.

Risk of Large Side-Wall Loads. One of the conditions that engineers carefully consider in designing rocket nozzles, particularly large ones such as the J-2X, is called separation phenomenon. This occurs when outside ambient air is sucked into the nozzle rim by the relatively low pressures of rapidly expanding exhaust gases. This separation of exhaust gases from the side wall imparts large asymmetric transverse loads on the nozzle, deforming the shape and thus perturbing exhaust flow to cause even greater separation. The resulting fluid-structure interaction can set up 10 g resonant vibrations with several inches of side-wall deflection - enough to potentially crack the nozzle or break actuator arms that control thrust direction.

Because of these risks, side-wall loads represent an important parameter in the design of rocket nozzles. These forces are extremely difficult to measure directly, and techniques were not available for accurately predicting the magnitude and frequency of the loads until now. In developing such a method, NASA researchers studied separation phenomenon in an MSFC test cell that shoots compressed air through 5:1 scale-model rocket nozzles. The goal was to use measured vibration on these nozzle replicas to calculate the unknown force causing the vibrations. Key to this approach was the creation of a computer model accurately representing the nozzle as well as the test cell.

Building a System-Response Model. Structural dynamic models of these complex configurations would have been difficult to prepare. Therefore, unique system-response models were created based on modal test data from LMS Test.Lab, which combines multi-channel data acquisition with a suite of integrated testing, analysis and display capabilities. The modal data were used in LMS Virtual.Lab, where system-level response analysis technology was applied to construct the overall representation. Assembly definition tools specified the various connections between individual parts and a component representation manager kept track of properties for each component.

The system response model defined the overall dynamics of the structure in terms of resonant frequencies, mode shapes and damping. One of the main advantages of the method is that very few key physical points of interest require testing. The sets of modes used in creating the system response model must all be independent. For this determination, LMS PoIyMAX was used in quantifying the amplitude and frequency of each mode, and LMS Modal Assurance Criterion (MAC) 3D plots were used to check these levels so modes that most accurately represent the dynamics of the structure could be selected.

Next, LMS Test.Lab was used to acquire vibration responses from a measured input force applied by electromagnetic shakers. Measurements were then compared to predictions for checking the validity of the system response model, with discrepancies corrected by scaling factors. Nozzle side loads were then computed in an inverse force determination in which nozzle power spectral density (PSD) responses were obtained for various inputs, with amplitudes and frequencies iterated until a good comparison between measurements and prediction was achieved.