Simulating a launch

Picture the launch of a space vehicle: The spacecraft sits on the launch pad, with its nose in the air. As you watch and listen, a rumbling noise quickly turns into a roar, an enormous cloud of smoke billows from the spacecraft, and flames shoot from the rockets as the vehicle blasts away from the launch pad and races toward the sky.

To ready a spacecraft for this moment—and ensure its components can withstand the high-intensity acoustic environment of a launch—space engineers must run countless hours of simulations on their designs. At the Korean Aerospace Research Institute (KARI; www.kari.re.kr), scientists use experimental test data to model satellites and spacecrafts, improving design and design flow while shortening schedules and reducing test costs. Key to their work are high-channel-count data-acquisition systems and advanced modeling software that combine to produce results second only to actual launch data.

KARI spearheads the South Korean National Space Program. Engineers at the institute focus on research and development of aircraft, artificial satellites, and rockets, including type certification of aircraft and quality assurance for space products. Their requirements for design and test mirror the needs of NASA and many other aerospace agencies around the world.

One goal of environmental launch testing is to ensure that space vehicles and their payloads survive the stresses of a launch. Another goal is to validate the models used to predict deployment and operational performance of payload devices, such as satellites.

During launch, sound pressure levels (SPL) in the nose cone, in fairings of launch vehicles, or in a space shuttle payload bay can approach 150 dB at frequencies ranging from 25 Hz to 10 kHz. This is 100,000 times louder than a normal conversation and loud enough to severely damage internal human organs. These high-intensity sound waves can also cause the failure of components with a large area and small mass. A high-intensity vibration can, for example, cause a solar panel to crack or a satellite to fail.

Generating such high acoustic levels in a chamber large enough to accommodate a full-size spacecraft is no small task. The reverberant chamber KARI uses for the test (Figure 1) has a volume of 1228 m³, and sound levels inside the chamber can reach 152 dB over a frequency range of 25 Hz to 10 kHz. In addition to the chamber, the test facility consists of an acoustic power-generation system driven by a high-capacity nitrogen supply, an acoustic control system, a shaker control system, and a high-channel-count dynamic-signal data-acquisition system.

To generate the high-intensity sound in the chamber, engineers use very large exponential horns (Figure 2). Valves dynamically control the flow of nitrogen gas in the horns' acoustic modulators to generate the acoustic energy. The KARI chamber uses two acoustic modulators and horns up to 2.1 m in diameter to cover a wide frequency range. The chamber's walls are shaped to minimize damping and create a diffuse high-level sound field.

Because KARI engineers must test many different launch vehicles—each with its own characteristic sound spectra—they must have precise control over the sound generation. The frequency content of the sound they produce must closely match that of the launch vehicle. Achieving a match is complicated by the frequency-dependent delay that occurs between the time the system makes a control request and the time the sound reaches the requested level.

To solve this problem, the engineers installed eight microphones in the acoustic chamber to measure the spatial sound level distribution. They then developed a scheme using a multifunction data-acquisition module to monitor the sound level and feedback command signals to the acoustic modulators. The module can measure up to 16 analog inputs and generate two analog outputs at 333 ksamples/s with 16-bit resolution.

A multithreaded LabWindows/CVI-based program controls the system. It compares the measurements with the acoustic spectrum of the specific launch vehicle and outputs a graphic display of the differences. Engineers can then choose to manually adjust the frequency content of the sound being generated or let the software do it automatically.

Because engineers use the data generated during a test to both confirm actual vibration levels on specific satellite subsystems and components as well as collect data to validate computer-aided engineering (CAE) models, the data-acquisition system must have a large number of vibration monitoring channels. The system must also have enough processing power to acquire all this data, provide interactive displays, and control the sound levels in the chamber.

Traditionally, a system such as this would have used a VME bus system and digital signal processor (DSP) boards for signal analysis. KARI, however, took a different approach and purchased a system from MTS Systems (Eden Prairie, MN; www.mtssystems.com) that uses PC and PXI technology. This system can accommodate anywhere from 8 to 5304 channels.

Currently, KARI uses a system with 192 channels, packaged in two 3U by 19-in. PXI chassis. The two chassis share the same clocks, and their sampling and triggering are synchronized through interconnected timing and control modules that reside in each chassis. This system can acquire data with less than 0.1° phase mismatch at 1 kHz between all channels, which is critical for the modal and other cross-channel analyses in these tests.

For the data acquisition itself, KARI uses modules with a resolution of 24 bits, which provides a dynamic range of more than 110 dB. These modules allow the engineers to make high-accuracy frequency-domain measurements using microphones and accelerometers placed on the spacecraft. High dynamic range is particularly important because of the high amplitudes the engineers must measure.

To analyze the data, KARI uses commercial modal-analysis software. This software saves real-time data to a PC, performs a variety of frequency response measurements (including swept sine, auto spectrum, and other frequency-domain and spectrum analyses), and post-processes prerecorded time signals. In addition to performing modal analysis, the software can also handle order tracking, acoustic intensity, fatigue evaluations, and other acquisition and post-processing applications that are required when testing satellites.

The KARI engineers use the data they acquire during their tests to improve their simulation models. Reliable test data is the key to validating model predictions and to understanding the physics so they can make meaningful design modifications based on their simulations. The problem is that test times are so short. For example, during acoustic tests, they can expose a spacecraft to the highest sound levels for only two minutes maximum or they risk damaging it. That's why their data acquisition must reliably capture data the first time. They rarely get a second chance to do it over.