Opening Remarks
Fred E. Saalfeld
Office of Naval Research
Good morning and greetings to our distinguished guests. I am pleased to represent the Office of Naval Research at this National Research Council Symposium on Large-Scale Structures in Acoustics and Electromagnetics. Large-scale structures are those whose dimensions are larger than the operational acoustic or electromagnetic wavelengths and which are complex, comprising many different systems and components with different length scales.
Large-scale structures are important for many missions of the Navy. A few examples are sonar and radar tracking and identification of ships and airplanes, microwave and optical communications, and geophysical remote sensing. The Office of Naval Research sponsors many diverse research programs that involve large-scale structures in acoustics and electromagnetics. In particular, the use of acoustics to ''see'' in the ocean is a Department of the Navy issue. In the littoral areas of the world, this problem is especially challenging technically since the acoustic reverberations are very complex. For mine location and mapping in the littoral environment, electromagnetics, especially ocean optics, holds the key to many mission requirements. Here, too, the technical challenges are impressive.
The Department of the Navy and the Office of Naval Research have undertaken support of many scientific and engineering endeavors, such as mathematics and the neural sciences, to meet these challenges. I will illustrate some of the Navy's science and technology work with two examples.
An electromagnetic example of a complex, large-scale structure is a very large scale integrated (VLSI) circuit, shown in Figure 0.1. This silicon chip is about 5 millimeters square and contains 108 photodetectors, which are seen as the horizontal line stretching across the center of the chip. Radar pulses modulate a laser beam that is then incident upon this chip. The circuit is a linear photodetector array that was developed for spectrum analysis in an environment with a high density of short radar pulses. The incident light beam wavelength is 830 nanometers, which is nearly 4 orders of magnitude smaller than the chip size. Information is extracted from the optical beam pixels at the focal plane of the photodetector array. Electromagnetic research problems arise, given that typical radar pulse widths may be less than 100 nanoseconds and a dynamic range greater than 60 dB is desired.
The electromagnetic field analysis of complex structures such as this VLSI chip can be accomplished with a combination of analytic and numerical methods. Asymptotic methods can be used depending upon the ratio of the characteristic dimensions of the complex structure to the operational wavelength. For wavelengths much smaller than the characteristic dimensions, geometric and physical optics approximations provide approximate analytic solutions. If the wavelength is much larger than the complex structure, then the Rayleigh approximation for large-scale volume scattering is valid. When the wavelength is approximately equal to the characteristic dimension, exact analytic solutions are possible only for simple, canonical geometries. Numerical integration directly from Maxwell's equations can be used for complex geometries if sufficient computational power is available.
This region of the spectrum, which is called the resonance region, is important for many applications. It is also important since the scattering characteristics are sensitive to small changes in wavelength, scattering angle, and polarization. Microscopic and quantum effects also need to be considered for yew short wavelength. We need a better understanding of electromagnetic interactions in this and other regions of the spectrum. This understanding can be obtained by the appropriate combination of exact and asymptotic analytic methods and also numerical methods.
Large-scale structures that use acoustic radiation and scattering are much larger in size and have complex infernal structures. An example is a submarine shown in Figure 0.2.
Acoustic waves that are incident on this submarine will have complex interactions with it. Inside the hull are many smaller structures and equipment. As is the case with electromagnetics, the analysis of the scattering and radiation of sound waves by such complex structures into the surrounding fluid medium or into its passenger cabin can also use asymptotically valid methods. This asymptotic analysis is possible for small or large values of the ratio of characteristic dimensions to acoustic wavelengths.
However, there is a large intermediate range in which this capability is severely limited. The difficulties involved can be illustrated by considering a simple example of an empty submarine hull submerged in water. The surrounding water supports the acoustic wavelength while the steel shell of the hull supports multiple elastic waves with wavelengths ranging from one-fifth of the acoustic wavelength to orders of magnitude larger than the acoustic wavelength. Thus the analysis of this system involves a very wide range of wavelengths that are amenable to widely differing analytic techniques. Due to the strong coupling between wave types, all interactions at different wavelengths must be treated accurately.
The comprehensive analysis of acoustic scattering by large complex structures has long had importance for the Navy. Furthermore, U.S. industry is under increasing pressure to develop analytic and modeling capabilities for complex structural acoustic systems with high strength, light weight, and low noise and vibration. The recognition of these needs has made it imperative to optimize the strength-to-weight and vibration-to-weight ratios.
Whether the interest is noise in submarines, aircraft, or automobiles, engineers and scientists are still unable to predict accurately and reliably the midfrequency acoustic behavior of large complex structures. These needs, coupled with the opportunities offered by broadening computational horizons, have renewed interest in analytical models of systems that in principle are understood, but the analysis of which is limited in practice by complexity.
Large-scale structures in acoustics and electromagnetics exemplify this trend. The VLSI chip and the submarine indicate the general characteristics of large-scale complex structures that need to be better understood:
-
First, models of these large-scale complex structures require the coherent combination of asymptotic, computational, empirical, and exact analysis.
-
Second, the analysis of large-scale systems generally is multidisciplinary in nature.
-
Third, efficiency and accuracy of the analytic techniques are essential for models of realistic large-scale systems.
At this symposium, the ONR looks forward to learning how cooperative research and development in different sciences and technologies will lead to a greater understanding of large-scale systems. I am impressed with the high quality of the abstracts and the breadth of the technical material to be presented. This symposium is indeed appropriate for the National Academy of Sciences because of the excellence of the speakers and their subjects. I am sure you will advance our knowledge of large-scale structures during your time here.
Thank you and best wishes for a successful symposium.
