Page 44 - Winter2018

From Sputnik to SpaceX
Figure 4. The Antares launch pad at the Wallops Flight Facility, Vir- ginia. The rocket is attached to the pad by the rocket launch mount, which is the large white ring that can be seen in the middle of the figure. The (square) rocket flame trench exit is shown at right.
al., 2005; Pico et al., 2016). Preliminary work (Tsutsumi et al., 2009) has indicated the possibility of substantial noise reduction, particularly if the initial inclination of the FD is steep and it is covered. The longer the trench, the greater the noise reduction (Gely et al., 2000).
Rocket nozzle configuration and shape also impact launch noise emission (Humphrey, 1957; Viswanathan et al., 2012), and tailored nozzles can provide a reduction in the direc- tional noise by providing a low-speed layer around the out- side of the primary jet that partially blocks sound transmis- sion. This layer can be further modified by the use of wedges, pairs of vanes, and flaps (Viswanathan et al., 2012). Finally, flat concrete (reflecting) surfaces are predominant on launch pads, and recent studies have indicated that the inclusion of perforations in these surfaces is effective at reducing noise (Natarajan and Venkatakrishnan, 2016).
To control the vibration levels on launch structures, their dynamic characteristics need to be thoroughly understood, and a significant amount of recent work has focused on this (Caimi and Margasahayam, 1997; Margasahayam et al., 2002). Whereas previous pad configurations have been de- signed based on reducing liftoff peak acoustic load, Caimi and Margasahayam’s (1997) work indicates that the duration of plume impingement is a far more damaging and crucial design parameter. However, it should be noted that the feasi- bility of utilizing such modifications in practical launch pad design still remains to be determined.
42 | Acoustics Today | Winter 2018
Theoretical Work and
Scale-Model Experiments
Noise mitigation techniques have been fairly successful and in some cases decrease peak acoustic levels by up to 5 dB. However, to achieve further reductions, much greater under- standing of the mechanisms by which rocket launch noise is generated and propagated is necessary. Although the impor- tance of acoustic loading in causing structural failure has been known for 60 years (e.g., Hess et al., 1957), only relatively re- cently have significant advances in sensors, data acquisition, and processing techniques, along with huge improvements in numerical simulation ability, allowed the measurement and prediction of launch noise with any degree of accuracy.
The main issue in accurately predicting rocket launch noise is determining the relationship between the aerodynamic characteristics of the flow and the spatial characteristics of the sound field. Most rocket noise models are semiempiri- cal and based on the classic NASA SP-8072 methodology (Eldred and Jones, 1971) in which the rocket plume is the primary noise source throughout launch. A major problem with this method is that it is not consistent with Lighthill’s (1952, 1954) generally accepted jet noise theory because the initial approach to the aerodynamics was far too simplistic. Nevertheless, modified versions of the NASA model con- tinue to be used. Revisions typically focus on improving the estimate of the laminar core length (Varnier, 2001) or on amending the acoustic efficiency by a factor that significant- ly improves the fit to the experimental data while accounting for the launch pad and FD geometry as well as for shielding (Plotkin et al., 2009).
Other empirical or semiempirical methods based on histori- cal data, engineering judgment, and/or acoustic measure- ments have also been used extensively (Arenas and Mar- gasahayam, 2006; Fukuda et al., 2009). For example, data were recently collected by the Japan Aerospace Exploration Agency (JAXA) during two static-firing tests of a solid rocket motor. The data were then compared with the results of the classical NASA SP-8072 empirical prediction method and a computational fluid dynamics (CFD) calculation (Herting et al, 1971). The former overestimated the sound pressure level at certain angles from the jet axis, although the prediction at other angles was reasonable. The CFD model was effec- tive for prediction of both the near- and far-field acoustic profiles.
Once the acoustic load generated by liftoff has been pre- dicted, it is then used to predict internal vibration responses of the vehicle, its payload, and the launch pad. Due to the