J.H. Janssen 337 
Of course, many questions remain. Obviously, one of the first questions is 
whether the Eq. (32) or the Eq. (33) "model" is representative for a given under- 
water noise problem; also, how large is 6, the width of the radiating part of the 
shell when excited by machinery vibrations? This raises another question, viz., 
can a shell with frames be considered as an orthotropic plate; or what is the 
direction of 4 ? Although answers cannot be given in this paper, some experi- 
ments seem to indicate that the approach to the underwater noise reduction 
problem as presented here offers rather good qualitative and sometimes even 
quantitative insight. 
17.7. CONCLUSIGNS 
It may be concluded that for the audio-frequency range, filtered noise meas- 
urements offer an efficient method for investigating the general response of a 
mechanical structure to vibratory excitation. 
The isolation of structure-borne noise by resilient mounts is understood 
quantitatively as long as the "normal" excitation model is representative for a 
practical situation. For simple configurations like beams or plates, computation 
of direct or transfer impedances is accurately possible; results agree with 
measurements. Qualitatively, it is understood how the "parallel" stiffness of a 
spring may spoil the "normal" isolation for practical resilient mounts. Flanges 
of I-beams near springs should be stiffened. 
The idealized point-force excitation model must be expected to fail if bending 
moments can be transmitted via a contact plane; assessment of improvement 
obtained by inserting a resilient mount between a source of structure-borne 
sound and a foundation is therefore very difficult. 
For frequencies f<f,,, the simple radiation theory of flat vibrating plates 
with bending waves qualitatively describes experimental facts. It may be con- 
cluded that increasing the loss factor 7 of the shell plates by applying damping 
layers is useless if the machinery is mounted rigidly, but it may reduce the 
radiated underwater noise considerably if resilient mounts are inserted (constant 
velocity vs constant-force excitation). 
CONVERSION FACTORS* 
Length lin. 2 0.0254m 
lim = 90.8! im, 
Force llbp = 4.448N 
IN 2 0,223 Ibe 
Pressure 1 d/cm? Sos N/m? 
X db re 20 pN/m? - X—74 db re 1 d/em? 
Ydbreld/em*’ = Y+74dbre 20 pN/m? 
1 lb¢ A 
Impedance meee) 175 kg/sec 
1 lb A ‘4 
ees, EES = 45 db re 1 kg/sec 
d=dyne 
