THE RESPONSE OF A POINT SOURCE IN A LIQUID LAYER OVERLYING A LIQUID HALF SPACE by SINSR TECA A "'%~~LRn ROY J. GREENFIELD S. B. Massachusetts Institute of Technology' (1958) SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE at the MASSACHUSETTS INSTITUTE OF TECHNOLOGY June, 196Z Signature of Author.. 4. --- -------. .. . . . . . . iences, May 19, i96Z Department of Earth - Certified by .. - % -- I Ther~s ~Sprvisor Accepted by ..... Chairman, Departmental Committee on Graduate Students , C~, The Response of a Point Source in a Liquid Layer Overlying a LAquid Half Space Roy J. Greenfield Submitted to the Department of Earth Sciences on June 196Z in partial fulfillment of the requirements for the degree of Master of Science ABSTRACT The response to a harmonic point source in a liquid layer overlying a liquid half space is computed as a function of frequency. Inclided are the contributions frorm all normal modes that occur, and the branch-line integral representing the refraction arrival. The value of the refraction arrival is given in terms of the complex error function. The effect of different velocity and density contrasts are considered, and the effect of source depth and the distance to the receiver are investigated. The results giving the behavior of the magnitude of the branchline show that it is much larger at the mode cutoffs than at other values of frequency. The total amplitude of the response shows a regular oscillation in the frequency range in which two modes are present, and somewhat irregular high and iow values over the range in which three modes are present. This behavior reflects the difference in amplitude at frequencies for which unodes reinforce each other. or interfere with each Thesis Advisor: Professor S. M. Simpson, Jr. Assistant Professor of Geophysics ACKNOWLEDGEMENT The author wishes to thank Professor S. M. Simpson Jr. for discussion towards formulating the thesis topic, and for his criticism during its writing. The IBM Corporation supported the research on this thesis for which the author is indebted. In particular, he wishes to thank Dr. J. P. Rossoni for his encouragement. The author also wishes to thank Mrs. S. F. Martin for the care she exercised during the typing of this work. TABLE OF CONTENTS Introduction 4 Derivation of Formulas 6 Computations Zz Results and Discussion Table I Appendix A U 31 4i Flow Chart of Program isting of Program 42 43 Appendix B Figures i-27 Bibliography 47 73 4 INTRODUCTION The problem of a point source in a liquid layer over a liquid half space was solved in terms of normal modes by Pekeris (1948). The solution, good for horizontal distances large with comparison to the depth of the layer, was expressed in terms of normal mode contributions. phase velocity. Each of the modes travels with frequency dispersed The arrivals of each mode at a given point were overlapping in time, but the frequencies for low frequency cutoff of the modes increases with mode number. By using recording systems that accentuated with low frequency part of the spectrum the arrival of the first mode could be seen on the time recording. In this paper the point of view will be to add the contributions from the several modes in the frequency domain. The results of the computation will give some idea how propagation through a layer affects the shape of the power spectrum of the recorded signal. The power spectrum of the motion is given by I g(w)f Id(t)(t, r, Z)[ Z where g(w) is the Fourier transform of the source and (w, r, Z) is the response at the observation point to a harmonic point source. 5 In his original paper, Pekeris (1948) derived an expression for the branch-line integral, representing the refraction arrival, which was valid away from the mode cutoff frequency. Officer (1953) showed that a different expression that give large contributions must be used at the cutoff frequencies. In this study an expression for the branch- line that is valid for all frequencies is derived. 6 DERIVATION OF FORMULAS In this section will be presented a brief outline of the approach used by Pekeris (1948) to solve the problem of a point source in a liquid layer over a liquid half space. The model is shown in Figure 1. b p d= depth of source Za depth of observation point C velocity in layer C = velocity in half space H= layer thickness r= horizontal distance from source to observation point 1 / p The density constrast (2. i) The problem was solved in terms of the velocity potential S(r,Z, W,t) a it (r,Z, ) (2.2) The pressure is given by Pa p at (2.3) The particle velocity is given by VI n V B (. The velocity potential obeys the wave equation. 4) The source is taken to be a simple harmonic source of the form (2z. 5) The solution must obey the boundary conditions i) at Z=O The pressure is ii) 0 at Z=H The pressure and the vertical particle velocity are continuous iii) Energy does not come in from - )Onor does the velocity become infinite as Z goes to - C . In the layer above the source the solution to this problem written in its integral form is o f with the following definitions b= p1 P2 J (Kr) is the zeroth order Bessel function F The period equation The K V XA- x4 V KL P cos H + ib K< j 1H rea. 2=) Lt"II" 's introduce branch points at KK The integral for sin iw/C l and KiK a will be evaluated by contour integration, Without proof (see Ewing et. al., 1957, p. 135) we state that all real poles are between KI and K.. just below the real axis. They may be considered to be Using the identity expressing the Bessel function in terms of the Hankel functions ZJ (Kr) a H' (Kr) + H o o (Kr) The integral for T is broken into two ingals the one containing HI is deformed upward in the K plane; the other is deformed downward as shown in Figure 2. None of the arcs contribute because of the behavior of the Hankel functions. /C 2 10 The integral along the real K dxis has been replaced by the contours shown, plus the residuais that arose by deforming the contour containing E and D. using -H C) o0 (iq) = H (2) the integral is even in branch cut. (-ig). Lines A and B cancel each other Lines D and C also cancel because and so equal on ooth si4es of the The contributions from lines E and F will be con- sidered following the discussion of the poie contributions. The residues of the poles contribute what is the normal modes contributions. of the period equation. For K known as The poles arise at the zero's < K< K. The period function may be written as tan A(, b 01.1- C2 z tan H f -9(Nj -(K) 0 1 (2.8) The contribution from each pole is given by the Cauchy residue theorem as - I VKL~ p- el R~s HH-b.9) (2. 9) 4N)is the Nth root of the period equation and is evaluated for K=eN). The Hankel function has been expanded in an asymptotic form good for large r H(2 (K (N) r) . r r.4 - K(N)r (2. 10) The total from all the poles is obtained by sumrning the contributions from each pole. T{JPOLES 7 JiVw) Ni= The number of poles that occur increases with w. The number of poles N(w) for a given w is given by the smallest integer greater than /"c L) - (/iC ) H ix + 1/2. (2. 11) There is no pole contribution if the fre uency is below the cutoff frequency of the first mode. 12 The branch line contribution is an integration path that comes in from (K , -iLN ) to KL on the left side of the branch line then returns on the right of the branch cut as shown in Figure 2. The value of is the same on both sides of the cut, but changes sign acr- ss the cut. The branch line contribution is then where L Officer (1956, p. 19o) shows 4 b inay be combined into a single integral. 1 3 IM IA/ SVH 6/rd eL kc/H 13 A new variable is introduced by the relation The asymptotic expansion of ( 10) 0. is again used for the Hankel Function. H(Kr) = H(K r ) _., 4 e -i(K r -f4) e -K_ rX (. 15) Because the Hankel function vanishes exponentially for negative inmaginary parts of K, for large r almost all the contribution to the branchline integral occurs in the vicinity of the point K=K . At this point X=0. In the change of variables only terms of the first order in X have been retained. this the denominator of F is written Cu Lc, bH -- K 1 4 ccs4l-f- A+15< 13X Doing 14 where A Bt cos (K MH) K M Lb K L( ij C ) sin K MH C The branch line integral may now be written as an integral over X. b TY ITK, ._, IC-X Cl. k I/TI2. s(KI A Kl .I(O<A 4C() 4 +-2 8 x. /4-~(X * -- <K K dX 4'?. IB) 4 b K/ - S/M K )A S In K4$ OT Pekeris (1948) evaluated the integral (2. 18) under an assumption equivalent to setting B=0. For much of the frequency range this is a reasonable procedure, because most of the contribution occurs for X small so A>> BX 15 Following from this the denominator may be approximated by A. Under this condition j=. 312 A 2 (KI Ir) (2. 19) AI-- oA(4)- U r-e (2z. 0) Officer (1953) pointed out that a different evaluation was necessary at frequencies for which A= 0. The frejuencies for which this occurs are given by These are the cutoff frejuencies for each mode. With A=0O o-Ulv o~I (~ [)j'~~-, (2. 2l) and /-;v ,) 2. cii(Kid) (e. 2) 16 At these fre uencies the magnitude of the branch line contribution decreases as I/r rather than as ir Z in the case of frequencies away from the mode cutoffs. Officer (1953) explains the significance of this as constructive addition to the refractive wave by the wave reflected in the layer. Consider a plane wave in the layer, travelling at the critical angle as in Figure 3. At D energy is put into the refraction wave The wave is reflected then travels travelling in the half space. to C where it is then reflected downward to D'. adds to the refracted wave. At D' it again The phase of the refracted wave that reaches D' will be the phase that was at D at a time L/C, earlier. For constructive interference, the wave in the layer and the refracted wave must differ in phase by ZrN at D'. The phase change over the path DCD' is given by CI CS &- + r(2.23) The change for the refracted path is 2_ TA A/- C . (2. 24) 17 Their difference is w 2-HH C, Cos -0 C1 T4+V (2.25) For the critical angle sin 0 = C / C So the condition that the phases differs by 2 w r (2z. 26) ( ct c -I - gives I C H-Q _N - - )-_ 7 "rr(N- "x- (2.27) (2. 2) Comparing with (2. 17) this is seen to be the necessary condition for A=O. The contribution from the branch line integral is large at the cutoff frequencies but falls off away from them. An ex- pression is desired that gives the value of the integral good for all values of frequency. 18 The following presents an evaluation of the branch line good for all values of A and B. The integral in question is from (2. 1c). J = (2. 29) 4- -Xjg ~k27 Q = 0 1-L (2 (2. 30) £S '" (Z. 31) Y OL - d- ;1 28 ,A --- A -+ 8 jB"~Y'Sd~ **5 (?. 32) ;8 v y ~~04 Be (L. 33) The emIuation of I follows. Consider I as a function of S. sP ' 0o -/ cLZ C +~y S2 Tf 1- I + 0 d e)y 5 (2. 34) 2_._. 19 assume a solution to (2. 34) of the form 1= e DS U(S) oij (*' d6 E-p-)t~ + 0~-Pd A(o) +M/1 77 t 2. .Q4W+ L((O) S WI Sdw+ (A(o) 2. rp U T(' ) +( (0) (2. 35) to get U(O) to get U(O) we have, using erf (0) = 0, t(O ) = (b) (Z. 36) Substituting (2. 36) in (2. 35) " C 02. (2. 37) 1. Equation (2. 37) is substituted into (2. 32) giving r='. -f I ~w(V$b)I~] (2.3b) 20 Using this evaluation of J, the general expression for the branch line integral is Ci Kt_)Y -Q 9jb K. stw(4Md) )W (k1AI) i; lSIrvyi( II (2. 39) 43J This reduces to the previous formulas 2. 19 or 2. ZI for J if B or A is 0. For A=O D=O and s= re VV S which is the result in (. 19) To compare the result for B=0O with (Z. 2I), we nust find the linit as B 4- 0. Js4- I11- -,7rTF ( PS I.-M ALVR POS-) 3 The error function is expanded in its asymptotic form good for large agreement. ~ gLV~ .. - rreL-P1. +S + + pn. 'OT I - iOs t' 1.1 24S Or .OMMW.W vFIFS "gob, k( Y~ (. 4o0) ' 21 This is the vaiue of the integrai found for B=0 and given in (2. i). The expression for the branch line given in (L. 39) is a good approxinmation at all fre juencies, and could be used in a nucerical integration to obtain the time response from the refracted wave. COMPUTATIONS COMP UTA TIONS Computations were made to obtain the vertical velocity of the liquid layer at the surface for a simple harmonic source, the response at the surface is itself a sinusoidal motion. The veritcal velocity is obtained by differentiating the veocity potential with respect to Z and letting Z=0. Previously the velocity potential was derived as the sum of the normal mode contributions (2. 9) plus the branch-line contribution. (2. 36) Although it is the magnitude of the response that is to be considered, the phase of each contribution must be used in combining the terms to find the total response ) (0). The contribution of the poles to the vertical particle velocity is given by (1. 9) as ?- :H 6-- r G -j) SN H g-s (sl (Pd) (3-1) - b(4 , ~ k,:~~c?~7_ The contribution from the branch-line is given by (2. 36) as ,-T{d)hVU ~Lc A [bf~ r ;:r A #-- , - (VP" 24 A and B are both real and positive so that D is negative imaginary. The complex quantity K rD has a phase of - r/4. In evaluating the branch line integral use was made of the Share subroutine BS WOFZ. 'The subroutine calling sequence has been changed to allow its use by FORTRAN. This subroutine computes From this the complex error function is obtained from For each frequency the contribution from the poles and from the branch-line are computed and added together to give result was printed as output. )VI . The In addition, the amplitude and phase of each pole, the total arnptade of all the poles, and the amplitude of the branch-line I bJ were printed. The computations were coded in FORTRAN and run on an IBM 7090 computer. of the program. Appendix A gives a listing and a flow chart 25 Solution of the Period Eluation Before the value of the pole contribution can be calculated, the values of K ( n ) must be obtained by solving numerically the period equation TBVAN(0 1) t- 7 > K> This form of the period equation is valid for K The graph iL given in Figure 4. is solved in terms of plotted in the graph. e I" A graph 139) suggests a method of solving given by Ewing et. al. (1957, p. this equation. K . The period equation Both sides of the period equation are Where they are equal give values of that satisfy the period equation. in terms of Smay be epressed (l, rather than in terms of K. On eliminating K .,,o I, o . r do The period equation is then expressed in terms of nN(PI): -I~iI as 26 The number of roots for a given frequency is given by () the smallest integer, greater than (N l)es ( ftK From Figure 4 it is seen that the Nth root in the interval between (n - 1 ) r and nr. (K(n)) lies With this knowledge, the equation is solved by starting with the end points of the interval enclosing the root, then dividing it in half. Evaluating the period equation at the midpoint deterrmines the half of the interval that contains the root. This opertion i is repeated 15 times so that the error in H (3 is less than The method of false position is then applied once to further decrease the error. 'The value of K is then obtained from the value of K(n) The computations have all been performed with C = 1 and Hl1. H This allows the results b be used for other values of CI and if the distance unit is taken as H and the time unit is taken as H/C . The model is then defined by giving r, d, and C2 in terms of these units. In the computations the model is defined by giving r, d, 27 C and b. As an example of how the program could be used to calculate the response in a model in which C1 or H are not one, consider the model, C1 Z km/ sec CZ 5 km/ sec d = km H = 3 km r a 100 km Use a distance unit of 3. The time unit is H C 1 2 3 sec. In these units HO I CI= 1 C d - 5/2 11/3 r = 100/3 If the response were desired at w = 10 radiani sec, the computation would be made at the value of - 3 RESULTS AND DISCUSSION 29 Branch- Lane As was expected, the magnitude of the branch-line integrated was much higher at the mode cutoff frequencies than at other points of the frequency curve. This is shown very strikingly in Figure 5. In this figure as in all the amplitude curves, the cutoff frequencies for four modes appear as vertical arrows. In Figure 6 a plot is presented for the same model, showing the behavior of the branchline contribution, near the cutoff of the third mode, using a smaller difference in frequency between points. In the derivation of the branch-line conitribution it was shown that at the mode cutoff frequencies the branch- line falls off as 1/r while away from the branch-line the frequency falls off as 1 /r Table I has the value of the branch-line contribution distances and also the ratios of the contributions. the general formulas does indeed give 1/r and 1/r I b for three This table shows 2 depe ence at the limits of frequencies close to and far away from cutoff frequencies. As a check on the computations, the magnitude of L has been computed using the formulas (2. 20) and (2. 22). This was done for the case C= 1. 5, b= .5, da .5, and Rz 50, which was plotted in Figure 6. At = 10. 53, the formulas (2. 39) used in the computed cal- culations gave / / = . 36Z7. The cutoff frequency is w = 10. 535. At 30 this frequency formulas (Z.22) gave j the rapidity at which 1IV seems satisfactory. At w = i =I.44381. Considering drops away from cutoff, this agreement . 31, a frequency far from cutoff equation (2. 20) gave . 0010596, while (2. 39) gave .001060. In formula (2. 40) the general formula was shown to go to Pekeris' formula (2. 20) away from cutoff by using a limit process. In the program this limiting process was not carefully programmed so that far from cutoff frequencies the branch-line contfribution is in error due to loss of significance in the machine computation. This is seenin Figures 5 and 6, by the uneveness of some of the values having small magnitude. is However, the magnitude of the contribution so small in these ranges that it was not -elt worthwhile to improve the program for the case of small values of the parameter B. 31 TABLE I Model C = 1.5 b= .5, (20) d= .5, (50) Rm 50 (100) (zW (100) (5o/ (100) 10.51 .6836 .Z09z .07812 8. 68 10.53 .9756 "3627 .1670 5. 17 10.55 .8936 .3141 .1360 6. 31 10.57 .6384 .1856 .06588 9, 82 10.59 .4609 .1138 .03499 13. 25 o10.61 .337b .0733 .02060 16. 56 10.71 .09628 .01612 . 004062 23. 97 10.81 .04207 .006791 .001700 24. 99 5. 0 25. 0 __________ __ I________________ Expected Ratios for I/r Expected Ratios for 0 - Cutoff frequency for the third mode is 10. 535. - . 1 32 Looking first at the modes individually we will consider The term several parts of the expression for particle velocity. G(K( n ) in equation (Z. 9) has been discussed by Pekeris (1948). He noted that for each mode this factor is 0 at its low frequency cutoff. This occurs because at the low frequency cutoff of the nth mode (n- 1l/) (n= H makes the term tan r iH in the denominator infinitely large. This behavior is seen in Figures 7, 8, and 9 where the amnpiare of each mode is 0 at cutoff; then rises rapidly as H 81 goes from (n-i ) toward nr with increasing frequency. (K(n)) was made for the first 5 In Figure 10 a plot of modes. 9 Notice in each case its cutoff value of (n--) . i( K( n ) ) initially rises rapidly from The second derivative of respect to w is negative; the slope of the increasing w. nr As H 1 curve decreases for the value of G(K(n) / The factor 1/ K( n ) causes the mode amplitude to fall off slowly as the frequency increases. V (3.1) 1Cj, (3.2) 33 This is expanded by the binomial theorem as 2 Kn)= C _+ 1to 2 W higher orde r in Therefore (3.3) \fW when w >> . The frequency where this approximation is valid modes is plotted in Figure I1. K(n)= t. (n) K(n ) is seen to go towards the line The fact that phase velocity curves go to C1 for large w reflects this. A more important factor in determining the amplitude of the mode is the factor sin d. This gives the amount that the mode is excited by a source at depth d. In Figure 9 a plot has been mnade of the amplitude versus frequency for the first six modes. depth in the case is .5. The source The amplitudes of the second and fourth modes start to fall off faster than that of the first and third modes. occurs because P 1d goes towards 2Tr This (. 5) = nr for the even modes while going toward ( 2 n+l) r (. 5) = (n + l/2) r for the odd modes, Several other plots were made to illustrate the behavior of this factor in controlling the mode amplitude. In Figure 7 the amplitude of the first mode was plotted against frequency for the three source depths d= . 25, . 5, . 75. At the low frequency part of the mode the deepest source gives the largest amplitude. As the frequency increases the source.near the bottom become less important. The curve for the source at d = . 5 gives the largest amplitude at high frequencies. The variance of the response with depth for the first two modes is examined more closely in F;igure 12. In this plot the amplitude of the mode is plotted against depth of source. This has been done for four frequencies in the range between the cutoff frequencies of the second and third mode. For both modes the behavior of the amplitude is similar at each of the four frequencies. The amplitude for the first mode varies fairly smoothly with depth and has its maximum value near d = . 65. shallow source. d. The amplitude is small for a The second mode has a small amplitude for small This amplitude reaches a maximum near .33 then falls to a value of 0 between . 6 and .7. occurs is that where FPd is equal to r. As an example, at (see Figure 10). be 0 at The depth at which the minimum = 7. 9, for the second mode is S. Then the amplitude of the second mode should d'= 5 / .63 This apparently occurs as seen in Figure 12. At depths greater than this minimum, the amplitude rises as the source is moved toward the bottom of the layer. To see if important changes occur when the velocity C2 in the lower layer is varied, the value of the amplitudes of the first two modes versus depth of source were plotted in Figure 13. By com- paring this curve for which C z 2 with Figure 12 in which C t 1.5, it is seen that the behavior of the mode amplitudes are almost identical. In this figure the mode amplitudes for three depths are plotted for a density contrast of b = .8 with C a 2. These points seem to indicate that this model has the same behavior as the models with density contrast of .5. The phase of each mode contribution is determined by i r 4 i If the first M modes are present with amplitudes Ai (w) and phase of e the square of the amplitude of their sum is given by A 1 4.II A/Ycs(-&) 1) K -VL .) (3.5) For Mr2, this formula gives for the square of the amplitude A + A2 + AIA 2 cos ( - (3.6) 2 The maximum value then taken is (A + A2 ) (3.7) occurring when the modes are in phase. The minimum is (A i - A2 ) (3. 8) which occurs when the modes are r radians out of phase. If the phase difference 0 - 92 is changing fairly uniformly with a then the difference between frequencies at which successive maximuma of the amplitude of the sum occurs is AW= 2V / d(9 - 02 dw (3.9) This alternating constructive and destructive interaction between the modes is the most striking feature of the curves computed. In the range above the low frequency cutoff of the second mode but below the cutoff of the third mode this behavior is seen to match the curves almost exactly. (see figures 14 through 24). The phase of the modes are given by e i LK )r + i 4J (3. 10) The rate of change is given by d(K( 1 r which is proportional to r. KZ)) (3. 11) dw It would then be expected that the frequency difference wald change much faster with increasing frequency. This is born out by comparing Figures 15, 17, and 18. The same model and source depth are used in all three curves. The horizontal distances from the source are 20, 50, and 100 respectively. The frequency difference between success maxima are approximately 2. 20 for distance 20, 90 for distance 50, and 45 for distance 100. These differences are roughly inversely proportional to the horizontal distance. It is noticed that the frequency difference between successive minima or maxima increases with frequency. dK(n) This occurs because as the frequency increases the value of (see Figure 11). frequency. Therefore d(K ) dK ) decreased with = I Plotted in Figure 19 is a curve for a small density contrast of d= . 8. This curve is quite similar to the curve in Figure 15. However, the maxima are slightly further apart. In Figure 20 the velocity contrast in the model is large; CZ 2. 5. Here the distance between maxima is very small when compared to Figure 17. Both of these curves are for the amplitude response at a distance 20. By a knowledge of the size of the mode amplitudes it is possible to make some surmise about the depth of source. However, the modes overlap in the frequency domain, so the magnitude of the contribution to Jw)) from the different modes cannot be measured separately. Still the low excitation of a mode does show up in a curve of amplitude of I ql(wl versus frequency. As was discussed, the amplitude of the second mode was very dependent on depth (see Figure 12). If the source depth is near the depth at which the second mode is not excited then the magnitude of t (W)does not start to show fluctuating variation in amplitude as the frequency rises above the cutoff of the second ciode. In Figure 22 the oscilla- tions of the amplitude are small at frequencies above the cutoff for the second mode. The source in this case is at depth d = . 65. At this depth the second mode is not greatly excited until well above the cutoff of the second mode, which is w = 6. 32, for the velocity of C 2 = 1. 5, As the frequency gets higher the second mode does start to contribute and the amplitude starts to oscillate. If the amplitude of the first two modes are about equal, as would occur if the source were at a depth near . 54 or . 70, then the modes would cancel each other ~am out of phase. These would then be values of w for which the amplitude of their sum would be very small. In Figure 22 a plot made for depth . 55 shows this to occur. For very shallow source depth the amplitude of the first mode is small compared to that of the second. Figure 12 shows that for a depth of . I the ratio of the second mode's amplitude to that of the first is approximately 4. For such a shallow source greatly above the curoff of the second mode. ) /I would increase The difference between successive maximum and minimums would be approximately twice the magnitude of the first mode. In Figure 24 the magnitude of has been plotted. From the computer output the first mode is seen to have amplitude of . 12. Unfortunately, a run was not made that covered the region below the second mode cutoff. However, considering other results, the curve was drawn in for the section between w = 5. 5 and w = 6. 1. The first 40 mode amplitude of. 12 is seen to be very close to half the difference between successive maxima and minima. The amplitude of the first mode changes very slowly at frequeny well above its cutoff. The first mode amplitude, then, would have about the same value below the cutoff frequency of the second mode as it has in the frequency range where the second mode is also present. Considering this a shallow source could be detected by a sudden jump from low amplitude to a much higher amplitude with fluctuations between maximum and minimurm of about twice the amplitude of the first mode. Three plots have been made for the frequency %rangeabove the cutoff frequency for the third mode. 25, 26, and 27. These are presented in Figures The three magnitude curves show an irregular pat- tern of highs and lows. There is a certain regularity between the local maxima of the curves, but the size of the maxima varies. It is probable that reflects the behavior that formula (3. 5) predicts. The maximum that could be obtained would be the sum of the amplitudes of the three modes. were in phase. This would occur if all three modes By comparing Figures 26 and 27 it is observed that the frequency difference between the maximat is smaller for the larger value of r, as was true in the frequency range where only two modes were present. 41 APPENDIX A FLOW CHART AND PROGRAM 1USTING ' 42 s TA RT AD ,CON* I * )DEL iLST Coq.puTe phnch o SN CNIIiTrib.i l S Coer e !te'rch SCor4ri 'uT'er S0 o1- oiA Co co vuTo o rT os f cF pT prRTs pNdga Ou CompTe T-t-J Pol CJTvl bT for Co' t/on"' - C!7-Te ToTl bcVa NC-h Pro computations p,.;T To t 1 ro doo MAiN PROGRAMV 0110) ,W(juO) 9Ai4I .(ju) ,v4(lU020)9,PiP(10O) , DIMEN61ON IRPOLE(100,)20),Fr-PUE(lU0,2.U), PQLEMv(2U),PLC.G(1L0U,2O),Rb(1.OU), 2FB(10O) ,XK1(lUO),BivV(lU'2Q),RP(2U),FP(2O)PULi~i(2O) I DiMEN,;,ION 6RNCri( iv I 1001 ONL=I.*Uo) N1=4 N2=2 PIE =3. 1415926 TPIE=2**PIE R EAD I NP UT FAP E NlLIOU tC I C,0 EN H 100 FORMAT(7F10.7) lol RE\D INPUT TAPt. i~lvlOlqNwqNR'NU)D FORMAT(515) XMU=.-QRTF( 1.-(C1/L2)**2) i 2 3 4 5 k( 10) vkRi<( I),.~wti<( 1 9Y( L)qEFCN( I) 9vLERO DEL.W READ INPUT TAPE. N.LIUi00(RJ)~jj=iNRud) NR=NRL) IF(NDD) 3vk,3 vii v100, (UJ) ,.J,i NUL.) READ INPUT TAP. NU=NDD DU 5 JW=1,NW Lw W(JW)W.t.RO+FLUA[F(JW-I)*U.JW=IPNW DO 9 WW=W (JW) CALL ROOT$(WWgrUENc1,C2,JWNP XNPIL) XK1 (JW)=1AW/C1 NPO=NP (JW) DO 8 JP1,tNPD CS=COSF ( bl*H) POLEG(JWJP)=B1**2* H/(DE*.'Qi-TF(XK(JWJP))) POLEG(JWJP)=PULL(JW9JP)l 5e001326 BIW (JWgJp)=B1 6 9 C04T INUE CONTINUE DO 50 JR=19NR SQR=,'QR-TF(R(JR)) DO 23 JW=19NW XK2=W (JW) /C2 NPL)NP (JW) DO 22 JP=19NPD PHASE=X(JWJP)*k(JR)-PIL/4o PStEJWJP)=MODF( PHA$E,-TPI'r')_____ RPOLE (JWtJP) =PUL&3( JW JP ) C~j. F (PHASL) /$Qwk 22 L. 6EC CONTINUE TO EVAL bRANCm A=IXKI(JW)*XMU* CUQDF(XKI(Jw)*AMU*Hi) )-k* rTF (Xi, 14+ GFAC=4**DEN*XK*AviU*,-) I. CG=GFAC*XI*EXP'4HAI*XK2*<(Jk)) S=XK1 (JWfl*R(JR) I Y=-XI*A/b (Jw(HPI E*k (VJR)-kX&I.(JW) k* SQRY=SQRTF(Y*S)*XI XX=SQRY ( 1) YY=SQRY (2) CALL CERR(XXYYo.RR1,E'RR2,,NO) ERR (1)=ERR1 ERR (2)=ERR2 E FCN=ONE-EXPF(-S*Y)*ERR FI=EFCZN-QNr'_ I QY=6QRTF (Y) I F2=QY*P E*F 1*XPF (Y*-z) IF3=(Q'QRTF(PIE/6)iF2)/ (XI*b) I BRNCH =F3*CG RB (JW) =5 RNCH ( 1.) FB (JW) =BRNCH (2) GO TO 23 CFOR 6=0 ia REL=2 *.*DEfi*XK2 )*R('JR[**Z*XMU DEiNNXK1IJW L)LNN RL3(JW)=k-EAL*C.F(Pt/2+r 2*k(JR)/ FB(JW)=-REAL*SINF(PI/2+K.*rd JR))/ DLNN 23 -'-ON TINU L D0 40 JD=J.PND NZ29112 'C.1 ,C2ZjDEN 9H t R(JR) WRITE OUTPUT TAPL 110 32 U(JU)) I 0*4,4X6HDEPTH=Fl10.4//I) WRITE OUTPUT TAPE N2 f131 FORMAT(IH F8*597X6F15*6) DO 39 JW1'sNW TRP=0. TFP=O* NPU=NP ( JW) DO032 JPitNPD DFP=,.'INF( 1W (JvwjP )*D( J0) RP (JP)I=RPOLE(JUw jP )*Dt-P FP (JP)I=FPULE (J0# JH )*UFP (JP) ; *2+FP (~)*~ POLM ( JP) =.$w RTF (RH"% TRP=TRPi-RP (JP) TFP=TFP+FP(JP) TOTP=SQRTF (TRP**, rTFP*,*2) DFt3=bINF!XK(JW)*XMU*oiJD)) TR3=RB (JW)I*DFB TFB=Fb (JW) *DFB BRMAG=SQRTF (TRb**?-+TF6**2) XMAG=6QNTF((TRP+fti~B)* ',2+(T'FP+iFb)**e-) WRITE OUTPUT TAPE. N2,119vj4,)(PLv(JP),P6E(JWJP)-JP=iNPD) TOT (K) =TOTP BR (K)= BRMAG XMG (K)=XMAG :3) CONTINUEN42'112'Ci<C2'DENqH iR(JR)qD(JD) WRITE OUTPUT TAPE N29135 OUTPUT TAPE WRITE DO 42 K1,PNW 42 WRITE OUTPUT TAPE N2,i1Uv4(s),TOT(NW3,R(KXG(K) 21,2.X6-iNU0E 3i2XbHMODE FO' MAT(9HOANG9 Fk.11X6HrAQUE ii2X6riiMQDL0 .L31 1HMODE 512X6HMOu)L 6) X5HTUTAL/) 135 FORMAT(T)H0ANGo Ft-.jiXiIHWkQTAL PU)L,-%~HBkACH 412A6 45 -40 50 CONTINUE CONTINUE GO TO 1001 FNU TOTAL 4 ft SUbROUTINERO D IIAEN6 ION XK( lUO,2O),NP(I UO) YR(I0OU 20) PE-DF(6ET)=8T /zR(iK-tL*2)+TAg4F(6ET*H-)*WEN SK1XKi**2 NP (JW) m*RTF NPO=NP (JW) DO 9 JP=IsNPD 3 i- /iL499~99 .NiT=0 RH=FL0ATF(JP)*PIL kL=RH-*49999*PI RHMAX=SQRT (SKi-z)k2) 10 RH=RHMAX-OOOO.L il CQNTINUE. DO 36 IIER=1i1it) RT=*5*(RH+RL) IF(PIERDF(RT)) 31 36 8 3O93Ov3i GO TO 36 RH=RT CONTINUt. PERL=PERD (RL) PERi-1PERDF (RH) SLJAP=(PERH-PERL) / (H-RL) TEM~4PR =Ru-PE~R L/ - LU AP YR(JWPJP)=PERDtr([ fvPR) XK ( J W PJP = OR T (~- r MPkR Rt TURN LIN L T0T AL L 2 - T SXff MIJ 9 XICINICICTV MODE L VACUu M d 14 , gIOUYCE - H C2 C2 )C, FI G URE -- - I ~ --1- K PL KN E FIGURE 2 - ---- ' 49 DIAGRAM TO EXPLAIN REFRACTION PHENOMENA C- K D' L 3 FIruRF. __ __ ____. ~_-___ _,___ GRAPH ILLUSTRATING SOLUTION OF THE PERIOD EQUATION i I FIGURE 4 __ IIU-~ -- UI 5.0 ; 1-= -,,....,,_..,----r~ + 5- - T I-T-;.-T 4"__- "! ! -4-1 --i-::. _ FIGURE -5 -Magmtude of theb-fanch- T .' _ __ T __ line versus frequ-nc 50-, .. I T il - -: ii-----I- -AV --- -:11 169-f -5 - I.r- - - __ d - i - - -. i- :J --- i_____, i ,--.---- i _i, _.. I ------ __- --- _'---J~--_2 ----I--------li+---l--------U ------ 6...... : -- ,Io 7- -- i ~---TIL^II-L~~C---L --------^-----e-ce --------------- I-~I I~ll~-C C~IICL--L( ; jL i .I I 4 I i _ _ _ __ _ _ _ _ r- _V - l-"), I 11 2 t Ii lr 9 ---1-- I - -- _ I ] -11 I-----~----TI~-'- ~--I i T- i--i-_ _ - -- -- _ - - - - - ., :-- i- - _ _ " --------~ i:_ 'i .. 2-' -- ,: ------ -I---- i ! _L-..-. "---V tz i .i. ,' Lf It- 1o~ I ' '--- sj _ -- - --_ -_ _ - _ ~I t .e i=x'iTp de; 3c; 3 -I_~ r L tt .- I--- B----- "'/C A-i- Clp C-- -- I----\- - - ~_ ---.-- ---- -I'- ---- - - --- - 4-- --- 3-~ - ---- - ---- _ ------- - - - - - - - -. i -- ___----- ___~_ __ - 4. - i- -------- -------- ________ 4. - . .,. - _ .~1 I -- -1-1-- ------- - -- -f__ , - 1 * - . -- - - 1 4-~ ~-~--I-c-. - - - - - -- - - - ,I-I-~ --------I -- ~ /C - - - - ~ I - - - - r - - - - 1----- / I" / K, I ¢ ,, ' ~ i')tA" I / --------- - 5 1' -1-- __ II __ - ! - 6 5~2____ - - : Kt. - |I - - 2 + . : "____. ' -"~--" FIGURE--64J I a gni tude-of the-branch l--ine versus-frequency , +] : 4 it4 Ii 3 ... 1'-. 2- 72 I-... 8 _ * __ ________I *~c______ 7 jfLb ----cl _ CI - 2 1' . . ; * t--------- I------------ --; I , -c-:-- -i -: I_______ -l- r- i- --- 1f-- I---- - -------..---i(.- : ------f- : ................ .... brn., - - I- - __ _ _ _ _ __ _ _ I - 1. :--,+"--- ic t++:-----(+-------!-------t-""---." --- - 4 1- 2 1,-, _~ /1 5- -: ___ " ... _ Ii _ -- ___ _ _____ _ I--- - -~ _ 4 1I- - -~ -- + .. . ~ - --- ) i_~ 5 t .-"_--' __.. . .. . I __ _ - - __ _ _ _ _ I . I .... ... -.. 'i i __ - 4 _ _ _ i _ _ _ _ _ _ :- IV'S5 I 'I- -- ---------- , -- -- /2.71 2* I . _ _ :'L 1-. .... _ _----_ I- ,c.11 - - ... i........ , +i-+ .. - _ _... . . .. . . , + i.- _______~--- . ~ c-;- , - Ur " ---- _ _ - ____ ____ ~ i-----__----'--- 1I- _ , --~-- 1 _ ________ - _ ._ .. "--"---+ . . .. ,i. .ii I 7--- 2. 7 -- - r= JL ITu (C , . ,-----------t--------- ------ I ~.- i __1--iT7 _ S Z--C: .L=* n 41~ ';'cr S~: . . 777f 6 lv") . I __- _ - -------- TH F 1 0 X 10 T Nt 4 J9 -1 2 G i ___________________________________ 1: it r T -r~i ~r + 17r T- -T I~I 77177 4. the. irest m'oae Mag itideof r-~-i T7j~7 Vr-sw§ Freguency-fofrthr~de 124b= .8, dr- .-5 -7 :-t t '74 .. .... (K - *I~. . *~1' -4--- 2. I, ill .2., , 4 7Tiz7 *~~~ I I ~45u~~i ~f7 I K 1+2 I, I ju; YO ?- VI i: ~: I '~ 1; on 10oz I , '11 Ij . Si .. . . '~~ , ' , . : ~ .. ' ," , I--.. --. / . . . .. , . . . ' ., + : .. -i. , i, . : li; 1. . ... ' + , "+ . ..... , . .' I + :i + : :: ....' i . - , , : , ;, I ' "' I • , .. . . . l, . . •' , + ,' ; , I + +.+. : ,'~ + .. . . . . . . . . . :, . ~(... ' , . . . . : j ' .. .... *+ .! ' + ++. .r . + ' . . . + ''+I : .... I + .. .. .. ' "i . ~i~ . ... : ' . , ' . .. . "//-. . ,; f 7 - ; - Le , :+ . . .+ + :':":: i.+.:t : + +I + ' :' . ' . i.. .. , + + , + .. . ... ~ i . . . .. . j " +, ; +. ; :I, . ; " . ' ;' + 'i ; + : J, : ' + :: + ::' ;+. ' ~~ . ++ i:L.. +,, . . . . . _i . :."..z , : '! .;+ i.: I .. ''' , ,+ '4, t~~-~-'1 L.'I + : I ;I -I-'' - -" .. ' :;T --. 4- __ ~ - .1. i 14)', ;: -- I,// - -1>. II, -r.. ILl 1' i- 11+: :! !I I - I ~~-"''''^- iTF :T! .1 i- '''1 i 1f 1 I i 1.4 /0,// 1 :LLI~_ ; , it I.1t ;1 ,, ~r, *+ I V., +, It I ii. - i .1,1 .- , Tl ;Ilr~ :1 i: jrti ~ tt ;4r.l 4f-ilI- -t 4-i : f.i - rw2 i- -~ -+ !-+- ' ++- iij r.- i-4 , 1-.' i ,-'r,1 , t . . I. - + +t~~ Ib If t 411- - r h7 t +r:~ ;"I~t~ --,~C I 414 2JI1~TV~T jit itt : 2-i :T!7I A tj ii i.Iij {j I tt, ~c~-~rrsn* _+ ?+LI _._ i -ilii -- ' T....... i K22 4.' i.,. II K.ii -- +, I ,+ *I +'*+''*+ I+ "- --" " ++. T,,i +;tT : I' ..-+ +., L.+. : +_ [ .. .. ... 4-;: L ''I - ,,.... 1 ::'i:iH S 40 MASS. AVE., CAMBRIDGE, MASS. +.. S.Iii.L .!:IL" ~-n-~-~ K:, .. ;+* i-* t . I I: ,~I ''r''+ + : . ~~~~' . + .: . ,' .. . ' ' I + .. . .. . . +. ° . . +f:". ' . J' . . ,I- 1I ' ,,+ , + , i ..... = +...+' *. ,l, . "I -. ! . ' .+ '+++,.+,'. ..! . '.. :,,:+ + II. ' ., + I + ' , '+• ' + ,,+. + - '' .+ , 1 , ' ,, L.C 1 , .,.++; T TT:,7 7T, T ~ ~ ~~:: . . . . . . . + .. . . :. ~~~~ ;' I. ::i+:. :: . ::.,. ' , . + . :-' I.~ . : '' , : j++ T, , TLCHiWGLOC-.Y STORE, H. C. S. .. . I.. ..: . + + i. i I: j + ." , ' -:' ' .I , , , ,:+. . . i. . .. - ., . • 'i - ', i ," ' "..] ' + ~ t j j . . , . +, . . I . I . . t~ , .1 . I . ,I - : +- I + .. - '.~ ,' i-' '++ . , I '1 , . ,,+ +., ,1,. / ."' "I 3,1! , t I *+ + . . .. .' ~ :1 •l +, : :+.t + I .... .... '1 J <::::' ++ " .• . .. : . .. : : V 7++ '7 7 >: t jeJ~ 3g/J . I. + , ~, ~ . I. . ,: + / • ... !t ' * F:F |foRM -- i, )7 ro ti CM f L AT ,Fr 35914G I 7: ~_ 2JIULL II:: 4141 V j 27~7~77I. I.... _ *' I- I. .12: :2 ttK FIGURE vs. Maknitude I--.-l ~ r fruecyor - .-. -5 medes S I S-2, r - * . -:iI KVVHT jiI ru - -.- IILI, i..... -r -~--~ ----- 1-- rr-C -c .... .. -- - c7- .-. r~ r- C I d dr=0 b V"T: ' I Elt '.2 ,'' . --1+- .... ' 1 ,' !... : ---,--.. .MO ..... .. -. .V~ H~~ i~iT1~~1+ 1±1 v/L.. i sK ...7.j..j..,{~ i I i~;f I~ 4 I rt ______ I- , TM i __ :L-~ .~I ~I I I $ KJ ~ f.4.f.i} 3 :7.lrii • r,.:~.,. " ,' . I. - ': . .. ! : j ill I - -7' f.... -: '~~~-~c~---~~~----~"~~l~'~~~~~~'"~c-~~""c'~'' ; Cl 0 b ' 1I ! : ---+I- ±fi e I ' '147; *1- 115 . .* t. .... I: 4214 ~ K L:i:7t1 '".!'i: e---( ce -- --j,-..i -~w---- +- -*I.-C-U-LY-eU-l~cl-+--Le---+LCIJI--CI.-Ci-rfCY----C---)... : L LG-I I. 4 1.-14WI.f2, ,H -I44 I~. I ,,..-f-T:7! 1- , I .~- /' ± L'LH)K ~~--L~+I ~---C-f-~*l-~*--f LI-i-L--~--CC-14-)--t~-Ct-Y1 t,~[ ,* ~-kCt-C ~31~ l 141-' 1' I i~i' 14 i i:T -4 ... ... .; '.',,,!- T ,. .. I, II ) __7 7- .. . i. t. ;l' "( - 1 *1 ,!. . . ... 1 . . . . . FU .l . .I . . :L, ... 4*, I, !::," i"I . .. 7 - ~~77 A ] ':7 7 I . 1-4 I, ;rCHliOLT,,Y STORE, H. C. S. , =f. 7 , .. . . .. ,.. I ' ' . 1 4 ,TI 14.2.4 4/. L... : ...- 1(.... I ... . i , , 1- ,,. I: I " *''1'' I I _i I - I Ir "-: 1 ': 4,1 ~. 4,W. - Vi , ,1.1 4- ..... '"-"' I ' K1K __J_ ___- _ 4-- 'I...-i--c-ciH 1.t u II .L : 2L_ J - - t •,".D .: 7 i1 j-_6__j-" .. 2 ", ! it. ' . . 2;;!i+ I." . r7T;77 40 MASS. AVE., CAMBRIDGE, MASS. ~74 IlL I ' i~i~~i*1 ~KiW h-tI~q)i!! tKT T': 'I:;.[4 1'-1.4 - , --' 4. i i - ~;U :.. ._ ;,'J *"- 7t77 c t Ti11' A'1 I 111 4 ____ ___ ______ 1 A 1. ,..A J --. K.. ------. Trequenc Sirst 2 r odes. i lor J:_t.:_ .-j. .... ..: -: fi _._s t I I 2Z. ..j::.:~:. ::--:::-- - K(N) ::.1:::..j {-_-: -first-:.7 .- _______ m I . .:., _h .. : . -I- - - : : " ' -': " :- L Ir--- .1 T 1. c- .:1 .. .... -.. . ... ..... : .. -...... t- . .-F- - ::. -: I. • , :- ,, .•__:- :.-: 7 --: : .- - _ . I4I '-. -" - -_ I . --F - . ,... . ...I:.: .... D.... .. ,3i. .. ,~~~o .. . ..- - -:.: 2-'- , " - . .. . ~h1iU~.Ji1.IUUJ 4UXLLi2 70. de : b - I - _--- . ; I , . : .... /.:; -1- : . -I{" : - !-: ...-:: : : . . - I" !- ,1-. .' - i i.. - i- --......... . .. .: .. . : . .. 7 ..: - . .!- i . 1 :.... .-:{ -:i i --i.-- - :-- -.- k-:I j .-- .. :. , -- - : - -- -. .. ....... - .. . . . .... - _ _ 1 _ - . , .. . . . . ... . . . 57 7it~+ d~wPotefo e I t of f re o~~r - - t - - Io~~- 7-7......... I K 1 .,: ft if j71 '" • ~~.. ~~ ~~ ~" ~ ... ~- *" ~":'. ~ ~... ~.. ~.. ~ ~ -; z:. .- ~ ~ ~ 711 ....... i~~~~~~~~~~~ .. ~ ~ i ,- , -~ --- . - .. .. . .. - • "' --'~-,~ . .. . . .. ...... 4~ . . . ...... .- -- --- I-"" : ....."-. "- --. ...... ~ ~ ... . "- .. . . I I ---; -*-.. -I-.... --- ,----- --.- I - -- - . :- _, -. .. ....__ __.... ... _ _I . . • • '" -I.. ... ' " " :. .--,-,----- . id ,--- , = - _ . , - . ... 5S --vs;- deptl -for .. : -I--" ! .: ::i ::.: : !' :.- M-ag-nmtude-of:=:: ::;-:::::::.-.-<-ii :: : -I::! :::: : ::: --C . .. Lo : .j -~z"... *1 -. : " "' " ._ C)~ IiO S, ..... .. ....... ... -- . -: --: 7 :- :•-:--'T --V:w 7. _1J ... with reode 1 . . . b= ==4- .. . I~ :I . - .. . ... . '.. . ... . i...... -.. ..... - .. .i.. . .. . t .. . .. .... ..... ..... . .................... _ ti - •~~~~!-"- -': i'Q H-. :. - :::'I:: -- -- I - .- 4' '- .:" .. . i." _. I - - -. .. - + ___ i "" : i I' -" . . .. " 1. I :... ". : ~ .• : . . 4. t .-. -:" ...... t-- -:---.. .. i ': -.- ... . -- - -. .. ' ...... ,_.... : ....... .. . ..:.... ..,- I .. . -- -: --... :-: -! ... •~ • :. -~~~~ :" - : ' - 1 ___ I ... .. . .. .. I......... L ... _-I 4 -.. : . ' - I- -. -....-----.. i ":-- :=br: _ . . modes ir si tw --. ____ __________ __ .~l - .. . . ~ . i:.-.:--'.-b=-"-: -5-i-"f .... 2 .... Znd . l _ . ndicate .. -. "':.L: : : : : : - 1 ".Z ...... - L. .. .. :: .--.. .., : : :: ' -1] ... _ .I i- -* - ....... .. ... .. . .. ..---. ..:.. :....:... ... :-.. ... :-:----i.. :__-_.l o.. 2-..: ./........ ...' .. - .. .- ... ! --:-:--- ---:-- -. ..-.... .. .. --:----b.. --=... ,... ... ...-:- -.- CO: ,.....: ....: . .. ,- . .. .. . .-,,. . . ....:.. . . ... :- -- - . . :. .. .. . .. ..:... _', .....:. . Illy I. oo S-I ---------- "1 ---I FORM 1 H 40 MASS. AVE, CAMBRIDGE, MASS. TECHNOLOGY STORE, H. C. S. 'ij Magitude of vs. .3 , S frequency. 1 5, b= 5, 2 4' s 5 , 25 I, I Vertical arrowv indicates mode cutoff 1 1 7 77. . .. . ' .. ?/ 7i "" 4 '- 07/ .1 'I l- " TECHNOLOGY STORE, H. C. S. FIGURE m agnitude of C2 CA__E *1OMASS. AVE.,I___ STRE H._C. S. ___ TEHOLG ___ I_ H~ FOR FORM 1 H MAS 40 MASS. AVE., CAMBRIDGE, MASS 15 Yvs, frequency 1. 5, b=. 5, r= 50, d= .5' Vertical arrow indicates mode ,', cutoff - I r, L-T,, ., ,. ', 1 ~ ... i... ' _ * ' ! \ i . I .. w3 . I . . i .. 7 " \. . *- . , . .. ' i +:+ 7 If/ / 7 4 l 7/ s 0,r 40 YASS. AVE., CAMI~RI03E, MASS. STORE, H. C. S. TECHNOLOGY TECHNOLOGY STORE, H. C. S. FORM I1 H 40 VASS. AVE., CAMBRIDGE, MASS. : -- ,. . ',,I RJ ' ';7 .. . I FIGURE 16 Magnitude of Wvs. frequency = 1.5, b=.5, r= 50, d= .75 1.C 'I ~ 4 tI -a " . i- , , 74!/,i '~i !i, . i ," ' :. ,,":i. . .: !,::, , , i! :! , , i " , '9/ J6./I , ! ,' * ., , - '7/ ,"r/ ' AS Ao' FOR FORM 2 H .40MSAECMRDE .C EHOOYSOE 2 TECHNOLOGY STORE, H. C. S. 40 MASS. AVE., CAMBRIDGE, MASS. -mcr t vs~egency, ragnituide of itI,_ II 4t M'v. ti C -] .5, b i eti. t 1ro 6 _ _ SaI _ __ 5, r ZO ,d= indicates , .,, . . .. i!q Z.bD I _ __ _ t _ ,,i.---1-- _ 5.. . . I. . . . .. I... . . . . . . . . .. . _I " t . ntiIt 144' It TI. 1 ';jT*," _.._ 71ii 71 ",,... .; ... _. 1 ,0 -5 :1 -7 ' -4 ','t .. ." ° ' - . t A-S i : .. .]. , I _ _ . I. .. . . ;, \ . . m "I .... *1.ii /1.7/ .. /0 .//. - i : FORM I H 40 MASS TECHNOLOGY STORE, H. C. S AVE, CAMBRIDGE, MASS. A3 l~~t:I.: 1'."' :; ' i I '-J .. 17 ' ' , ' ' , ' . '''.4 TTr TTT 4 FIGURE 1 t 18 t IT Magnitude o vs. frequency I -::-i7i-C = 1.5, b= 5, r= 100, d= .5 2 Vertical arrbw indicates mode cutoff . F T... .4 , -, ,'I : . rJ, ,',,, "., ".4. 3 / .71 4:: 77 .1. i j ?/J , 7 '1 to 7 , 17 71 7 ,, . /OP' .Tfl FORM 2 H TECHNOLOGY STORE, H. C. S. 40 MASS. AVE., CAMBRIDGE, MASS ,, 7.4 . -,, . ~_~~__ __I 4 4 *,,*4 4t '., :GREt :, 49 tt . . . tI I I I :1 M nitude al g Vs..'free - - -- gu e ncy ..... c t=1 5! b=. 8, r= 50, d= 5 i.'._. .. ... . ..... ............ , .. . ........ I.....4 "• , -Verticalarrowindicates mode cutoff' 4' rlYI 11 ,.I,,4 . .=,,, . ..... ... t4.4, <JI-;:'I; T;"ITT I: . ... ' ii ,, I .... . tL. ... . . _ ... ., ... 4 4." ji ' *- TT 1 iftr" -,* .. .. I ,.: l . , / -t I.: . .' . *L' ' 4,1.i4,! ,. ,, 4 - TjT . + i I . . ,. . . I : ',, , : ' " . ,- . . . .! , , .-::., .... 4 , L K.. ... .... 4 .1,T4.. ITI 1 i 74 ' 7,71r ,7t r' .31/5 s7t r ,s 7/ 7.Hi 3s S), - II-W1LLI dLU - 4. . . . ?/ /.4 I.__~_ 3/1 715" 40 MASS. AVE., CAMBRIDGE, MASS. TECHNCLOGY STORE, H. C. S. FORM 2 H 1 T-1RE 7-11 - . [.Lii 1 Magnitude of vs. freuenc !t .... .. = 2 5 b= .. , ., . O , d=.5 f. . , r =.25 , i .:, . .. . . . . .. . .. . . , i . . . . . [ .. . 11 T !_ .. .... t--4i.. .,-. -i.. .... ItI . .. ..... ,.. .-....... . . .. .. - •-. ..... . . 1+r t t.. I t. 7/ 7 . , . .,, . .. .. . . ' . I . ' ' ' ' ' - t . I, -- i ,,,I.{. ' , I 41 . ... 2. , , , I ' ! .iiI , . , . , If_ i _7 z, _- /7 L. : ?7. 6 C. €.7 , . f - . #- ,z 'z. 7) '131 . ' ,? , ,'z , , ,., .' ~~ TRH .S P~~~~~~~~.PM~~~~~~~ F,.PM 2 Ht 2 ! {, , ; :: : ., . . . . , . I .. . . l. . . . . . vs- fzreguency ________________-+ '5 AS O A_ 40 MASS. AVE., CAMBRIDGE, MASS. TECHNOLOGY STORE, H. C. S. !: :: F IG U R E . V. AS 4 ITCNLG b= .5, ri 1'00 .~j1 d~c 175 tI = I . ..5 1 4.1 jqij It . . . I >. I I 4 . .. , , I , . 1 . ; ' .. < ,. I<, . " ' ' * .., . . ' ' ' ,.. . . , + ' . . . . ., , . . . ' 9 , . .. . ' . t. . . . I. . . . , . i I l , it o 4 7.9 071 OF,$16*I .',. . ,(",... //., I,) I 1.W 2 FORM 2 H i1:~ :j Ii ~ ii 2LIft I 40 MASS. AVE., CAMSRIDGE, MASS. 't~it K II TECHNOLOGY STORE, H. C. S. II -1 I I t z --.i ;i:. Il , , 1.t l .I I '4 1' '4 LUt4.22 lit I~'II -I I it *4t4 . C 1 _: I I1: :' .4 '' I. I . = -' . - , r LI 1 !! ' 'f; 1 .. 4 . T . -1T.. 1''''' A I -'.'" * '~~1''' .LI 1,4, ii'I 4' .1..~-.- LI 4 I LI 41 ''~ I 17}77 7 -7 .. . . .' . '' ( ' I . .. .. ! Ii i ' ! +,++ , +I:!+\ '' ' . II Itt hti- 4{. i7i':'j 11 c'I" + S , . .I Il t I '+ • '1 +' Ii'4 ' + ..... j , + . . ' :-\ .. 4 :~~.. i121. E~ + j~I~1 I., ''I -~ ':1 ---- 7.30 7 170 ./ rORM 2 H ILI TECHNOCLOGY STORE, H. C. S, I' 40 MASS. AVE., CAMBRIDGE, MASS. T 12,~ Magnitude,6 off 1 s .iIiQitI _ I _ _ _ c2 T:~iiT~~4t i~~ _ _ b 1 _ reguencXy_ r=50, 11 1 4sz _ - 1 5 4 I t 00 ' I1 4' it. T1.' TI7. > :J _ _ _ _ 4o5 C~7o nvp T + -7 ~ ~ ic' ~ _ _ _ _ __ _ __ _ __ _ _ __ _ __ _ _ __ _ _ ' /4'.$ 4 -i4 r- T: 'I t r - t i. .FIGURE .24 7 -f-7Magnitude'of 44 .. ,, . . . . +* - reunv +- =':vv 4-1reguenicy -: b-..5,r 50, d= 5 Vertical arrow indicates mrxode cutoff I, 1 T. + . '-17 :7. C_ __~X I . . . . .. . .. ,+, t .. .... . .. . . . . . .14I+ ' .. + I'+ ' '+ . . . . 4 . i ' ' I'' i f~ 4 T+ .,'+-"' - ! , 141 I'IP1 ' + + * f i + - it: ' S. - : : 4, ''4 -;" , : - ':++: +I" + - : . ,I'' :' : . ,I T 4 I4 I... 4'.. , . 4-'-.. . .. ''' ' t;tl t ' . . .: :' ...+ I+ I 71 ''1 ' + tt "" ... . ... .; + . .. ''4 . . .i. ... ', j it Ii tI t i''f t '' F i iI. '. t .. , :" + . '., I : . ... . | <,, + i1 ' t4 ; :+ : '!,4. ': + ++:," t " . .. .. _ _ _ 4 ' __.. _ _ _ _ _ _ ' . ..If __'_'r *,I I ' '++ ' 4I. . . . I.. + ..'+ . . 4" 4 . . '' . ' ', ' 4 ' 4't ' . 'i. . ' .. 1 . + ' . ' , ' ,. . . ... +. + . . [. .4+,+7,1++,r ++.,+'+t. +' ,+ . ., 4 . +... + +. ' . ' ' t . + ' ,t ' ' ' li i t' 4 4i 4-i 73 Ii FIGURE Magnitude of C I k .1 'f vs. frequency 1.5, b= .5, ' I 25 , " 1 r= 100, d= .75 '" , . • . . I 2 ,A\ . .. WI ,-*. + :/ / rV ,- . .. .. 358-14 10 X 10 TO THE CM. K'UFLL • LSS CO. V: ---__-_--------^----r ----- -7---1--7- -- -~--------~*-I----- 4' lip' II~ S : - ga itw' .....e& . IiiJ. uec . -" r 100 4= .. 5 Verical arrqw-indicat b 1.~ 1 I 4 { I.. rnde c toff /9' .-.---1. I t I, I I I ~ .. I I' i . I ... .. . . . ...... I. ~. '-~1 ~1 . t. - I i I, lo , / .i/ ,7 ,4/ F1 - '/ //.P/ I// 121 -- ./ ___ill_ ___ , / .3-f ~--~----~ -~-I- - ,e7/ ,?/ , i/ 12(/ #I ,.1 .. i 11_ 0 OWN__FIFI Kiv.:17 :1 IL i7 K1FIGi -7- Magnitudeo iLd'±calar4, :4:;. T 7-t I Z7 LJRE 1* T *1 71 *I.. I, -~ .1 Ll :1 :1 -77i It'.31 1015'1 10,7 1 73 BIBIOGRAPHY Jardetzky, W. S., and Press, F, Elastic Waves in Layered Media, 1957, McGraw-Hill. 1. Ewing, W. M., 2. Officer, C. B., 'The Refraction Arrival in Water Covered Areas", Geophysics, Vol. 18, pp. 805-819, 1953. 3. Officer, C. B., Introduction to the Theory of Sound Transmission, McGraw-Hill, 1958. 4. Pekeris, C. L. "Theory of Propagation of Explosive Sound in Shallow Water found in Propagation of Sound in the Ocean, Geol. Soc. Amer. Mom. 27, 1948.