SELF

22

S.B. Karavashkin and O.N. Karavashkina

The obtained solution (15) essentially differs from the known solutions both in amplitude and phase parts. In the amplitude part, instead an indefinite coefficient A which is known to be the basis of boundary conditions, the vibration amplitude obtained on the basis of exact analytical solutions is clearly determined relatively to the frequency omegacut.gif (838 bytes), line density rocut.gif (841 bytes) and line stiffness T. The vibration phase lags from the external force variation phase by picut.gif (836 bytes)/ 2 , which is determined by the complex unity in the right part of (15). As a result of limit transition the multiplier (2n - 1) in (12) has transformed into xo.gif (847 bytes), and the trigonometric relationship between taucut.gif (827 bytes) and betacut.gif (852 bytes) disappeared; this is non-restorable with the reverse transition from the distributed line to that lumped. Furthermore, at the limit passing the parameter betacut.gif (852 bytes) is transformed as

Due to this in a distributed line for the critical (betacut.gif (852 bytes) = 1) and aperiodical  (betacut.gif (852 bytes) > 1) vibration regimes the conditions of their existence will be invalid in the entire range from zero to infinity. So, should we try making the reverse transition from (15) to (12), using the solutions of the wave equation to which (15) naturally satisfies, it would fail, the same as we could not describe the critical and aperiodical vibration regimes. At the same time, to obtain (15) on the basis of wave equation, we would need to express the initial and boundary conditions through the parameters of external force and elastic line. However it is known that we can take as the initial and boundary conditions only numerical values of location and velocity of the picked out region of an elastic line. Consequently, despite (15) satisfies the wave equation, we cannot obtain this solution immediately from the equation. True, in some simple cases one can obtain the solutions for a wave equation having the right part [16, p. 264- 266], but with the complication of the initial conditions (for example, in inhomogeneous lines; see, e.g., [17]), or in lines with an elastically fixed end (see, e.g., [18]) these particular techniques prove to be invalid too. But seeking the solution by way of limiting process, we will always obtain the exact determined solutions.

On the basis of obtained solution, determine the regularity of line linear density rocut.gif (841 bytes)(t). Note that rocut.gif (841 bytes)  is strongly real value; it means that we cannot substitute (12) into (13) immediately. To make a substitution, represent the external force regularity as

(17)
where ficut.gif (844 bytes)0 is some initial phase. With it the solution (12) will take the form

(18)
Substituting (18) into (13), we yield

(19)
where .

We see from (19) that, though we consider in this problem a linear model of an elastic line, rocut.gif (841 bytes)(t) has non-sinusoidal, though periodical pattern. Furthermore, at F0 = T the ruptures form in the rod, they mean the density discontinuity, and this is unexpected for linear models of such class of problems that still supposed the absence of any limitations on the affecting force amplitude or limiting the amplitude by the linearity of rod stiffness T. In using the approaches based on the determinacy of exact analytical solutions, (19) shows the upper boundary of allowable load on the line equal to the line stiffness itself.

Contents: / 17 / 18 / 19 / 20 / 21 / 22 / 23 / 24 / 25 / 26 / 27 /

Hosted by uCoz