The periodic rotary motions of a rigid body in a new domain of angular velocity

Introduction Some asymptotic perturbed techniques [1–3] are widely used by many authors for solving the ordinary linear and nonlinear systems for differential equations in different problems of engineering, mathematical physics, and astronomy. As an extension of this type of problem, we use some perturbation and numerical techniques in the movement of coherent bodies around a fixed point in the presence of new conditions to obtain periodic solutions of different scales to those obtained before. As in [4], let the fixed z-axis of the ellipsoid of inertia of the body makes an angle

sufficiently small instead of sufficiently large value in [4]. We define a large parameter µ proportional to 1/r o instead of the small one in [4]. Consider A, B, C represent the moments of inertia of the body, p, q, r are the components of the angular velocity vector and γ , γ ′ , γ ′′ are the direction cosines of the unit vector in direction of the Z-axis. So, the equations of motion and their three first integrals are derived in the form: Such that:

The formal construction of the periodic solutions for a rational value of the natural frequency ω
We achieve the periodic solutions (14) when: The generating system of (14) is obtained when µ → ∞ in the form: So the periodic solutions of system (17), when the period T 0 = 2π n , become: where M i , i = 1, 2, 3 are constants to be determined. Assuming the following solutions of system (14) [9]: with a period T ( µ −1 ) = T 0 + α( µ −1 ) which reduces to (18) at µ → ∞ . Let us define the quantities M i , i = 1, 2, 3 as follow: where β i are functions of µ −1 which represent the deviations of the initial values of p 2 ,ṗ 2 , γ 2 for system (14) from their initial values of generating system (17) such that β i ( 0) = 0. Let us express the initial conditions (16) by the relations: We rewrite the periodic solutions (18) in the form: where E = M 2 1 + M 2 2 and ε = tan −1 M 2 M 1 . Using (22) and (12), we get: Substituting (22) and (23) into (16), we get: where: Using (24) and (25), the following functions are obtained: Substituting by the initial conditions (21) into the first integration (17) when τ = 0 , we get: Let γ 0 ″ depends on μ −1 , we get: (26) Taking into consideration, Eqs. (27) and (28), we get M 3 , β 3 as follows: The independent conditions for periodicity are: where L 1 (ω), N 1 (ω) are obtained from L(ω), N (ω) replacing M i by (M i + β i ), i = 1, 2, 3 to get: where: For zeros approximation for power series of 1/µ , Eq. (30) give: Since the z-axis is directed along with the major or the minor axis of the ellipsoid of inertia of the body, we get: W 0 (ω) > 0 for all ω under consideration.
Assume that: Using (30), we get β 1 , β 2 in power series expansions of powers less than µ −2 . Then for the rational values of the natural frequency ω does not equal to (1, 2, 1/2, 3, 1/3, . . .) , we get the required periodic solutions and the correction of the period α(µ −1 ) as: The obtained solutions (35) and (36) are considered as the generalization of the corresponding problem in gravity field which studied in previous works [10] (when k = 0), the deviations between them are given by:

Geometric interpretation of the motion
The geometric interpretation for the motion of the body at any instant of time to Euler's angles definitions θ, ψ, φ is given by [11]: where:

Numerical solutions
In this section, we use a computer program to determine the obtained solutions (19) and their derivatives for the time in the interval t ∈ [0, 300]. On the other hand, we use the fourth-order Runge-Kutta method [12] through another program to obtain numerical solutions for the autonomous system (14). In the end, we compare both solutions to check the accuracy of the method of solutions. These results are obtained through Tables 1 and 2. From these Tables, we deduce that the numerical solutions are in agreement with the analytical ones which prove the accuracy of considered methods. (36)

Conclusions
We conclude that the problem of the motion of a rigid body about a fixed point is studied in many works [13][14][15][16][17][18] in both the uniform and gravity fields. We study our problem in case of a right angle of nutation θ 0 when its center of mass does not necessarily coincide with the fixed point. The equations of motions of the problem are obtained and reduced to a quasilinear autonomous system. The obtained system is solved by assuming a large parameter achieved from an angular velocity component tends to zero. The obtained solutions are treated through computer programs in a bounded interval of time. The autonomous system is treated with the Runge-Kutta method in the same interval of time to obtain the numerical solutions of the motion. Both obtained solutions are in full agreement with others which prove the accuracy of both numerical and analytical techniques used in solving the problem. For the geometric interpretation obtained, we note that: The precession angle ψ From (38) when µ → ∞ , we deduce that the precession angle ψ is sufficiently large because r 0 is sufficiently small, that is, we obtain a case of large precession

The nutation angle θ
We obtain a case of steady regular permutation: θ = θ 0 .

The pure rotation angle φ
The case of a large pure rotation is obtained which depends on 1/r 0 in the form: The large parameter technique used here is considered as the only one suitable for this problem in the origin domain of r o tends to zero. Poincaré-Lindstedt method or Krylov Boboliubov Mitropolski one is failed to solve this problem because they depend on achieving a small parameter in domain r 0 tends to infinity. We conduct a comparison of the results of this manuscript with the results of the previous work. The results were obtained in [19] deals with the disk problem which satisfies the symmetry moments of inertia about two principal axes of the ellipsoid of inertia but our results here treat the general rigid problem in a limiting value of the Euler's angle θ o ≈ π 2 . The advantage of our used technique [20] depends on a large parameter µ → ∞ . The obtained solutions are checked using two programs to assert their accuracy through Tables 1 and 2. The main results in our work are the obtained analytical solutions in Eq. (35) which is represented through computerized digital data in Table 1. The secondary results are proving the validity of these solutions which are given through the Runge-Kutta method in