Analytical Mechanics This page intentionally left blank
Download 10.87 Mb. Pdf ko'rish
|
« 3
N u G x 3
1
2
Fig. 7.9 7.12
Precessions of a heavy gyroscope (spinning top) The system under consideration is depicted schematically in Fig. 7.9; the constraint at O is assumed to be smooth. Setting OG = ρ 0 , the gravitational potential is U (θ) = −pρ 0 cos θ. Recalling equation (7.33) and setting J 1 = J 2 = J , we arrive at the following expression for the Lagrangian: L =
1 2 J ( ˙θ 2 + ˙
ψ 2 sin 2 θ) +
1 2 J 3 ( ˙
ϕ + ˙ ψ cos θ)
2 − pρ
0 cos θ.
(7.99) This expression does not contain explicitly the two variables ϕ and ψ. From it we can deduce the two first integrals ∂L/∂ ˙ ϕ = constant, ∂L/∂ ˙ ψ = constant, i.e. ˙ ϕ + ˙ ψ cos θ = ω (0)
3 , (7.100) J ˙ψ sin 2 θ + J 3 ω (0) 3 cos θ = Λ (0)
3 . (7.101) 262 The mechanics of rigid bodies: dynamics 7.12 In addition, we have the energy integral 1 2 J ( ˙θ 2 + ˙
ψ 2 sin 2 θ) +
1 2 J 3 [ω (0) 3 ] 2 + pρ 0 cos θ = E. (7.102) Equation (7.100) expresses the fact that the gyroscopic component of the angular velocity ω is constant. This also follows directly from the third Euler equation (7.50), because of the identity J 1 = J 2 and the fact that the torque of the weight p is normal to the gyroscopic axis. Equation (7.101) expresses the conservation of the vertical component of the angular momentum L, due to the fact that the torque (G −O)×p is horizontal; to check this, compute the product σ(0)
ω · ε 3 using equations (6.35) and (6.38). The constants ω (0)
3 , Λ (0) 3 , E are to be determined using the initial conditions for ˙ ϕ, ˙
ψ, ˙ θ, θ (the initial conditions for ϕ and ψ are not essential, as the axes ξ 1
2 can be chosen so that ψ(0) = 0 and the axes x 1 , x
2 can be chosen so that ϕ(0) = 0). We exclude the case that θ(0) = 0. The system (7.100), (7.101), (7.102) can be rewritten in normal form ˙ ψ =
Λ (0)
3 − J
3 ω (0) 3 cos θ
J sin 2 θ , (7.103)
˙ ϕ = ω
(0) 3 − Λ (0)
3 −J 3 ω (0)
3 cos θ
J sin 2 θ cos θ, (7.104)
˙ θ =
± 1 J 2E − 2pρ
0 cos θ
− J 3 [ω (0) 3 ] 2 − ( Λ (0)
3 − J
3 ω (0) 3 cos θ)
2 J sin
2 θ 1/2 . (7.105) We analyse equation (7.105) under the assumption (consistent with θ(0) = / 0) J
ω (0)
3 = / Λ (0)
3 , (7.106) which implies that in equation (7.105) the expression in parentheses is always positive for the θ variable in an interval (θ , θ ), with 0 < θ < θ < π. If we exclude the trivial solutions θ = θ , θ = θ , the function θ oscillates in the interval (θ , θ ). These oscillations can be determined by integrating (7.105), with sign inversion at the endpoints. It is interesting to note that in equation (7.103) it may happen that ˙ ψ vanishes for some value of θ. In this case the precession stops momentarily, and its direction may be inverted if ˙ ψ changes sign in a certain interval. For example, if we consider θ(0) = θ 0 ∈ (0, π/2) and ˙ψ(0) = 0, we have Λ (0)
3 = J 3 ω (0) 3 cos θ
0 , and equation (7.103) becomes ˙ ψ = J 3 ω (0) 3 J sin 2 θ (cos θ 0 − cos θ). (7.107)
7.13 The mechanics of rigid bodies: dynamics 263
3
3
3
2
2
2
1
1
1
Ј
Ј
Ј
Љ
Љ
Љ (a) (b) (c) Fig. 7.10 Motion of the trace of the gyroscopic axis on a sphere: (a) momentary stop, (b) inversion of the precession, (c) precession without stops. If in addition we choose ˙ θ(0) = 0, equation (7.105) can be written ˙ θ =
± ⎧ ⎨ ⎩ 2pρ
0 J (cos θ 0 − cos θ) − J 3
(0) 3 J sin θ 2 (cos θ
0 − cos θ)
2 ⎫ ⎬ ⎭ 1/2
, (7.108)
which shows that necessarily cos θ ≤ cos θ
0 , or θ
≥ θ 0 , and hence θ = θ 0 . To fix
ideas, choose J 3 ω (0)
3 > (2pρ
0 J / cos θ 0 )
. (7.109)
We then have θ < θ < π/2. It follows that this case is characterised by a momentary halt of the precession at instants when θ reaches its minimum (corresponding to an inversion of the nutation). On the other hand, when θ takes the value θ (again the nutation is inverted), ˙ ψ takes its maximum value (verify that ˙ ψ is an increasing function of θ), see Fig. 7.10a. It is possible to choose initial conditions for which ˙ ψ changes sign, e.g. 0 < Λ (0)
3 < J 3 ω (0)
3 cos θ , see Fig. 7.10b, or such that the precession is never inverted, e.g. Λ
> J 3 ω (0)
3 > 0, see Fig. 7.10c. 7.13 Rotations We conclude our discussion of the dynamics of rigid bodies by considering briefly the case of a rigid system with a fixed axis. The physical realisation of such a constraint can be obtained by a spherical hinge (triple constraint at the point O, see Fig. 7.11) and with a collar (double constraint at the point A). 264 The mechanics of rigid bodies: dynamics 7.13
3 ≡ x 3 j 2
1
2
1 Fig. 7.11 It is convenient to fix the reference frame (O, x 1 , x 2 , x
3 ) and the body frame (O, ξ 1
2 , ξ
3 ) with the axes x 3 and ξ
3 coincident with the rotation axis. We take as the Lagrangian coordinate the angle ϕ between the axes x 1 , ξ 1 , measured counterclockwise. Hence we write ω = ˙ϕe
3 . (7.110) If the constraints are smooth, their action is expressed through a force φ O applied at O and a force φ A applied at A and orthogonal to the axis of rotation. Otherwise, we must additionally consider a friction couple µ, directed as the axis of rotation, to be specified (as an example, µ = −kω). The expression for the angular momentum in a rotation is given by L(O) = σ(O) ⎛ ⎝
0 ˙ ϕ ⎞ ⎠ =
⎛ ⎝ I 13 I 23 I 33 ⎞ ⎠ ˙ϕ, (7.111)
which reduces to L(O) =
J 3 ˙ ϕe 3 (7.112) if the axis of rotation is also a principal axis of inertia (Proposition 7.2). The projection onto the axis of rotation of the second cardinal equation σ(0) ˙ ω + ω × σ(0)ω = M(O) + (A − O) × φ A + µ, (7.113) namely
I 33 ¨ ϕ = M 3 + µ, (7.114) 7.14 The mechanics of rigid bodies: dynamics 265 can normally be integrated independently of the others. Starting from initial values ϕ, ˙ ϕ, it yields the motion of the system. The other two scalar components of equation (7.113) can be written as I 13 ¨ ϕ − I 23 ˙ ϕ 2 = M
1 (0)
− |A − O|φ A 2 , (7.115)
I 23 ¨ ϕ + I 13 ˙ ϕ 2 = M 2 (0) +
|A − O|φ A 1 , (7.116)
and yield the components of φ A . The reaction φ O can be determined using the first cardinal equation. In the simple case of uniform rotations ( ¨ ϕ = 0), equations (7.115) and (7.116) illustrate the dynamical effect of the products of inertia, which produce an additional stress to the constraint at A. In this case, the constraint must balance not just the component of M normal to the axis of rotation, as happens when I 13
23 = 0, when the axis of rotation is the principal axis of inertia. We note also that the latter case is characteristic of principal axes of inertia. Indeed, requiring that the left-hand sides of (7.115) and (7.116) vanish, one easily obtains (I 2
+ I 2 23 ) ˙ ϕ 2 = 0, (7.117)
and hence I 13 = I 23 = 0.
7.14 Problems
1. Let R be a rigid body with an axis γ with the following property: R takes the same geometric configuration after any rotation of 2π/n around γ, with n > 2 integer. Prove that: (a) γ contains the centre of mass; (b) γ is an axis of rotation of the ellipsoid of inertia with respect to any of its points.
2. Find the centre of mass of the following homogeneous systems: circular sector, circular segment, spherical cap, pyramid, cone and truncated cone, arc of ellipse, semi-ellipsoid. 3. Given a fixed reference frame, compute the tensor of inertia of the follow- ing homogeneous systems for a generic configuration: segment, general triangle, rectangle, circle, disc, regular polygon, sphere, cube. 4. Solve the problems in Examples 4.2, 4.3 of Chapter 4, replacing the two point particles P 1 , P
2 with a rigid homogeneous rod of mass m. 5. In Example 7.2 suppose that the point A is subject to a friction force given by −λ ˙x (λ > 0). Prove that the system tends asymptotically to a configuration in which the point A again takes its initial position, independently of the initial value of ˙ ϕ. Hint: take the projection along x of the first cardinal equation, and let x 0 be the x-coordinate of P 0 . Then m¨ x 0
−λ ˙x, and by integration it follows that x 266 The mechanics of rigid bodies: dynamics 7.15 must tend to zero (start by proving that ˙ x must tend to zero and deduce the asymptotic relation between x 0 and ϕ).
Write down the complete system of equations of motion. 6. Two homogeneous equal rods AB, BC, of length l and mass m, are hinged at the common endpoint B. The system lies in a vertical plane with the point A fixed and the point C moving along a fixed circle passing through A, of radius l and with centre O lying on the horizontal line through A. All constraints are smooth. Find the configurations of stable equilibrium and the normal modes of the system. 7. A heavy homogeneous circle of mass M and radius R rotates without friction around its centre O. A point particle P of mass m is constrained to slide without friction on the circle, and it is attracted with an elastic force by a point A fixed on the circle. Write down the first integrals of the cardinal equations. 8. Study the motion of precession of a gyroscope around its centre of mass O, assuming that the only torque with respect to O is due to the constraint friction and that it is proportional to the angular velocity. 9. In a vertical plane, a homogeneous equilateral triangle ABC with weight p and side l has the vertices A and B sliding without friction along a circular guide of radius l (the point C is located at the centre of the guide). A horizontal force of intensity p/ √ 3 is applied at C. Find the equilibrium configurations and the corresponding constraint reactions. Study the stability of the (two) configurations and the small oscillations around the stable one. (Remark: the first part of the problem can be solved graphically.) 10. In a vertical plane, a homogeneous rod of length l and mass m is con- strained without friction so that its endpoints lie on the parabola with equation 2ay = x
2 (y vertical, directed upwards). Study the equilibrium of the sys- tem and describe the small oscillations around the configurations of stable equilibrium. 7.15 Additional solved problems Problem 1 In a vertical plane, two rods AB, BC are hinged at the common endpoint B and the point A is constrained to be fixed. The respective lengths and masses are
1 , 2 and m 1 , m 2 . The plane on which the rods are constrained rotates with constant angular velocity ω around the vertical line through A (Fig. 7.12). Determine the configurations of relative equilibrium in the plane of the system, as well as the constraint reactions. Solution
In the rotating plane, a field of centrifugal forces normal to the axis of rotation is established, whose intensity per unit mass is ω 2 r, with r being the distance 7.15 The mechanics of rigid bodies: dynamics 267
1
2
Fig. 7.12 from the axis of rotation. We can associate with this field the potential 1 2 ω 2 r 2 . It follows that the total centrifugal potential of the system is 1 2 ω 2 m 1 1 1 0 (ξ sin ϕ) 2 dξ + 1 2 ω 2 m 2 2 2 0 ( 1 sin ϕ + ξ sin ψ) 2 dξ.
In addition to this we have the gravitational potential, yielding for the total potential U (ϕ, ψ) = 1 2 1 m 1 g cos ϕ + m 2 g 1 cos ϕ +
1 2 2 1 cos ψ
+ 1 2 ω 2 2
1 1 3 m 1 + m 2 sin
2 ϕ +
2 1 m 2 sin ϕ sin ψ + 1
2 1 2 m 2 sin 2 ψ .
Requiring to vanish the first derivatives of U , we find the equilibrium equations, which can be written in the form sin ψ = a 1 tan ϕ − b 1 sin ϕ, (7.118) sin ϕ = a 2 tan ψ
− b 2 sin ψ, (7.119) 268 The mechanics of rigid bodies: dynamics 7.15 with
a 1 = 2g ω 2 2 1 +
1 2 m 1 m 2 , b 1 = 2 1 2 1 + 1 3 m 1 m 2 , a 2 = g ω 2 1 , b 2 = 2 3 2 1 . Equations (7.118), (7.119) always admit the solutions sin ϕ = sin ψ = 0. What are the corresponding configurations? To determine the other possible solutions, we study the function β = β(α) defined by sin β = a tan α − b sin α. We have β(0) = 0 and α ∈ (−α 0
0 ), with a tan α 0 − b sin α 0 = 1, α
0 ∈ (0, π/2). Moreover β (α) cos β = a cos
2 α − b cos α, and hence β (0) = a − b. If a > b, then β (α) > 0 for α ∈ (0, α 0 ), while if a < b then β (α) < 0 in a neighbourhood of α = 0. In addition β (α) cos β = β 2 sin β + 2 a sin α cos
3 α + b sin α, and hence β > 0 for β > 0, α > 0. We can summarise this discussion in the two graphs shown in Fig. 7.13. b p 2
2
2
2
–a 0 –a 0 a 0
0
Fig. 7.13 7.15 The mechanics of rigid bodies: dynamics 269
2
p 2
2
2
a 1
1 , a 2 > b 2
1 > b 1 , a 2
2 Fig. 7.14 Graphs of the curves (7.118), (7.119). The solution of (7.118), (7.119) yields a variety of different cases. If, for example, a 1
1 (which happens for ω sufficiently large), but a 2 > b
2 (which
can be achieved by diminishing 2 ) there exists a unique non-zero solution (the symmetric solution corresponds to the same configuration). The same happens in the symmetric case, a 1 > b
1 , a
2 < b 2 (Fig. 7.14). We leave it to the reader to complete the analysis of all remaining cases in the whole variability interval ( − π, π) for ϕ and ψ. Note that adding term by term equations (7.118), (7.119) we find the second cardinal equation written with respect to the point A. To find the constraint reactions in a given configuration of relative equilibrium, we can proceed as follows: (a) write the second cardinal equation for the rod AB with respect to A, to obtain φ
B ; (b) write the first cardinal equation for the whole system (then knowledge of φ B is not necessary), to obtain φ A .
A homogeneous disc of mass M and radius R is constrained to rotate around the normal axis passing through its centre O. In addition, a point A of the axis is constrained to rotate on a fixed circle with centre O (Fig. 7.15). Determine the motion of the system starting from a generic initial condition and compute the constraint reaction at A.
270 The mechanics of rigid bodies: dynamics 7.15
3
1 ≡ N j 2
3
2
1
Download 10.87 Mb. Do'stlaringiz bilan baham: |
ma'muriyatiga murojaat qiling