NivaarExam PrepOfficial exam papers ↗

22-Mec-A2 Kinematics and Dynamics of Machines · December 2016

Question 6 of 7: Double pendulum with a stopper (impact)

Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)

Notes on this paper

Paper: National Exams — December 2016 · 07-Mec-A2 Kinematics and Dynamics of Machines · 3 hours, open book · Answer five of seven questions, at least one from Part B. Marks: 20 each. All seven questions are solved here.

Reference texts (subject). R. L. Norton, Design of Machinery, 6th ed. (mechanisms, cams, gear trains); C. E. Wilson & J. P. Sadler, Kinematics and Dynamics of Machinery, 3rd ed.; Uicker, Pennock & Shigley, Theory of Machines and Mechanisms, 5th ed.; S. S. Rao, Mechanical Vibrations, 6th ed. (Part B). Balancing follows Norton Ch. 13; planetary trains Norton §9.7–9.9.

Question 6: Double pendulum with a stopper (impact) (20 marks)

Question text not reproduced: the examination questions are © Engineers and Geoscientists BC. Open the official past paper (linked at the top of this page) to read the question, then follow the worked solution below.

Given. Two uniform rods (each $m$, $l$); rod 1 pinned to ground at $O$, rod 2 pinned to the end of rod 1. A rigid stopper on the equilibrium vertical contacts rod 1 at $0.9l$ from $O$ when rod 1 returns to vertical ($\theta_1=0$). Restitution $e$, impact duration $\Delta\tau$. Small initial angles $\theta_{1,0},\theta_{2,0}$, released from rest.

Find. The free-vibration response of the system in the interval between the first and second stopper impacts (i.e. the motion "just prior to the second impact").

stopper0.9lg ↓llO
Double pendulum: absolute angles $\theta_1,\theta_2$ from the downward vertical. The stopper at $0.9l$ arrests rod 1 as it swings back through $\theta_1=0$; each contact is an impulsive event with restitution $e$.

Approach. Build the linearised 2-DOF equations, get the modes, propagate the free response from the release, apply the impulsive restitution law at $\theta_1=0$ to jump the velocities, and propagate again to the next contact.

Step 1 — Linearised equations of motion

Using absolute angles $\theta_1,\theta_2$, the kinetic and potential energies of two uniform rods expand (small angles) to a mass and stiffness matrix

$$\mathbf M=ml^2\!\begin{bmatrix}\tfrac43 & \tfrac12\\[2pt]\tfrac12 & \tfrac13\end{bmatrix},\qquad \mathbf K=mgl\!\begin{bmatrix}\tfrac32 & 0\\[2pt] 0 & \tfrac12\end{bmatrix},\qquad \mathbf M\ddot{\boldsymbol\theta}+\mathbf K\boldsymbol\theta=\mathbf 0 .$$

Step 2 — Natural frequencies and modes

Setting $\det(\mathbf K-\omega^2\mathbf M)=0$ with $\lambda=\omega^2 l/g$ gives $7\lambda^2-42\lambda+27=0$, so $\lambda_{1,2}=0.732,\ 5.268$ and

$$\boxed{\omega_1=0.856\sqrt{g/l},\qquad \omega_2=2.295\sqrt{g/l}.}$$

The mode-shape ratios $\theta_2/\theta_1$ follow from $(\tfrac32-\tfrac43\lambda)\theta_1=\tfrac12\lambda\,\theta_2$:

$$\boldsymbol\phi_1=\begin{Bmatrix}1\\ 1.43\end{Bmatrix}\text{(in phase)},\qquad \boldsymbol\phi_2=\begin{Bmatrix}1\\ -2.10\end{Bmatrix}\text{(out of phase)}.$$

Step 3 — Free response before the first impact

Released from rest at $\boldsymbol\theta(0)=\{\theta_{1,0},\theta_{2,0}\}$, the modal expansion $\boldsymbol\theta=\eta_1(t)\boldsymbol\phi_1+\eta_2(t)\boldsymbol\phi_2$ with $\dot{\boldsymbol\theta}(0)=0$ gives

$$\boldsymbol\theta(t)=c_1\cos(\omega_1 t)\,\boldsymbol\phi_1+c_2\cos(\omega_2 t)\,\boldsymbol\phi_2,$$

where $c_1,c_2$ solve $c_1\boldsymbol\phi_1+c_2\boldsymbol\phi_2=\{\theta_{1,0},\theta_{2,0}\}$, i.e. $c_1=\dfrac{2.10\,\theta_{1,0}+\theta_{2,0}}{3.53},\ c_2=\dfrac{1.43\,\theta_{1,0}-\theta_{2,0}}{3.53}$. The first impact occurs at the earliest time $t_1$ with $\theta_1(t_1)=0$ (rod 1 back to vertical); the pre-impact velocities are $\dot\theta_1^-=\dot\theta_1(t_1),\ \dot\theta_2^-=\dot\theta_2(t_1)$.

Step 4 — Impulsive restitution at the stopper

At contact the stopper applies a horizontal impulse $\hat P$ to rod 1 at $0.9l$; positions are frozen ($\theta_1=0$) and only velocities jump. In generalised coordinates the impulse appears as $\hat Q_1=0.9l\,\hat P,\ \hat Q_2=0$ (the contact point lies on rod 1, independent of $\theta_2$), so $\mathbf M\,\Delta\dot{\boldsymbol\theta}=\{0.9l\hat P,\,0\}$. The second row gives the internal constraint

$$\tfrac12\Delta\dot\theta_1+\tfrac13\Delta\dot\theta_2=0\;\Rightarrow\;\Delta\dot\theta_2=-\tfrac32\Delta\dot\theta_1 .$$

Restitution at the contact point (horizontal speed of the $0.9l$ point reverses and scales by $e$): $0.9l\,\dot\theta_1^+=-e\,(0.9l\,\dot\theta_1^-)$, hence $\dot\theta_1^+=-e\,\dot\theta_1^-$ and $\Delta\dot\theta_1=-(1+e)\dot\theta_1^-$. Therefore

$$\boxed{\dot\theta_1^+=-e\,\dot\theta_1^-,\qquad \dot\theta_2^+=\dot\theta_2^-+\tfrac32(1+e)\,\dot\theta_1^-.}$$

(The finite duration $\Delta\tau$ enters only through the mean contact force $\bar P=\hat P/\Delta\tau$; it does not affect the impulsive velocity jump.)

Step 5 — Free vibration up to the second impact

Restarting the clock at the first impact ($\tau=t-t_1$), the state is $\theta_1=0,\ \theta_2=\theta_2(t_1)$, with velocities $\dot\theta_1^+,\dot\theta_2^+$ from Step 4. Projecting onto the modes,

$$\boldsymbol\theta(\tau)=\sum_{i=1,2}\Bigl[A_i\cos(\omega_i\tau)+B_i\sin(\omega_i\tau)\Bigr]\boldsymbol\phi_i,$$

with $A_i$ set by the post-impact position $\{0,\theta_2(t_1)\}$ and $B_i=\dfrac{\text{(post-impact modal velocity)}_i}{\omega_i}$ set by $\{\dot\theta_1^+,\dot\theta_2^+\}$ (modal coordinates obtained from the same $\boldsymbol\phi_1,\boldsymbol\phi_2$ inverse used in Step 3). This expression is the free vibration during the interval; the second impact is the next root $\tau_2>0$ of $\theta_1(\tau)=0$, and "just prior to the second impact" means evaluating the boxed response at $\tau\to\tau_2^-$.

Check (closed form of the impact times): because $\omega_1$ and $\omega_2$ are incommensurate, the impact instants $t_1,\tau_2$ are roots of a transcendental equation $\theta_1(t)=0$ and have no elementary closed form; they are found numerically once $\theta_{1,0},\theta_{2,0},e$ are given. The framework above (modes, modal fit, restitution jump) is exact and fully determines the motion between impacts for any data.

QuantityResult
Natural frequencies$\omega_1=0.856\sqrt{g/l}$, $\omega_2=2.295\sqrt{g/l}$
Mode shapes $\{\theta_1:\theta_2\}$$\{1:1.43\}$, $\{1:-2.10\}$
Impact velocity jump$\dot\theta_1^+=-e\,\dot\theta_1^-$, $\ \dot\theta_2^+=\dot\theta_2^-+\tfrac32(1+e)\dot\theta_1^-$
Motion before 2nd impact$\displaystyle\sum_i[A_i\cos\omega_i\tau+B_i\sin\omega_i\tau]\boldsymbol\phi_i$, evaluated at $\tau\to\tau_2^-$