Question 3 of 6: Rotating shaft with an overhung load — fatigue diameter and deflections (20 marks)
Nivaar worked solution (AI-drafted; not reviewed by a licensed engineer)
Notes on this paper
Paper format. National Exams, December 2018 — 07-Mec-B1 Advanced Machine Design. Three hours, open book, 100 marks. Part I (Problems 1 and 2) is compulsory; the candidate answers three of the four Part II problems (Problems 3–6). Any non-communicating calculator is permitted, assumptions must be stated, and tabulated values must be sourced. All six problems are worked here, so the set is complete as a study resource.
Reference texts. Budynas & Nisbett, Shigley's Mechanical Engineering Design, 11th ed. (shaft fatigue §7, power screws §8-2, clutches and brakes §16, journal bearings §12); Norton, Machine Design: An Integrated Approach, 6th ed. (impact loading §3, brakes §16); Juvinall & Marshek, Fundamentals of Machine Component Design, 6th ed. (screws, clutches, brakes); Hibbeler, Mechanics of Materials, 10th ed. (beam bending, impact factors); CSA/ISO 14006 Eco-design management and ISO 14040 Life-cycle assessment for the green-design criteria.
Question 3: Rotating shaft with an overhung load — fatigue diameter and deflections (20 marks)
Given. A rotating steel shaft on two self-aligning bearings, loaded transversely beyond the right-hand bearing and carrying a pulsating torque; the figure gives the three stations along the shaft.
Given data — Problem 3
Quantity
Symbol
Value
Bearing A (left) at
$x_A$
0
Bearing B (right) at
$x_B$
14 in
Transverse load station
$x_P$
18 in
Shaft overall length
$L_{tot}$
20 in
Transverse load (constant magnitude)
$P$
1000 lb
Torque range
$T$
0 to 2000 lb·in
Ultimate / yield strength
$S_{ut}$ / $S_y$
108 ksi / 62 ksi
Required fatigue safety factor
$n$
2
Stress concentration
$K_f = K_{fs}$
1 (stated: none)
Find. (1) the shaft diameter that gives a fatigue safety factor of 2, and (2) the corresponding maximum bending deflection and maximum angle of twist.
Figure 3.1 — Bearings A (0) and B (14 in) act as simple supports; the 1000 lb load sits 4 in outboard of B, so the bending moment peaks at bearing B and is fully reversed as the shaft turns.
Approach. Recognise that the load overhangs bearing B, so the bending moment peaks at the bearing; because the shaft rotates under a constant-direction load, that bending is fully reversed while the torque pulses about a non-zero mean. Size the shaft with the DE-Goodman shaft equation, iterating on the size factor, then compute the overhang deflection and the twist at the selected standard diameter.
Support reactions. Bearings at $x = 0$ and $x = 14$ in are simple supports; the 1000 lb load sits 4 in outboard of the right bearing.
$$R_B = \frac{P\,x_P}{x_B} = \frac{1000(18)}{14} = 1286\text{ lb}\ (\uparrow), \qquad R_A = P - R_B = -286\text{ lb}\ (\downarrow)$$
The left bearing carries a downward (hold-down) reaction — typical of an overhung layout and the reason the left bearing must be retained axially and radially in both directions.
Critical bending moment. Beyond bearing B the moment builds linearly from zero at the load; between the bearings it decays linearly to zero at A. The maximum is therefore at bearing B, where the overhang is $a = 18-14 = 4$ in.
$$M_{\max} = P\,a = 1000 \times 4 = 4000\text{ lb}\!\cdot\!\text{in}$$
Resolve the loading into mean and alternating components. The transverse load keeps a constant magnitude and direction while the shaft turns, so every surface fibre sees tension and compression once per revolution: the bending is fully reversed. The torque, by contrast, pulses between 0 and 2000 lb·in about a positive mean.
$$M_a = 4000\text{ lb}\!\cdot\!\text{in}, \qquad M_m = 0$$
$$T_a = \frac{T_{\max}-T_{\min}}{2} = 1000\text{ lb}\!\cdot\!\text{in}, \qquad T_m = \frac{T_{\max}+T_{\min}}{2} = 1000\text{ lb}\!\cdot\!\text{in}$$
Endurance limit with Marin factors. For steel with $S_{ut} = 108\text{ ksi} \lt 200$ ksi the rotating-beam limit is $S'_e = 0.5S_{ut} = 54$ ksi. Take a machined surface, room temperature, and 50 % reliability as the baseline of the shaft equation:
$$k_a = a\,S_{ut}^{\,b} = 2.70\,(108)^{-0.265} = 0.781, \qquad k_b = 0.879\,d^{-0.107}\ \ (0.3 \le d \le 2\text{ in})$$
$$S_e = k_a k_b S'_e$$
Because $k_b$ depends on the diameter being sought, the sizing is iterative.
Assemble the von Mises moment groups. The DE (distortion-energy) shaft equation combines bending and torsion into equivalent alternating and midrange terms:
$$\sqrt{4(K_f M_a)^2 + 3(K_{fs}T_a)^2} = \sqrt{4(4000)^2+3(1000)^2} = 8185\text{ lb}\!\cdot\!\text{in}$$
$$\sqrt{4(K_f M_m)^2 + 3(K_{fs}T_m)^2} = \sqrt{0+3(1000)^2} = 1732\text{ lb}\!\cdot\!\text{in}$$
DE-Goodman diameter, iterated on the size factor. Shigley's first-cycle-free design form is
$$d = \left\{\frac{16n}{\pi}\left[\frac{\sqrt{4(K_fM_a)^2+3(K_{fs}T_a)^2}}{S_e} + \frac{\sqrt{4(K_fM_m)^2+3(K_{fs}T_m)^2}}{S_{ut}}\right]\right\}^{1/3}$$
Starting from a trial $d = 1.00$ in and re-evaluating $k_b$ each pass, the iteration converges in a few cycles to
$$\boxed{\;d_{req} = 1.356\text{ in}\;}$$
Select a standard size and confirm the factor. Round up to the next standard fractional shaft diameter, $d = 1\tfrac{3}{8}$ in $= 1.375$ in. At that size $k_b = 0.879(1.375)^{-0.107} = 0.847$, $S_e = 35.7$ ksi, and back-substitution gives
$$\boxed{\;d = 1\tfrac{3}{8}\text{ in, } n_f = 2.09 \;\gt\; 2 \;\checkmark\;}$$
First-cycle yield check. Fatigue sizing must be backed by a static check at the peak of the load cycle. At $d = 1.375$ in,
$$\sigma_b = \frac{32M_{\max}}{\pi d^3} = 15\,674\text{ psi}, \qquad \tau = \frac{16T_{\max}}{\pi d^3} = 3919\text{ psi}$$
$$\sigma' = \sqrt{\sigma_b^2 + 3\tau^2} = 17\,067\text{ psi} \;\Longrightarrow\; n_y = \frac{62\,000}{17\,067} = 3.63$$
Yielding is not the governing mode, as expected for a fatigue-driven design.
Bending deflection at the overhung load. For a simply supported beam of span $L$ with a load $P$ on an overhang $a$, the tip deflection is $\delta = Pa^2(L+a)/(3EI)$. With $E = 30\times10^6$ psi and $I = \pi d^4/64 = 0.1755\text{ in}^4$,
$$\delta = \frac{1000(4)^2(14+4)}{3(30\times10^6)(0.1755)}$$
$$\boxed{\;\delta_{bend} = 0.0182\text{ in} = 0.463\text{ mm (downward, at the load)}\;}$$
Angle of twist at peak torque. With $G = 11.5\times10^6$ psi and $J = 2I = 0.3509\text{ in}^4$, taking the torque as carried over the 18 in from the drive end to the load station,
$$\theta = \frac{T_{\max}L_T}{GJ} = \frac{2000(18)}{(11.5\times10^6)(0.3509)} = 0.00892\text{ rad}$$
$$\boxed{\;\theta_{max} = 0.511^\circ\;}$$
Both deflections are comfortably inside normal machinery limits. The bending deflection of 0.018 in at the overhang is well under the 0.001 in-per-inch-of-span guideline often applied to gear and pulley shafts, and the 0.51° twist is far below the 1° per foot of length rule of thumb for power-transmission shafting — here 0.51° over 18 in is 0.34° per foot.
Check: two stated assumptions. (1) The torque-carrying length is taken as the full 18 in from the drive end to the load station, because the figure shows the torque applied to the shaft but does not locate the driving coupling; a shorter torque path reduces the twist in direct proportion. (2) The question states "assume no stress concentration", so $K_f = K_{fs} = 1$. A real bearing seat with a shoulder fillet or a keyway at the load station would carry $K_f \approx 1.6$–$2.2$, which would push the required diameter to roughly 1.6–1.75 in. The 1 3/8 in answer is therefore the idealised-geometry result the question asks for, not a size to build from without a detail-geometry check.