Mechanical Engineering Paper II: Core Topics
Learning Objectives
- Understand the fundamental concepts of Mechanical Engineering Paper II: Core Topics
- Apply key formulas and techniques to solve problems
- Practice with exam-level questions to build speed and accuracy
Mechanical Engineering Paper II — Manufacturing, Machine Design, Heat Transfer and Industrial Engineering
Manufacturing Processes
- Casting: sand casting, patterns, gating and risering, shrinkage, defects; investment, die and centrifugal casting.
- Forming: forging, rolling, extrusion, drawing, sheet-metal operations; hot and cold working.
- Machining: turning, milling, drilling, grinding; tool geometry, tool wear, cutting forces, machinability.
- Welding: arc, gas, resistance, TIG, MIG, submerged arc; weld defects and inspection.
- Metrology and automation: limits, fits, tolerances, gauges; NC, CNC and robotics.
Machine Design
- Design philosophy: static and fatigue design, factor of safety, stress concentration, S–N curve, endurance limit.
- Joints: bolted, riveted, welded; eccentric loading.
- Components: shafts, keys, couplings, bearings, gears, springs, belts, brakes and clutches.
- Codes and standards: preferred numbers, fits, materials selection.
Heat Transfer
- Conduction: Fourier’s law, composite walls, fins, transient conduction.
- Convection: Newton’s law of cooling, forced and free convection, dimensionless numbers (Re, Pr, Nu, Gr), correlations.
- Radiation: Stefan–Boltzmann law, blackbody and greybody, view factors.
- Heat exchangers: LMTD and NTU methods, effectiveness, fouling.
Industrial Engineering
- Work study: method study, time study, work measurement, ergonomics.
- Production planning and control: forecasting, scheduling, MRP, lean manufacturing.
- Inventory control: EOQ, safety stock, ABC analysis.
- Quality control: control charts, acceptance sampling, Six Sigma; operations research basics.
Common Traps
- Forgetting the riser in casting questions — it compensates solidification shrinkage.
- Confusing endurance limit with yield strength.
- Using arithmetic mean temperature difference instead of LMTD.
- Mixing EOQ carrying and ordering costs.
- Ignoring tool wear and surface finish in machining answers.
Advanced Synthesis: Interdisciplinary Failure Modes and Optimization
The core syllabus teaches you the isolated principles of manufacturing, design, heat transfer, and industrial engineering. However, the UPSC Mains examiner frequently tests your ability to synthesize these topics into complex, real-world scenarios where one discipline dictates the failure of another. This section moves beyond standard textbook problems to address "boundary condition" questions.
Mechanism-Process Coupling: Consider a high-speed stamping operation (Manufacturing). Standard texts discuss tool wear and blank holder forces in isolation. However, an advanced question might ask you to analyze the thermal distortion of the die (Heat Transfer) caused by frictional heating during cold working, and how this distortion alters the part geometry (Machine Design tolerance stack-up). The examiner may present a case where the material's yield strength increases due to strain hardening (Machine Design), but the cooling rate is insufficient to prevent residual stresses that lead to springback (Manufacturing/Forming). You must identify the coupling point: Thermal expansion of the tool altering clearance, leading to increased cutting forces and accelerated wear.
System Optimization Trade-offs: In Industrial Engineering, you learn EOQ for inventory. An advanced scenario introduces a constraint from Machine Design: the warehouse floor must support specific dynamic loads from automated guided vehicles (AGVs). If you optimize purely for storage density (Industrial Engineering), you might violate the structural safety factor of the facility (Machine Design). Similarly, in Heat Transfer, optimizing a heat exchanger for minimum surface area (cost) often conflicts with maintaining a specific LMTD to prevent thermal shock cracking in the tubes (Material failure/Design). The examiner expects you to articulate these trade-offs rather than just calculating an optimal value.
Landmark Data and Exceptions: Textbooks state that ductile materials fail by shear. However, under high strain rates (impact loading) or low temperatures, the fracture mode shifts to brittle cleavage. This exception is critical in crashworthiness analysis. Similarly, while Fourier's law assumes constant thermal conductivity, in advanced manufacturing like laser additive manufacturing, the effective conductivity varies spatially due to rapid phase changes. Recognizing these exceptions demonstrates a depth of understanding that separates average candidates from high scorers.
1. Identify the anomaly: Failure below endurance limit implies a stress concentration or material defect not accounted for in standard design.
2. Manufacturing Link: During installation, if the bolt hole clearance was excessive (poor machining tolerance), it creates a notch effect.
Critical Thinking: Examiner's Twists and Conceptual Traps
The UPSC examiner often uses "twist" questions to test if a student has merely memorized formulas or understands the physics. These twists usually involve changing the boundary conditions of a standard problem.
The Twist in Heat Transfer: A standard question asks for heat loss through a composite wall. The twist? Introduce a phase change material (PCM) layer or a variable thermal conductivity k(T). Standard Fourier law integration fails here; you must use the concept of enthalpy flux during melting. Another twist: Ask for the temperature distribution in a fin with an insulated tip, but then ask how the answer changes if the tip is exposed to convection. The examiner wants to see if you can modify the boundary condition equation (dT/dx = 0 vs -k dT/dx = h(T - T_∞)).
The Twist in Machine Design: Standard problems assume static loading or simple bending. A twist involves a "shock load" scenario where the dynamic factor K_d is applied to a fatigue calculation. Or, consider a gear train where one gear has a slight eccentricity (manufacturing defect). How does this eccentricity induce vibration and alter the contact ratio? The examiner expects you to link the manufacturing error (eccentricity) to the dynamic loading (Machine Design) and the resulting failure mode (pitting or breakage).
The Twist in Industrial Engineering: EOQ models assume constant demand. A twist introduces "uncertain demand" using safety stock logic combined with a learning curve effect where production time decreases over batches, affecting the setup cost assumption. The question might ask: "If the learning rate is 80%, how does the unit cost reduction affect the optimal order quantity?" This requires integrating the learning curve formula with the EOQ derivative.
1. Initial State: Let H0 be the initial slab thickness and Δ Hroll be the geometric reduction set by the mill.
2. Deformation Effect: Under load F, the rolls flatten, increasing the contact area. The effective roll gap increases by δ = (F)/(E I) (simplified elastic deformation model), where E is Young's modulus and I is moment of inertia.
3. Combined Effect: The final thickness H_f is not just H0 - Δ Hroll. It must account for the back pressure from the roll flattening.
4. Expression: H_f = H0 - (Δ Hroll + δ). Here, δ depends on the rolling force which itself depends on the friction and material flow stress. This shows how Machine Design (elasticity) modifies Manufacturing (rolling reduction).
Comparative Analysis: Process Selection for Extreme Conditions
A high-level answer often requires comparing processes not just by cost, but by their suitability for extreme material conditions or geometries.
Casting vs. Additive Manufacturing (AM): Traditional casting relies on fluid flow and solidification shrinkage. AM relies on layer-by-layer melting. A comparative question might ask: "For a turbine blade with internal cooling channels, why is investment casting insufficient compared to Direct Metal Laser Sintering (DMLS)?"
Answer Logic: Investment casting cannot create complex internal geometries without expensive ceramic cores that are prone to breakage. DMLS can print the channel directly. However, AM introduces residual thermal stresses and anisotropic grain structures (Heat Transfer/Design issue), which might require post-processing heat treatment (Thermal processing) not needed in traditional casting. The comparison must highlight this trade-off between geometric freedom and microstructural integrity.
Welding Methods for High-Pressure Vessels: Compare Submerged Arc Welding (SAW) vs. Tungsten Inert Gas (TIG). SAW is fast and cheap but leaves a slag layer and has limited access, making it unsuitable for complex joints in pressure vessels where inspection is critical. TIG offers superior penetration control and cleaner welds but is slower. The decision matrix must include: Weld Speed, Defect Probability, Inspection Accessibility, and Thermal Distortion of the parent material.
1. Material Constraint: Aluminum has high thermal conductivity, making it difficult to weld with conventional arc methods without excessive heat input.
2. Process Evaluation:
- MIG (GMAW): Good speed, but risk of porosity in aluminum; requires AC or pulsed DC.
- Laser Welding: Deep penetration, low heat input (good for distortion), but expensive equipment.
- Riveting: No thermal cycle (no HAZ), excellent fatigue resistance if done correctly, but adds weight.
3. Selection: For critical aerospace structures where fatigue is paramount, Laser Welding or advanced Riveting (blind rivets) is preferred over MIG to minimize Heat Affected Zone (HAZ) which weakens the aluminum grain structure.
4. Conclusion: Laser welding is selected for its ability to create a narrow HAZ, preserving the base material properties essential for fatigue life, despite higher initial cost.
Common Traps in Advanced Sections
- Ignoring Anisotropy: Assuming isotropic material behavior in AM or rolled sheets when the grain structure is directionally dependent (critical in fatigue design).
- Overlooking Thermal History: In heat transfer problems, neglecting the transient cooling rate which dictates residual stress distribution in castings.
- Misapplying Safety Factors: Using a static factor of safety for dynamic or fatigue loading scenarios without converting to an equivalent static load first.
- Neglecting Frictional Heating: In high-speed machining or braking systems, ignoring the conversion of kinetic energy into heat which alters material properties (annealing or tempering).
- Assuming Constant Demand in Inventory: Failing to adjust EOQ models when lead time varies due to supply chain disruptions or seasonal demand fluctuations.
Finite Element Modeling for Structural Integrity Assessment
Finite element analysis (FEA) is the cornerstone of modern machine design validation. The method discretises a continuous domain into a mesh of elements connected at nodes, allowing the solution of the governing equilibrium equations K u = F where K is the global stiffness matrix, u the nodal displacement vector and F the applied load vector. For linear elastic problems the element stiffness matrix is derived from the strain‑displacement matrix B and the material constitutive matrix D as kₑ = ∫ Bᵀ D B dV. Convergence is verified by mesh refinement studies; a typical criterion is that the strain energy norm changes by less than 1 % between successive refinements.
Boundary conditions must reflect the physical constraints: fixed supports impose zero displacement, symmetry planes enforce zero normal displacement and zero shear traction, while contact interfaces require either penalty or Lagrange multiplier formulations to prevent penetration. In rotating machinery, centrifugal body forces are introduced as ρ ω² r acting radially, and thermal loads are incorporated through the thermal strain vector εᵗ = α ΔT where α is the coefficient of thermal expansion.
Post‑processing focuses on stress intensity factors for fracture mechanics, von Mises equivalent stress for yield assessment, and fatigue life prediction using the S‑N curve corrected for mean stress (Goodman or Gerber relations). A common examiner twist is to ask for the effect of mesh distortion on stress concentration factors; the answer emphasises that poorly shaped elements (aspect ratio > 5, Jacobian < 0.2) artificially stiffen the model and underestimate peak stresses.
Non‑Traditional Machining: Electrochemical and Laser Processes
Electrochemical machining (ECM) removes material by controlled anodic dissolution. The material removal rate (MRR) follows Faraday’s law: MRR = (I · M) / (n · F · ρ) where I is the current, M the atomic weight, n the valency, F Faraday’s constant and ρ the density. The inter‑electrode gap (IEG) is maintained typically between 0.1 mm and 0.5 mm by a servo system that adjusts the feed rate to keep the voltage constant. ECM produces no heat‑affected zone, no tool wear, and can machine complex cavities in hard alloys such as Inconel 718.
Laser beam machining (LBM) uses a focused high‑power density beam (10⁶–10⁸ W/cm²) to melt and vaporise material. The key parameters are pulse energy, pulse duration, repetition rate and assist gas pressure. For cutting, the kerf width is approximated by w ≈ 2 · √(α · tₚ/π) where α is thermal diffusivity and tₚ pulse duration. Short pulses (picosecond, femtosecond) enable cold ablation with minimal recast layer, essential for micro‑fluidic channels in aerospace components.
Both processes are selected when conventional cutting forces would cause distortion, when material hardness exceeds tool capability, or when intricate internal geometries are required. The trade‑off lies in capital cost, electrolyte handling for ECM, and the need for precise beam delivery optics in LBM.
Thermal Stress Evaluation in Heat Exchanger Tubes
Heat exchanger tubes experience cyclic thermal gradients that generate stresses beyond the mechanical pressure load. The thermal stress in a long cylinder with a radial temperature distribution T(r) is derived from the thermo‑elastic equations: σ_θ(r) = E α / (1 − ν) · [ (1/r²) ∫₀ʳ T(ρ) ρ dρ + ν ∫ʳᴿ T(ρ) ρ dρ / (R² − r²) ]. For a linear gradient T(r) = T_i + (T_o − T_i)(r − r_i)/(r_o − r_i) the maximum hoop stress occurs at the inner surface and can be approximated by σ_max ≈ E α ΔT / (2 (1 − ν)) where ΔT = T_i − T_o.
Creep‑fatigue interaction is assessed using the time‑fraction rule: ∑ (t_i / t_r_i) + ∑ (n_i / N_i) ≤ 1 where t_r_i is the rupture time at the stress level of the i‑th hold period and N_i the fatigue cycles to failure at the corresponding strain range. The ASME Boiler and Pressure Vessel Code Section III provides allowable stress intensity values S_m that already incorporate a factor of safety of 1.5 on yield and 3.0 on ultimate tensile strength at temperature.
Design mitigation includes using bellows expansion joints to absorb axial growth, selecting low‑expansion alloys (e.g., Incoloy 800H) for high‑ΔT service, and applying a thermal barrier coating on the fire‑side to reduce the surface temperature gradient. A frequent examiner trap is to ignore the effect of temperature‑dependent modulus E(T) on stress redistribution; the correct approach is to iterate the FEA with updated material properties at each temperature step.
Reliability Centered Maintenance and Failure Mode Effects Analysis
Reliability Centered Maintenance (RCM) structures maintenance tasks around the functional failures of each asset. The process begins with a Failure Mode and Effects Analysis (FMEA) that lists every component, its possible failure modes, the local effect, the system‑level effect, and a Risk Priority Number (RPN) = Severity × Occurrence × Detection. Typical severity ratings follow a 1–10 scale where 10 denotes loss of safety or environmental breach.
RCM decision logic then classifies each failure mode into one of four task categories: (1) On‑condition tasks (predictive monitoring such as vibration analysis, oil spectroscopy), (2) Scheduled restoration (overhaul at fixed intervals), (3) Scheduled discard (replace‑at‑life), (4) Run‑to‑failure (no proactive task, accepted for non‑critical functions). The selection criteria are based on the P‑F interval – the time between potential failure detection and functional failure – which must be longer than the inspection interval for on‑condition tasks to be effective.
Quantitative reliability models feed the RCM: the Weibull distribution R(t) = exp[−(t/η)^β] provides the reliability function, where β (shape) indicates infant mortality (<1), random failures (=1) or wear‑out (>1). The characteristic life η is obtained from field data or accelerated life testing. Maintenance optimisation uses the cost‑per‑unit‑time function C(T) = (C_p + C_f · F(T)) / T where C_p is preventive cost, C_f failure cost, F(T) cumulative failure probability at interval T. The optimal interval T* satisfies dC/dT = 0.
Multi‑Objective Optimization in Production Planning
Production planning often requires simultaneous minimisation of makespan, total tardiness, and energy consumption. The problem is formulated as a mixed‑integer linear program (MILP) with decision variables xijk = 1 if job j follows job i on machine k. Objectives: min C_max (makespan), min ∑ T_j (tardiness), min ∑ E_k (energy). Constraints include machine capacity, sequence‑dependent setup times, and release dates.
Pareto optimality is used: a solution is non‑dominated if no other solution improves one objective without worsening another. The ε‑constraint method converts the multi‑objective problem into a series of single‑objective MILPs by fixing upper bounds on all but one objective. For example, minimise makespan subject to ∑ T_j ≤ ε₁ and ∑ E_k ≤ ε₂. By varying ε₁, ε₂ the Pareto front is traced.
Heuristic approaches such as Non‑Dominated Sorting Genetic Algorithm II (NSGA‑II) are employed when the MILP size exceeds solver capability. NSGA‑II maintains a population, applies crossover and mutation, and uses fast non‑dominated sorting with crowding distance to preserve diversity. Convergence is monitored by the hyper‑volume indicator. In examinations, a typical question asks to formulate the ε‑constraint model for a three‑machine flow shop with given processing times and due dates; the answer must show the decision variables, the three objective functions, and the linearised constraints for sequence‑dependent setups.
- Assuming linear material behaviour in FEA when the problem involves plasticity or large deformation.
- Neglecting the temperature dependence of thermal conductivity and specific heat in transient heat‑transfer calculations.
- Using a single safety factor for combined mechanical and thermal loads without separating the load cases.
- Overlooking the P‑F interval requirement when proposing on‑condition maintenance tasks.
- Treating all objectives as equally weighted in a weighted‑sum method, which can hide non‑convex Pareto regions.
- Ignoring the effect of sequence‑dependent setup times on energy consumption in scheduling models.
- Applying the Weibull shape parameter β = 1 for all components without verifying the failure mode distribution.
Additive Manufacturing: Powder Bed Fusion and Directed Energy Deposition
Additive manufacturing (AM) builds components layer by layer from a digital model, enabling geometries that are impossible with subtractive methods. Powder Bed Fusion (PBF) uses a high‑energy laser or electron beam to selectively melt a thin layer of metal powder spread across a build platform. The process repeats until the part is complete, after which excess powder is removed and the part may undergo stress‑relief heat treatment. Key process parameters include laser power (P), scan speed (v), hatch spacing (h), and layer thickness (t). The volumetric energy density E = P / (v × h × t) governs melt pool stability; typical values for Ti‑6Al‑4V range from 50 to 100 J/mm³. Excessive energy density causes keyhole porosity, while insufficient density leads to lack‑of‑fusion defects.
Directed Energy Deposition (DED) feeds metal powder or wire into a melt pool generated by a laser, electron beam, or plasma arc that moves relative to the substrate. DED is suited for repair, feature addition, and large‑scale near‑net‑shape fabrication. The deposition rate R (mm³/s) is proportional to the feed rate and the melt pool cross‑section. Process control relies on real‑time monitoring of melt pool temperature and geometry using infrared cameras or photodiodes. Both PBF and DED produce anisotropic microstructures due to rapid solidification and directional heat flow; columnar grains grow parallel to the build direction, affecting mechanical properties such as yield strength and fatigue life. Post‑process hot isostatic pressing (HIP) at 920 °C and 100 MPa for 2 h is commonly applied to close internal pores and homogenise the microstructure.
Design for AM requires consideration of support structures, residual stress distortion, and surface roughness. Supports are needed for overhangs exceeding 45° from the horizontal; they are removed mechanically or by chemical etching. Residual stresses arise from steep thermal gradients and can cause warpage or cracking; strategies include optimized scan strategies (e.g., island scanning, stripe rotation), pre‑heating the build plate to 200–400 °C, and topology optimization to minimize material volume while maintaining stiffness. Surface roughness (Ra) of as‑built PBF parts typically lies between 5 and 15 µm, necessitating machining or polishing for critical interfaces.
Gear Design and Contact Stress Analysis
Gear design begins with selecting the module m (mm), pressure angle φ (usually 20°), and number of teeth z for pinion and gear. The pitch diameter d = m × z, and the base circle diameter db = d × cos φ. The Lewis bending stress formula σ_b = (F_t × K_a × K_v × K_m) / (b × m × Y) estimates tooth root stress, where F_t is tangential load, K_a application factor, K_v dynamic factor, K_m load distribution factor, b face width, and Y Lewis form factor dependent on tooth count and pressure angle. For high‑precision gears, the AGMA 2001‑D04 standard replaces Y with the geometry factor J and introduces the fatigue stress concentration factor K_f.
Contact (Hertzian) stress at the line of action is given by σ_c = C_p × √( (F_t × K_a × K_v × K_m) / (b × d_p × I) ), where C_p is the elastic coefficient (√(1/π × ( (1‑ν₁²)/E₁ + (1‑ν₂²)/E₂ ) ) ), d_p is pinion pitch diameter, and I is the geometry factor for pitting resistance. The factor I accounts for the ratio of base circle radii and the contact ratio. For a 20° pressure angle, standard full‑depth teeth, I ≈ 0.1–0.12. The allowable contact stress σ_c,allow is obtained from material pitting resistance curves (e.g., AGMA 2001) and must exceed the calculated σ_c with a safety factor of at least 1.2 for industrial gearboxes.
Lubrication regime influences both bending and contact fatigue. The specific film thickness λ = h_min / σ_composite (ratio of minimum elastohydrodynamic film thickness to composite surface roughness) determines whether the gear operates in boundary (λ < 1), mixed (1 ≤ λ ≤ 3), or full‑film (λ > 3) regime. Full‑film lubrication dramatically reduces micropitting risk. Gear rating also requires checking scuffing criteria using the flash temperature method (AGMA 925) and verifying that the gear mesh alignment tolerance (typically ≤ 0.02 mm for precision gears) does not introduce edge loading that elevates local contact stress beyond the allowable limit.
Transient Heat Conduction in Composite Walls
Transient heat conduction through multilayer walls is governed by the one‑dimensional heat equation ∂T/∂t = α ∂²T/∂x² in each layer, where α = k/(ρc_p) is thermal diffusivity. At interfaces, temperature and heat flux continuity apply: T_i = Ti+1 and k_i ∂T_i/∂x = ki+1 ∂Ti+1/∂x. For a wall of N layers subjected to a step change in surface temperature or convective boundary condition, analytical solutions use the Laplace transform method, yielding infinite series of exponential terms with eigenvalues determined from the transcendental characteristic equation involving layer thicknesses L_i, conductivities k_i, and diffusivities α_i.
In practice, the finite‑difference or finite‑element method is preferred for arbitrary boundary histories. An implicit Crank‑Nicolson scheme provides unconditional stability: (T_in+1 - T_i^n)/Δt = 0.5 α ( (Ti+1n+1 - 2T_in+1 + Ti-1n+1)/Δx² + (Ti+1^n - 2T_i^n + Ti-1^n)/Δx² ). The thermal contact resistance R_c (m²K/W) at imperfect interfaces adds a temperature jump ΔT = q″ R_c, which can be modeled as an additional thin layer of equivalent conductivity k_eq = Δx / R_c. For a typical steel‑ceramic interface, R_c ≈ 1×10⁻⁴ m²K/W, causing a measurable lag in heat‑flux response.
The concept of thermal penetration depth δ = √(π α t) indicates how far a thermal wave travels in time t. In a composite wall, the effective penetration depth is limited by the lowest diffusivity layer. For fire‑protection design, the time to reach a critical temperature at the unexposed face is estimated by solving the transient problem with the standard fire curve (ISO 834) as the hot‑side boundary condition. The result is often expressed as a fire‑resistance rating (e.g., 60 min, 120 min) that must satisfy building code requirements.
Lean Manufacturing and Value Stream Mapping
Lean manufacturing seeks to eliminate waste (muda) across the value stream by applying the five principles: define value, map the value stream, create flow, establish pull, and pursue perfection. Value Stream Mapping (VSM) visualises material and information flows from raw material to finished product. The current‑state map captures cycle time C_t, changeover time C_o, uptime U, inventory levels I, and lead time L_t for each process step. The future‑state map targets reduced lead time by implementing continuous flow, kanban pull systems, and quick changeover (SMED) techniques.
Key metrics include Process Cycle Efficiency (PCE) = Value‑Added Time / Total Lead Time. In many batch‑oriented shops, PCE is below 5 %; lean transformations aim for > 25 %. Takt time T_t = Available Production Time / Customer Demand sets the pace of production. Each work cell must complete its cycle within T_t; otherwise, line balancing or additional resources are required. Heijunka (production leveling) smooths the schedule by sequencing product variants in a repeating pattern, reducing demand amplification (the bullwhip effect) upstream.
Standardised work documents the optimal sequence, timing, and work‑in‑process for each operation, enabling quick identification of deviations. Visual management tools such as Andon lights, shadow boards, and 5S (Sort, Set in order, Shine, Standardise, Sustain) reinforce discipline. Continuous improvement (Kaizen) events use the Plan‑Do‑Check‑Act cycle to test countermeasures on a small scale before full deployment. The effectiveness of lean initiatives is tracked through Overall Equipment Effectiveness (OEE) = Availability × Performance × Quality, with world‑class targets above 85 %.
Vibration Analysis of Rotating Machinery
Rotating machinery vibration is characterised by synchronous (1×) and non‑synchronous components. The 1× vibration amplitude is primarily due to mass unbalance; the force F_u = m_e ω² r_e, where m_e is equivalent mass, ω rotational speed (rad/s), and r_e eccentricity. Balancing standards (ISO 21940‑11) define permissible residual unbalance U_per = G × M / ω, where G is the balance grade (e.g., G2.5 for precision rotors) and M rotor mass. Two‑plane balancing corrects both static and couple unbalance by adding correction masses at specified angular positions.
Bearing defects generate characteristic frequencies: Ball Pass Frequency Outer (BPFO) = n/2 × (1 - d/D cos α) × f_r, Ball Pass Frequency Inner (BPFI) = n/2 × (1 + d/D cos α) × f_r, Ball Spin Frequency (BSF) = D/(2d) × (1 - (d/D cos α)²) × f_r, and Fundamental Train Frequency (FTF) = 1/2 × (1 - d/D cos α) × f_r, where n is number of rolling elements, d ball diameter, D pitch diameter, α contact angle, and f_r shaft rotational frequency. Envelope demodulation of the acceleration signal isolates these defect frequencies from the broadband noise.
Rotor‑dynamic analysis uses the Jeffcott rotor model to predict critical speeds ω_cr = √(k/m) where k is bearing stiffness and m rotor mass. For multi‑disk rotors, the finite‑element method assembles mass, stiffness, and gyroscopic matrices; the eigenvalue problem (K + iωG - ω²M) φ = 0 yields complex natural frequencies and mode shapes. Campbell diagrams plot damped natural frequencies versus spin speed, highlighting intersections (critical speeds) and instability zones due to bearing cross‑coupled stiffness. Active magnetic bearings can shift critical speeds and provide damping control, but require robust controller design to avoid subsynchronous instability.
- Assuming uniform energy density across the build chamber in PBF without accounting for powder layer thickness variations.
- Using the Lewis form factor Y for high‑contact‑ratio gears where the AGMA geometry factor J is required.
- Neglecting thermal contact resistance at layer interfaces in transient composite‑wall calculations.
- Calculating takt time based on total available minutes instead of net available time after planned downtime.
- Applying a single balance grade G to all rotor components regardless of their individual mass and speed.
- Interpreting envelope‑demodulated spectra without verifying the carrier frequency band matches the bearing defect frequency range.
- Each of these traps can be avoided by systematically checking the underlying assumptions against the specific problem data.