Castilla y León PAU Technology and Engineering II common mistakes
Castilla y León PAU Technology and Engineering II · Understand the traps and how to avoid them.
When asked to classify a material or justify a selection, the candidate lists properties (density, hardness, conductivity) without linking them to the service requirement, and picks the material with the 'best' numbers overall instead of the one that satisfies the limiting constraint (e.g. specific stiffness per unit mass for an aerospace bracket).
The selection-criteria topic is taught as a catalogue of properties, so candidates memorise values but never practise the constraint-first reasoning (objective, constraints, free variable, material index) that the question actually rewards.
What to do instead
Before answering, write three lines: function of the part, non-negotiable constraints (must not fail, must not exceed mass/cost), and the objective to optimise. Then eliminate candidates that violate a constraint, and only compare the survivors using the relevant index (E/rho, sigma_y/rho, etc.). Practise with 5-6 typical parts (bracket, shaft, spring, beam, pressure vessel, electrical contact).
In a static equilibrium problem the candidate writes the force equations but forgets the moment equation, or takes moments about a point that eliminates the unknown they were asked to find, then reports a reaction that does not balance the system.
Free-body diagrams are drawn hastily, distributed loads and self-weight are omitted, and the sign convention for moments is chosen inconsistently between lines of working.
What to do instead
Adopt a fixed routine: draw the FBD with all reactions, replace distributed loads by their resultant at the centroid, state a sign convention, then write sum Fx, sum Fy and sum M about a point chosen deliberately (the point through which the unwanted unknown passes). Finish every problem by substituting the reactions back into the unused equation as a check.
In kinematics/dynamics problems the candidate mixes units (mm with m, rpm with rad/s, g with m/s^2) or applies a constant-acceleration formula to a case with non-constant acceleration, producing answers off by factors of 60 or 1000.
Formulas are recalled without their unit assumptions, and rotational speed is taken straight from the statement in rpm into omega = 2pin without conversion.
What to do instead
Convert every datum to SI at the top of the answer and write the conversion explicitly (n rpm -> omega = 2pin/60 rad/s). Before using a formula, state its validity condition (constant acceleration, rigid body, no slip). Keep a one-line 'given / asked / units' header on every dynamics answer.
When analysing a DC circuit with several sources or a bridge, the candidate applies series/parallel reduction where it does not hold, or uses superposition/mesh analysis but forgets to restore a removed source correctly (replacing a current source by an open circuit instead of a short, or vice versa).
Series/parallel rules are over-generalised from simple textbook circuits, and the source-suppression rules are memorised without understanding that an ideal voltage source has zero internal resistance.
What to do instead
First redraw the circuit marking nodes; only combine elements that carry the same current (series) or share both nodes (parallel). For superposition, state the rule aloud each time: voltage source removed -> short circuit; current source removed -> open circuit. Verify with a power balance or by checking one branch with an independent method (Thevenin).
In combinational or sequential logic the candidate writes the truth table correctly but then simplifies the Boolean expression by 'cancelling' terms that are not cancelling (e.g. A + A'B = A + B is fine, but A*B + A'*C = B + C is not), or ignores don't-care states.
Boolean algebra is treated as ordinary algebra, and Karnaugh maps are filled without grouping the largest possible power-of-two blocks or wrapping edges.
What to do instead
Use a Karnaugh map for every simplification of four variables or fewer, group 1s (or 0s for POS) in the largest power-of-two blocks, allow edge wrapping and don't-cares as 1s, and then verify the simplified expression against two or three rows of the truth table. Never cancel a variable that does not appear in every term of the group.
When asked whether a system should be open or closed loop, the candidate answers 'closed loop is always better' without discussing cost, stability, sensor availability or the consequence of a sensor failure.
Closed-loop control is presented as the 'correct' answer in lessons, so the trade-off reasoning (disturbance rejection versus complexity, cost, instability risk) is never practised.
What to do instead
Answer with a two-column comparison: disturbance rejection, accuracy, cost, complexity, stability risk, behaviour on sensor failure. Then justify the choice by the process requirement (e.g. a domestic water tank with a float valve is fine open loop; a temperature-controlled furnace needs feedback). Name at least one concrete risk of the chosen architecture.
In pneumatic/hydraulic circuit questions the candidate draws the actuator and valve but omits the return line, the filter-regulator-lubricator unit, the relief valve or the exhaust silencer, and mislabels the 5/2 valve ports.
Circuits are memorised as pictures rather than as functional chains, and ISO 1219 symbol conventions (numbered ports, arrow directions, spring return, pilot lines) are not drilled.
What to do instead
Build every circuit from the supply outwards in a fixed order: source -> FRL/filter -> pressure relief -> directional valve -> flow control -> actuator -> return. Label ports 1(P), 2(B), 4(A), 3(S), 5(R) on the valve. Check that the drawn valve has the right number of positions and ways for the required actuation, and that the return path to tank is complete.
In manufacturing questions the candidate confuses processes that look similar on paper (turning versus milling, forging versus stamping, arc versus resistance welding) or gives a machining parameter without units or without relating it to the material and surface finish required.
Processes are studied as definitions in isolation, without a mental model of the relative motion between tool and workpiece or of the physical mechanism (chip removal, plastic flow, fusion, pressure).
What to do instead
Anchor each process to one sentence about relative motion or mechanism: turning = workpiece rotates, single-point tool translates; milling = tool rotates, workpiece translates. For parameters, always state cutting speed in m/min, feed in mm/rev or mm/tooth and depth of cut in mm, and say how the choice trades off tool life against productivity and finish.
In project-methodology questions the candidate produces a budget or schedule that omits indirect costs, contingencies, taxes or the critical path, and presents a Gantt chart without dependencies or float.
Budgeting and scheduling are treated as arithmetic rather than as planning under uncertainty, and the distinction between direct/indirect cost and between critical and non-critical activities is not consolidated.
What to do instead
Build the budget as direct costs (materials, labour, machinery) + indirect costs (overheads, energy, administration) + contingency percentage + taxes, and show the total explicitly. For scheduling, list activities with duration and predecessors, compute earliest/latest start and float, and identify the critical path before drawing the Gantt. Mention at least one sustainability or environmental criterion where the question invites it.
In ICT/programming questions the candidate writes code or a network description that ignores boundary conditions (empty list, division by zero, index out of range) or confuses hardware addressing concepts such as IP address versus MAC address and switch versus router.
Programs are written for the happy path shown in the statement, and network devices are memorised by name without their layer and forwarding criterion.
What to do instead
For any algorithm, state the input domain and test the empty, single-element and maximum cases before finalising. For networks, attach each device to its OSI layer and forwarding rule (hub layer 1 broadcast; switch layer 2 by MAC; router layer 3 by IP), and always distinguish logical (IP) from physical (MAC) addressing when the question asks.