PEC EPE Metallurgical common mistakes
PEC EPE Metallurgical · Understand the traps and how to avoid them.
When a stoichiometry or mass-balance item gives a concentrate assay, a recovery percentage, and a product grade, the candidate computes the metal in feed and stops there, reporting the feed metal content as the answer instead of applying the recovery factor and then the grade divisor to get mass of final product.
Under time pressure the arithmetic chain is compressed; the recovery step is mentally noted but skipped because the feed calculation already produced a plausible-looking number with the right units. The candidate also forgets that the product mass must be divided by grade, not multiplied, so answers come out near the feed tonnage.
What to do instead
Adopt a fixed three-line template for every balance item: (1) metal in feed = feed mass x assay; (2) metal in product = metal in feed x recovery; (3) product mass = metal in product / product grade. Write all three lines even when the answer seems obvious, and sanity-check that product mass is always smaller than feed mass.
On Ellingham-diagram and Gibbs-energy items, the candidate reads the standard free energy of formation at the stated temperature and reports it as the actual driving force, ignoring that the reaction quotient term RT ln Q shifts the operating line, or ignores that the diagram plots per mole of O2 so the slope must be rescaled for a different stoichiometry.
Ellingham diagrams are taught visually, so candidates memorise 'lower line wins' and treat the plotted value as the equilibrium condition. They also forget the per-mole-of-oxygen convention, which silently doubles or halves the value they extract.
What to do instead
Before reading any value off an Ellingham diagram, state in writing the normalisation basis (per mol O2) and convert the reaction to that basis. Then apply delta-G = delta-G-standard + RT ln Q explicitly for the stated gas ratios, and check the sign of the correction against Le Chatelier: raising CO/CO2 ratio should make reduction easier.
When asked to predict or explain a microstructure after a heat treatment, the candidate names the phase constituents from the phase diagram alone, without applying the kinetics: they say a steel quenched from the austenite field is martensite even when the section size or the stated cooling rate implies a bainitic or pearlitic product, or they ignore that a slow-cooled alloy may retain coring from incomplete solid-state diffusion.
Phase diagrams are the most heavily drilled topic in revision, so the equilibrium answer is retrieved first and the kinetic qualifier in the stem is treated as decoration. Candidates also conflate 'possible' with 'obtained'.
What to do instead
For every microstructure item, first extract the thermal path (temperature, time, cooling rate, section size) and only then select the transformation product. Use a CCT-versus-TTT distinction check: if continuous cooling is specified, CCT curves govern; if isothermal holding is specified, TTT curves govern. State the governing curve in the answer.
On mechanical-testing items the candidate confuses engineering and true stress-strain values, or reports hardness as a strength without applying the correct empirical conversion, and mixes up the definitions of proof stress, yield strength, and tensile strength when a 0.2 percent offset is specified.
These quantities share units and similar magnitudes, so the labels blur under exam conditions. The 0.2 percent offset construction is remembered as a rule but not as a geometric operation, so candidates default to the upper yield point.
What to do instead
Draw the offset construction on the answer sheet (a line parallel to the elastic slope, displaced 0.2 percent strain) before reading the value. Always state the basis in the answer: engineering stress uses original area, true stress uses instantaneous area. When converting hardness, cite the conversion standard being assumed rather than applying a remembered factor.
On corrosion items the candidate identifies the corrosion type correctly but gives a mitigation that does not address the controlling mechanism, for example recommending a coating for a crevice-corrosion problem that is driven by geometry and stagnant solution, or recommending cathodic protection for an acid environment where it will not polarise usefully.
Mitigation lists are memorised as a block ('coatings, inhibitors, cathodic protection, material selection') and recited without being matched to the specific mechanism named in the stem.
What to do instead
For each corrosion item, first name the mechanism and its controlling variable (oxygen differential, galvanic couple, chloride breakdown of passive film, hydrogen embrittlement). Then choose the mitigation that changes that variable: geometry and drainage for crevice, avoiding dissimilar couples or insulating them for galvanic, alloying or inhibitor for pitting, and control of potential plus hydrogen charging for embrittlement.
On welding and casting items the candidate recommends a process or parameter change without linking it to the specific defect, such as answering a solidification-shrinkage porosity question with a preheat recommendation, or answering a hydrogen-induced cold-cracking question with a pouring-temperature increase.
Defects and remedies are revised as separate lists, so the retrieval cue in the stem ('porosity', 'cracking') triggers the nearest remembered remedy rather than a mechanism-based choice.
What to do instead
Build a defect-mechanism-remedy table for yourself with three columns and practise filling it from memory: shrinkage porosity (feeding, riser design, thermal gradient), gas porosity (degassing, gas content of melt, mould permeability), hot tearing (composition, mould constraint, thermal gradient), cold cracking (hydrogen control, preheat, interpass temperature). Answer by naming the mechanism column first.
On mineral-processing items the candidate calculates a liberation or recovery figure from a particle-size distribution without accounting for the two-product formula, or reports concentrate grade and recovery without checking that they are mutually consistent with the feed grade via the mass balance.
The two-product formula is a short equation that candidates substitute into mechanically, and the consistency check is not part of the memorised procedure, so impossible grade-recovery pairs pass unnoticed.
What to do instead
After any grade and recovery answer, plug both back into the balance: feed grade = concentrate grade x mass yield, with recovery = concentrate grade x mass yield / feed grade. If the numbers do not reconcile within rounding, the answer is wrong. Practise deriving the two-product formula once from the balance rather than memorising it.
On engineering-economics items the candidate applies a simple payback or straight-line depreciation when the stem specifies a discount rate, or compares alternatives by present cost without putting both on the same time basis, and on project-management items they draw a network diagram but do not recompute the critical path after a duration change stated in the stem.
Simple methods are faster and are the ones most practised, so the stated discount rate is treated as extra information. In network questions the first critical path is computed and then frozen, even though the question explicitly changes a duration.
What to do instead
Scan the stem for the words 'discount rate', 'interest', 'present worth', or 'annualised' before choosing a method; if present, use discounted cash flow and state the factor used. For network items, redraw and recompute float for every activity after any stated change, and explicitly list total float for the activities the question asks about.
On engineering-ethics and professional-practice items the candidate answers with the technically optimal option rather than the one consistent with the duty to protect public safety and to report, or they accept a client instruction to suppress a safety finding because 'the client decides'.
Candidates prepare technically for the metallurgy domains and treat ethics as common sense, so they do not apply the specific professional obligation hierarchy: public safety first, then employer and client, then self. The word 'confidential' in a stem is read as an absolute bar to disclosure.
What to do instead
For ethics items, rank the obligations explicitly in the answer: public safety and life, then legal and regulatory duty, then employer or client interest, then personal interest. Where disclosure is required, the correct answer is normally to escalate internally first and then to the regulator if unresolved, documenting the technical basis. Never select an option that conceals a known safety hazard.
On failure-analysis items the candidate jumps from the observed feature (a beach mark, an intergranular fracture path, a corrosion pit) straight to a root cause without stating the examination sequence or distinguishing the failure origin from the final fracture zone, and they omit the distinction between the initiating stress raiser and the overload that completed the failure.
Fractography features are memorised as labels attached to causes, so pattern recognition replaces the investigative sequence. Candidates also answer the 'why did it break' question with the mechanism rather than the root cause.
What to do instead
Structure every failure-analysis answer in the fixed sequence: background and service history, macroscopic examination to locate origin, microscopic examination to establish mechanism, then mechanism-to-root-cause mapping (design, material, fabrication, service). Always separate the initiation site from the final overload zone, and state which evidence supports each step.