IESL Part III common mistakes
IESL Part III · Understand the traps and how to avoid them.
In structural design questions, the candidate applies partial factors from the wrong limit state (e.g., uses ultimate limit state factors when the question asks for serviceability deflection or crack width checks), producing a design that is safe on paper but fails the stated serviceability criterion.
Candidates memorise a single load combination table and apply it uniformly, without re-reading whether the question is asking for strength, stability, or serviceability. Sri Lankan practice references both BS 8110/5950 and Eurocode, so factor sets get mixed.
What to do instead
Before calculating, write one line stating the limit state and the code/factor set you are using, then check each load case against it. Practise paired problems: the same beam checked for ULS bending and for SLS deflection, so the factor difference becomes automatic.
In building services questions, the candidate sizes a pipe, duct, or cable from a rule-of-thumb figure (e.g., a fixed W/m² cooling load or a standard cable current rating) without adjusting for the actual ambient temperature, diversity, or derating factors given in the question.
Rule-of-thumb values are heavily rehearsed and feel authoritative; candidates skip the derating tables because they are time-consuming under exam pressure.
What to do instead
Treat every given ambient temperature, grouping factor, or diversity figure as a signal to use the derating table. Build a one-page checklist (ambient correction, grouping, thermal insulation, altitude, power factor) and run it on every sizing question in practice.
In electrical and electronics questions, the candidate writes the final numeric answer without units or with the wrong unit (e.g., mA reported as A, or kW reported as kWh), and loses marks even when the method is correct.
Under time pressure candidates focus on getting a number and treat units as cosmetic; many also carry values through calculations without converting prefixes consistently.
What to do instead
Convert every given quantity to base SI units at the start and write the unit beside every intermediate line. Reserve the last 30 seconds of each numerical question to re-derive the unit dimensionally from the formula.
In process and chemical engineering questions, the candidate sets up a mass or energy balance without first drawing and labelling the system boundary and stream table, then loses track of a recycle or purge stream and produces a balance that does not close.
Candidates try to save time by going straight to equations, but recycle, bypass, and purge streams make the topology non-obvious and errors compound silently.
What to do instead
Always sketch the boundary, number the streams, and build a stream table with knowns and unknowns before writing any equation. Do a closure check (total in = total out) before moving to the next part.
In IT and communication questions, the candidate answers a subnetting or addressing question with the host range or broadcast address included, or misstates the number of usable hosts by using 2^n instead of 2^n − 2.
The arithmetic is familiar, so candidates rush the boundary conditions; the network and broadcast addresses are easy to forget when the mask is not a familiar /24.
What to do instead
For every subnetting question, write the block size, network address, first usable, last usable, and broadcast explicitly, then read the question again to see which of these it actually asks for. Drill non-octet masks such as /26 and /27.
In mechanical engineering questions, the candidate selects a standard component (bearing, belt, pump, motor) from a catalogue rating without applying the service factor, load spectrum, or efficiency chain specified in the question.
Catalogue selection is taught as a lookup exercise; candidates treat the catalogue rating as the answer rather than the starting point for correction.
What to do instead
Adopt a fixed selection sequence: required duty, then service/application factor, then correction for speed, temperature, and efficiency, then catalogue choice. Practise by deliberately over-specifying a drive train and checking each correction step.
In agriculture and plantation engineering questions, the candidate calculates irrigation or machinery requirements using a design value (e.g., peak evapotranspiration, field efficiency, or application efficiency) chosen without justification, and never states the assumption.
These problems are open-ended and candidates fear committing to a number, so they pick one silently; markers cannot award method marks for an unstated assumption, and the final figure is often inconsistent with it.
What to do instead
State every assumed design value in one line with its basis (e.g., 'ETo = 5 mm/day, dry zone peak; field efficiency 75%, typical for the stated implement'). Use a consistent set of efficiencies through the chain and sanity-check the result against a known local benchmark.
Across all domains, the candidate spends disproportionate time on one long calculation and leaves later short-answer or definition questions unattempted, or writes a single number with no working for a multi-mark question.
Marks-per-minute are not estimated during the exam; candidates follow question order rather than value, and method marks are invisible to them.
What to do instead
In the first five minutes, mark each question with an estimated time and mark value, then answer in descending order of marks-per-minute. For any calculation worth more than two marks, show the formula, substitution, and result as separate lines.
In code and standards questions, the candidate cites a clause or standard number from memory that is wrong or superseded (e.g., quoting an old edition of a Sri Lankan or British standard), or asserts a legal requirement without naming the authority.
Candidates memorise clause numbers as isolated facts rather than the requirement they contain, and editions change while study notes lag behind.
What to do instead
Learn the requirement and its context rather than the number: state the principle, then cite the standard family and edition you are confident about. If the exact clause is uncertain, describe the requirement correctly and flag the edition; accuracy of the engineering requirement earns more than a confident wrong citation.
In design questions, the candidate produces a technically correct calculation but omits the practical checks the examiner expects: constructability, maintenance access, safety, environmental impact, or cost, and so loses the discussion marks attached to the design.
Candidates treat the exam as a pure calculation test and stop at the number, not recognising that Part III rewards engineering judgement and communication as much as arithmetic.
What to do instead
After every design calculation, write two or three short sentences covering the non-numerical checks the context implies (buildability, safety, maintenance, sustainability, cost). Keep a reusable prompt list and apply it to every design-style question in practice.