PE Mechanical Thermal Fluids common mistakes
PE Mechanical Thermal Fluids · Understand the traps and how to avoid them.
Solving a Rankine, Brayton, or refrigeration cycle problem by hunting for the right equation instead of first sketching the cycle on T-s and h-s coordinates and labeling each state with its known properties.
The exam presents cycles in prose or simple tables, so candidates jump straight to formulas. Because state points are not enumerated, they mix up which enthalpy belongs to which stream, apply the pump work sign inconsistently, or forget that turbine and compressor efficiencies apply to the actual work, not the isentropic work. The result is a numerically plausible but wrong answer, and under time pressure the candidate often cannot see where it went wrong.
What to do instead
Before computing anything, draw the cycle and build a state table with one row per numbered state and columns for pressure, temperature, quality, enthalpy, and entropy. Fill only entries you can justify from the problem statement. Then traverse the table device by device (pump, boiler, turbine, condenser) writing an energy balance for each. This makes sign conventions and efficiency placement explicit and turns a search for a formula into a bookkeeping exercise. Practice until the table takes under two minutes.
Treating a heat exchanger or mixing problem as a simple arithmetic average of temperatures, or applying LMTD without checking for counterflow versus parallel flow and without the correction factor for multipass or crossflow arrangements.
The LMTD formula is memorized in its simplest counterflow form, and candidates plug in terminal temperature differences without confirming the configuration. When the problem describes a shell-and-tube unit with multiple passes or a crossflow arrangement, the uncorrected LMTD overstates the driving force, so the required area comes out too small. Arithmetic-mean approximations compound the error at low effectiveness.
What to do instead
For every heat exchanger problem, first identify the configuration and the hot and cold terminal temperatures at each end. Compute LMTD from the two terminal differences in the correct orientation, then apply the correction factor F when the configuration is not pure counterflow. As a cross-check, compute effectiveness from the capacity rates and compare with the effectiveness-NTU relation for that configuration; the two methods should agree. If they do not, the configuration assumption is wrong.
Using the wrong reference state or the wrong specific heat when converting between enthalpy, internal energy, and temperature for ideal gases and steam, including mixing Cp and Cv or assuming constant specific heats over a large temperature range.
Reference states differ between tables (steam tables use saturated liquid at the triple point, ideal gas tables use 0 K or 298 K), and candidates subtract enthalpies from different references. Similarly, they use Cp where Cv belongs (closed rigid tank versus flow device) or apply a room-temperature specific heat across a 600 K span. The arithmetic looks clean, so the error survives review.
What to do instead
Write the governing first law for the specific device before substituting numbers: closed system at constant volume needs Cv and internal energy; steady-flow device needs Cp and enthalpy. For large temperature changes, use ideal gas tables or an integrated average Cp rather than a single value. For steam, always read enthalpy and entropy from the same table family and never mix saturated and superheated references. State the reference explicitly in your scratch work.
Confusing absolute and gauge pressure, or dropping the atmospheric term, when computing compressor work, boiler pressure, or cavitation margin at a pump inlet.
Problems often quote gauge pressures in the narrative and absolute pressures in tables. Candidates substitute gauge values into relations derived for absolute pressure, or compute NPSH available without adding atmospheric pressure and subtracting vapor pressure. The error is a constant offset, so the answer is close to a distractor and feels right.
What to do instead
Adopt a habit: convert every pressure to absolute in the first line of the solution and note the conversion explicitly. For pump inlet problems, write NPSH available as the sum of static head and atmospheric pressure head minus vapor pressure head minus friction, and confirm each term is in absolute pressure. When the answer choices are separated by roughly one atmosphere, that is a signal the gauge-versus-absolute trap is being tested.
Applying the Darcy-Weisbach or Hazen-Williams relation with a friction factor read from the Moody chart at the wrong relative roughness, or using the laminar friction factor formula in turbulent flow and vice versa.
The Reynolds number must be computed first, and candidates skip that step or use diameter in the wrong units, so they land in the wrong flow regime. They also forget that relative roughness is dimensionless and divide by diameter inconsistently, or they interpolate the Moody chart carelessly. Because friction factor changes slowly, a wrong reading still yields a plausible head loss.
What to do instead
Compute Reynolds number explicitly and compare with 2300 and 4000 before choosing a friction factor relation. In laminar flow use f = 64/Re; in turbulent flow use Colebrook or an explicit approximation such as Swamee-Jain, and verify the relative roughness is dimensionless. Keep units consistent in the same unit system throughout the head loss calculation. A quick sanity check: doubling flow rate should roughly quadruple head loss in fully turbulent flow.
Neglecting minor losses, entrance and exit effects, or the velocity head at a free surface when applying the energy equation to a piping system, and misassigning the datum elevation.
Minor losses are small in long pipelines but dominate in short, fitting-heavy systems, and candidates trained on textbook long-pipe problems drop them by default. They also assume a large tank surface has zero velocity but then forget that the exit loss coefficient accounts for discharging into a reservoir. Datum confusion flips the sign of elevation terms.
What to do instead
Write the energy equation between two clearly labeled points with the datum drawn on a sketch before substituting. Include a sum of minor losses with each fitting's K value, and remember that a pipe discharging into a tank contributes an exit loss of K = 1.0. Check whether the problem states the pipe is long enough that minor losses are negligible; if it does not say so, include them.
Choosing a pump or fan by flow rate alone, or reading the operating point off the pump curve at the design flow without intersecting the system curve.
Candidates treat the rated flow and head as the operating point, ignoring that the actual operating point is where the pump curve crosses the system curve. They also forget affinity laws when speed or impeller diameter changes, or compare specific speed in inconsistent units. The result is a pump selected at the wrong point on its curve, often far from best efficiency.
What to do instead
Always construct the system curve (static head plus flow-dependent losses) and intersect it graphically or algebraically with the pump curve to find the operating point. When speed changes, apply the affinity laws to the entire curve before intersecting. Check that the operating point falls near the best efficiency point and that NPSH available exceeds NPSH required. Report the operating flow and head, not the rated values.
Mishandling dimensionless groups by mixing unit systems, forgetting the conversion factor in the convection coefficient, or using the wrong characteristic length in the Reynolds and Nusselt numbers.
Dimensionless correlations are unit-agnostic only if every input is consistent, yet candidates plug Btu/hr-ft2-F into correlations expecting W/m2-K, or use pipe diameter where hydraulic diameter of a noncircular duct belongs. They also forget that the Nusselt number yields a convection coefficient only after dividing by the correct characteristic length. The resulting coefficient is off by an order of magnitude.
What to do instead
Convert all inputs to SI before entering any correlation, then convert the result back if needed. Identify the characteristic length explicitly for each geometry (diameter for internal pipe flow, hydraulic diameter for noncircular ducts, length along the plate for external flow) and write it down. Confirm whether the correlation is for local or average Nusselt number and whether properties are evaluated at bulk, film, or wall temperature.
Treating a transient conduction or lumped-capacitance problem as steady state, or applying the lumped-capacitance model without checking that the Biot number is small.
The problem statement may mention a time or a sudden change, but candidates default to steady-state resistance networks because those are more familiar. Conversely, they apply the exponential lumped model to a large object where internal gradients matter, producing a temperature that is wrong in both magnitude and trend.
What to do instead
Scan the problem for time dependence first; if any property changes with time, it is transient. Compute the Biot number using the correct characteristic length (volume over surface area for lumped analysis) and compare with 0.1. If Bi is below 0.1, use lumped capacitance; otherwise use the one-term approximation for a plane wall, cylinder, or sphere with the appropriate Biot and Fourier numbers. State the criterion in your solution so the grader sees the check.
Rushing the unit conversions and significant figures in combined thermodynamics and fluids problems, then selecting the answer choice closest to an intermediate result rather than recomputing from the final expression.
The exam mixes English and SI units, and candidates chain conversions through several steps, dropping a factor of 60 for minutes to seconds or 32.2 for pound-mass to pound-force. Under time pressure they round intermediate values aggressively, then pick a nearby distractor that matches the rounded intermediate. This is the most common source of missed points among otherwise competent candidates.
What to do instead
Carry units through every line as if they were algebraic symbols and cancel them explicitly; a unit mismatch reveals the missing factor. Keep full precision in intermediate values and round only the final answer. Before selecting a choice, re-derive the final expression symbolically and substitute once. Budget a final two-minute check per problem to confirm the answer's order of magnitude against a physical estimate.