Structural Engineering — Good-to-Know Rules of Thumb

Structural Engineering — Good-to-Know Rules of Thumb

Practitioner judgment for structural engineers and S.E. candidates. These are heuristics, not code — verify every design against the governing standard.

Load estimation (back-of-envelope)

• Building weight: offices/residential ~150–200 psf of floor area total (dead + partitions + MEP) for concrete; ~100–130 psf for steel frame with composite deck. Use for quick base-shear estimates: V ≈ Cs × W.
• Seismic base shear sanity: for ordinary buildings, V is typically 5–15% of W in moderate seismic zones, 10–25% in high-seismic zones. If your computed Cs gives 40%, something is wrong (check R, Ie, period).
• Wind base shear sanity: roughly 1–3% of W for low-rise, up to 5% for tall flexible buildings. If seismic and wind differ by an order of magnitude, recheck both — they're usually within 3× of each other for mid-rise.
• Live load reality: offices 50 psf design, but actual sustained live load is ~10–15 psf. The code value covers the rare full-loading event plus uncertainty — don't "optimize" it away.

Concrete

• Slab thickness: one-way slabs ~L/24 (interior spans), two-way flat plates ~L/30 to L/33 for deflection control. A 30-ft flat plate wants ~11–12 in.
• Beam depth: ~L/16 to L/21 for continuous beams; deeper is cheaper than more steel — concrete is cheap, labor for congested rebar is not.
• Column size: axial capacity ≈ 0.5 f'c Ag for a quick check of a tied column at moderate eccentricity. A 24×24-in column at 5 ksi carries ~0.5×5×576 = 1,440 kips — then check slenderness and moments properly.
• Development length: #9 bar in 4 ksi concrete ≈ 40–50 bar diameters for tension development; hooks cut it to ~15–20 db. When in doubt, add the hook.
• Punching shear: if the factored reaction exceeds ~0.5 × φVc at an interior column, start thinking about drop panels or stud rails early — retrofitting punching shear is miserable.
• Crack control: 0.016-in crack width is the corrosion threshold people design around; more cover beats smaller bars for durability in marine exposure.

Steel

• Beam weight: a floor beam spanning 30 ft at 10-ft spacing carrying 100 psf needs roughly a W18–W21 in the 40–60 lb/ft range. The rule "depth in inches ≈ span in feet × 2/3" gets you close (30 ft → ~W20).
• Column curve: KL/r = 50 → Fcr ≈ 0.8 Fy; KL/r = 100 → Fcr ≈ 0.4 Fy; KL/r = 150 → Fcr ≈ 0.2 Fy. Memorize these three points and you can sanity-check any column.
• Lateral-torsional buckling: unbraced length Lb > ~15× flange width (roughly Lp) starts costing you moment capacity. Brace the compression flange at every purlin and stop worrying.
• Connection philosophy: a connection that is 10% stronger than the member is money well spent; a connection that governs the member size is money wasted. Design connections for the required strength, not the maximum conceivable.
• Bolts vs. welds: bolts are for erection tolerance, welds are for stiffness. A bolted moment connection in the field will slip before it yields — if drift matters, weld or slip-critical.

Seismic

• Period: T ≈ 0.1N seconds for steel moment frames (N = stories), ≈ 0.06N for concrete shear-wall buildings. A 10-story steel frame at T = 1.0 s — if your analysis says 2.5 s, your model is too flexible or your masses are wrong.
• Drift: 1% story drift is where nonstructural damage gets expensive; 2% is where structural damage starts in ductile frames. If your drift is 0.2%, you're leaving money on the table (or your R is wrong).
• R values mean damage: R = 8 doesn't mean the building feels 1/8 the force — it means you're counting on the building to yield and dissipate energy. Detail for the ductility you're claiming, or the R is a lie.
• Soft story: if one story is visibly more open (parking, lobby), it IS a soft story until proven otherwise. Strengthen it or add walls — the earthquake will find it.
• Diaphragms: flexible wood diaphragms span like deep beams — aspect ratio > 3:1 starts needing collector scrutiny. Rigid concrete diaphragms distribute by stiffness — check that your shear walls can actually take what stiffness attracts.

Foundations

• Bearing: 2–4 ksf for medium clays/sands, 6–12 ksf for dense sands/gravels, 20+ ksf for sound rock — then verify with a geotechnical report, always.
• Settlement: 1 in. total, 1/2 in. differential is the serviceability budget for ordinary buildings. If your computed settlement is 3 in., change the foundation system, not the superstructure.
• Piles: end bearing in dense stratum is worth 3–5× the shaft friction in soft soils per dollar of pile. Get to bearing if you can reach it economically.
• Liquefaction: if the water table is within 50 ft and the soil is loose saturated sand, assume it liquefies until a proper triggering analysis says otherwise. The foundation design starts from that assumption.

Wind

• Pressure scales with V²: a 10% increase in wind speed is a 21% increase in pressure. Never interpolate pressures linearly with speed.
• Corners are killers: roof corner (Zone 3) suctions are 2–3× field-of-roof values. Cladding budgets should be set by the corners, not the field.
• Flexible structures: if the natural frequency is under 1 Hz, dynamics matter — gust factors, vortex shedding, across-wind response. A 1-Hz rule of thumb separates static from dynamic wind design.

Professional judgment

• The 10% rule: if a refinement changes the answer by less than 10%, it wasn't worth the engineering hour. Spend effort where the sensitivity is.
• Draw what you design: if you can't sketch the load path on a napkin — gravity from slab to foundation, lateral from diaphragm to ground — you don't understand the structure yet.
• Symmetry is cheap insurance: symmetric, regular structures outperform clever irregular ones in every earthquake on record. Regularity is a design choice, not a constraint.
• Peer review your own work: run one independent check by a different method (hand calc vs. software, or two software models). The errors that kill are the ones no one re-computed.
• Document assumptions: the next engineer (or the forensic engineer) will judge your design by what you wrote down. A calculation with stated assumptions is engineering; without, it's guessing.
• Know when to say S.E.: in full-restriction states (IL, HI) there's no judgment call — but in partial-restriction states, the triggers (heights, occupancies, irregularities) are minimums. If the structure scares you, it needs an S.E. regardless of whether the statute demands one.

File: structural/good-to-know-rules-of-thumb.md — heuristics for study and practice; always verify against the governing code edition.

Original ETS reference writing, verified against public sources on 2026-09-30. Thin sourcing is flagged inline. Not legal advice — confirm current editions with the issuing body.