I usually design mid-rise RC frames in ETABS, but I’m about to tackle a 65 m steel pedestrian truss and want to sanity-check serviceability and stability assumptions. For those who do both buildings and small bridges, do you target vertical f1 > 3 Hz, run P-Delta with eigen-buckling factors > 10, and use AASHTO LRFD’s 90 psf pedestrian load, or is a lower crowd load (around 0.5 kN/m²) more realistic for a campus span?
For a 65 m truss I try to push vertical f1 to about 3.5–4 Hz and lateral >1.5 Hz; , chasing that last bit of frequency can snowball, so a small TMD has saved me more than once. I keep “90 psf” per AASHTO unless the owner signs off on lower densities, and I run P-Delta with φ_cr > 10 but still check chord slenderness/K so the global factor doesn’t hide member issues. Coming from ETABS, make sure truss panel joints are released so you’re not inflating stiffness; quick accel sanity check here: https://www.fhwa.dot.gov/publications/research/infrastructure/structures/03054/.
Quick example: on a 65 m steel truss we landed f1 ≈ 3.2 Hz and lateral about 1.6 Hz — , chasing another 0.2 Hz cost more steel than a 1% mass TMD, so we added the TMD and checked footfall per FHWA 04–041 (https://www.fhwa.dot.gov/publications/research/infrastructure/structures/04041/04041.pdf). I do run P-Delta and like eigen-buckling >10, and I’ll stick with “90 psf” unless the owner explicitly signs off on less; small caveat in ETABS: double-check mass on deck/rails so modes aren’t optimistic. Are you expecting any synchronized crowd events or just commuter traffic?
@OP On a ‘65 m’ truss we ran in ETABS, vertical f1 cleared 3 Hz but lateral fell to about 1.4 Hz after adding the real handrail/glass mass — , lateral mode always bites — and pedestrian accel governed; a midspan K-frame and one extra lateral tie bumped it to about 1.8 Hz and passed SETRA/HiVoSS checks (https://civil.columbia.edu/files/seasdeptcivil/butcher_setra_footbridges.pdf). I still use AASHTO for strength, but for serviceability I’d prioritize the accel check and member LTB/bracing over chasing more global f1.
On my last about 65 m truss in ETABS, vertical f1 cleared “3 Hz” until we added the real handrail/glass as line masses; that pulled vertical about 0.3 Hz and lateral by almost 1 Hz, so we switched to Ritz with about 30 modes. I run P-Delta, but I don’t sweat an eigen-buckling factor >10 — 8–9 was fine once second‑order drifts were tiny. @OP I still use AASHTO’s 90 psf for the envelope, but the best payoff was getting the mass source right; if weight is tight, that’s where you’ll win.
Quick tip from a similar job: model the real handrail/glass and conduit as continuous line mass before you chase frequencies — ours pulled the lateral mode down about 0.8 Hz, so we left a midspan TMD pocket just in case (the glass is a stubborn roommate: lots of mass, little stiffness). I still use AASHTO’s 90 psf for strength, but if the owner wants lower crowd load, get it in writing; @sjohns19 is right that lateral is the trap — have you checked torsion coupling with the parapet eccentricity?