Recorded workshop · CIVIL

Structural Design and Analysis with STAAD

Held 18 Mar 20223 videosCertificate available

Notes written by Skill-Lync from the recording · Updated 17 Nov 2026

What this workshop covers

A three-part STAAD.Pro bootcamp: seismic and dynamic analysis of industrial structures, a steel design workflow, and a steel warehouse modelled from scratch.

Part 1, given by an R&D manager at Bentley Systems, covers dynamic analysis in STAAD.Pro: the eigenvalue problem, mass modelling, eigen solvers and Ritz vectors, seismic analysis of industrial structures to IS 1893 Part 4 with a site-specific response spectrum, the 100% plus 30% rule, floor response spectra and forced vibration from machines. Part 2, by a Bentley structural engineer, walks through an eight-step steel design workflow under IS 800:2007: physical versus analytical modelling, custom sections, wind and mass model generators, load envelopes, choosing first- or second-order analysis, design parameters, SELECT optimisation, drawings and OpenSTAAD automation.

Part 3 is a beginner demo that builds and designs a small pinned steel frame with bracing. This Skill-Lync bootcamp was recorded in March 2022 on STAAD.Pro CONNECT Edition. Parts 1 and 2 suit engineers who already use STAAD.Pro, and part 3 suits first-time users. Load values in the part 3 demo are placeholders rather than code values, load combinations are skipped there, and connection design is only mentioned. The first four minutes of parts 1 and 2 are a course introduction you can skip.

Chapters

Part 1

  1. 7:07 Equation of motion and the eigenvalue problem
  2. 10:13 Mass modelling: lumped masses in all three directions
  3. 14:07 Subspace iteration vs Lanczos eigen solvers
  4. 15:39 Ritz vectors: 90% mass participation in fewer modes
  5. 20:39 IS 1893 Part 4 categories and when dynamics is required
  6. 26:40 Site-specific response spectrum setup in STAAD.Pro
  7. 35:21 100% plus 30% rule and spotting torsional modes
  8. 39:35 Floor response spectrum and synthetic time history
  9. 45:34 Forced vibration: steady-state checks for machines

Part 2

  1. 8:21 Eight-step steel design workflow
  2. 14:53 Physical vs analytical modeller for tapered PEB frames
  3. 22:21 Custom sections: Section Wizard and user tables
  4. 27:07 Wind load generator, mass model, IS 800 combinations
  5. 33:25 Choosing first-order, P-delta or buckling analysis
  6. 38:53 Design parameters: STP, FYLD and seismic checks
  7. 43:31 SELECT optimisation and fabrication drawings
  8. 49:41 OpenSTAAD automation with Python and VBA
  9. 53:41 Q&A: time history, plate girders, lifting analysis

Part 3

  1. 0:06 What a steel warehouse is and why steel suits it
  2. 2:36 Modelling options: grids, Structure Wizard, nodes
  3. 5:02 Grouping members and releasing moments
  4. 8:39 Load types, IS 875 and wind vs seismic combinations
  5. 14:13 Demo: new analytical model and beam grid
  6. 20:00 Translational repeat and mirror to build frames
  7. 32:50 Assigning sections and pinned supports
  8. 38:00 Defining dead, live and wind load cases
  9. 48:18 Steel design: check code, select and take-off

Key points from the session

  • STAAD.Pro builds the mass matrix from load items in any load case that carries dynamic commands, dividing them by g. Enter those loads as absolute values in all three directions, or define a separate mass reference load case. The speaker says wrong mass modelling is the most common cause of wrong frequencies and mode shapes.
  • On a long pipe rack, natural modes needed 53 modes to reach 90% mass participation in X and 144 in Z, because many modes are local flutter. Switching the eigen method to Ritz vectors reached 90% in 13 and 18 modes respectively.
  • IS 1893 Part 4 sorts industrial structures into categories 1 to 4 by the consequence of failure. Per the speaker, dynamic analysis is required for category 1 in every zone, category 2 in zones III to V and category 3 in zones IV and V. Other cases can use the simplified static method.
  • For an unscaled site-specific spectrum (Sa/g of 1 at zero period), fold the scaling into the STAAD.Pro direction factor: peak ground acceleration (0.24g in the example) x 0.5 x I/R, with a scale factor of 9.81. The speaker prefers CQC combination, then compares base shear with the code spectrum and scales up if the code value is higher.
  • If one mode shows large mass participation in both horizontal directions, or a torsional mode appears among the first five, mass and stiffness are unsymmetric and you should combine 100% of the earthquake in one direction with 30% in the other.
  • Under IS 800:2007, the moment amplification method is acceptable only while the amplification factor stays below 1.4; above that, run a P-delta (second-order) analysis. A buckling analysis gives the lambda cr/lambda p ratio, which the speaker compares with 10 for frames without modelled infill and 20 when infill walls are modelled.
  • In the IS 800 design block, set STP to tell STAAD.Pro a member is welded rather than rolled, and add the Section 12 seismic parameters only if you need ductile detailing checks. Use SELECT for a single-pass resize and SELECT OPTIMIZED for repeated resize-and-reanalyse cycles, because member forces change as stiffness changes.

Why use Ritz vectors instead of natural mode shapes in STAAD.Pro?

Long or flexible industrial structures often have many local modes that add almost no mass participation, so the 90% participation a response spectrum analysis needs can take well over 100 modes. Load-dependent Ritz vectors target modes the loading actually excites. In the pipe rack demo, 20 Ritz vectors were enough in both directions.

When should I run a second-order (P-delta) analysis instead of a linear static analysis?

The session's rule from IS 800:2007 is to stay with first-order analysis and moment amplification only while the amplification factor is below 1.4. Above that, or when a buckling analysis shows a low lambda cr/lambda p ratio, switch to STAAD.Pro's P-delta analysis, which reapplies the loads on the deflected shape iteratively.

Is a STAAD.Pro response spectrum enough for vibrating machine foundations?

No. Checking that natural frequency is away from the forcing frequency is not enough, because a structure that is not stiff enough can still vibrate excessively. The speaker runs a time history with the machine forcing function long enough to reach steady state, or uses the steady-state frequency sweep, and checks peak amplitude against tolerable limits.

Where to go next

Who should watch

For Civil Engineers.

A 18 Mar 2022 workshop recording from Skill-Lync on CIVIL. For a structured course with projects on this topic, see Skill-Lync's civil engineering courses.

Talk to a career counsellor

Talk to a career counsellor

Free 20-minute call. We'll map the right course to where you want to work.

By submitting you agree to our Terms and Privacy Policy.