What the program is
FEA OpenSees Building is FEA Studio for buildings with OpenSees as its only solver. Every study runs on the OpenSees.exe you installed.
- Left out: Project Chrono, MFEM and the CBFEM connection solver. There is no Solver choice on a study. Connections are designed by the hand checks only.
- Bridges are in FEA OpenSees Bridge. This program has no bridge templates, bridge pushover or bridge time history.
- Kept: templates and samples, IFC/BIM, RC beams, columns, slabs and walls, steel and composite design, foundations, connections, drawings, the section designer and staged demolition.
Install and find OpenSees
Install FEA OpenSees Building from the Download page. Install OpenSees separately from opensees.berkeley.edu. Our verification runs used OpenSees 3.8.0.
The program looks for OpenSees.exe in this order:
| Where | How |
|---|---|
| FEA_OPENSEES | An environment variable holding the path of OpenSees.exe, or of its folder. |
| Your choice | The OpenSees.exe chosen with Analysis › OpenSees program… › Browse… (kept in %LOCALAPPDATA%\FeaStudio\opensees.txt). |
| Beside the program | A folder named opensees next to the application. |
| PATH | Any folder on the PATH. |
| Common folders | Folders named OpenSees… under Program Files, C:\ and your user profile. |
If OpenSees is not found, the study stops with a message saying so.
The OpenSees licence
OpenSees is separate software, distributed by the Regents of the University of California under its own licence. It is free for educational, research and non-profit use. Commercial use, and redistribution, may require a licence from UC Berkeley (Office of Technology Licensing). You are responsible for holding the licence that applies to your use.
- The program does not include, install or download OpenSees. It writes the model as an OpenSees input script and runs your OpenSees.exe.
- The first time a study runs on OpenSees, the dialog OpenSees is separate software says this. Choose I understand to go on. It is asked once on each computer.
Analysis types
The Analysis menu lists ten types. All run on OpenSees.
| Menu item | What it does |
|---|---|
| Linear static | One linear solve per load case or combination. |
| Modal | Periods, mode shapes and mass participation. |
| Linear dynamic | A linear time history: loads, a force history or a ground motion. |
| Nonlinear static | Large-displacement static analysis in load steps (P-Δ or corotational). |
| Pushover | Gravity, then a lateral pattern pushed to a target roof drift: capacity curve, performance point, hinge checks. |
| Nonlinear time history | The building under a record or a record suite: drifts, hinge rotations, floor spectra. |
| Response spectrum | Modal combination (CQC / SRSS) of a design spectrum. |
| Incremental dynamic analysis (IDA) | The time history at rising intensity: drift against intensity, collapse level and fragility. |
| Cyclic | Reversed cycles of rising drift: hysteresis loops and dissipated energy. |
| Collapse | Progressive collapse by column removal (GSA 2016 / UFC 4-023-03 alternate path). |
Pushover, Nonlinear time history, IDA, Cyclic and Collapse each run their own study, and add one if the project has none. IDA copies the time history and Cyclic copies the pushover, with the same records, hazard and loads. Staged demolition (see the Demolition manual) also runs on OpenSees.
Running a study
- Select a study in the Explorer and press F5 (Analysis › Run analysis), or use the Analysis menu item.
- A new pushover, time history or collapse study takes the gravity loads of a seismic study: dead load in full and a quarter of the live load. Wind and other cases are left out.
- The RC members need their bars. Solve the static study and run Design… first; the study tells you when it finds no bars.
- Record suites, IDA and collapse scenarios run several OpenSees processes at once: one per two logical cores, within 70 % of the memory.
- Results › OpenSees › OpenSees log shows what OpenSees printed, condensed: a repeated message is counted once.
The nonlinear model
The nonlinear studies build the OpenSees model at run time from a copy of the project. Nothing of it is saved in the project.
| Member | In the pushover, time history and collapse |
|---|---|
| RC beams (beam lines) | Fibre sections with the designed bars, and the T or L flange of the slab. |
| RC columns (column stacks) | Fibre sections with the storey’s bars and ties, rectangular or circular. |
| Steel beams, columns, braces | I, box and pipe profiles as Steel02 fibre sections. Option Steel members, on by default. |
| RC wall piers (wall design) | MVLEM_3D panels with Concrete02 and Steel02 fibres and an elastic shear spring. Option Wall piers, on by default. |
| Everything else | Elastic. Members without designed bars stay elastic. |
- Rigid joint zone (default 0.25): members start a fraction of a half-depth from the joint, so hinges form near the faces.
- Bar slip at the member ends (off by default): a bond-slip section at each hinge adds the fixed-end rotation. It converges less reliably in a whole frame.
- Check member shear compares end shears with Vc + Vs (ACI 318-19). Stop at shear failure ends the push there, because shear failure is brittle.
Section models
One Section model property on the pushover, the time history (and IDA) and the collapse study sets how beams and columns carry their nonlinearity:
| Choice | What it is | When to use it |
|---|---|---|
| Fibre sections along the member | Every integration point is a fibre section of cover, confined core and bars, so yielding can spread anywhere. The default. | The most general model. |
| Fibre hinges at the member ends | Fibre sections only over the plastic hinge length Lp (Paulay–Priestley) at each end, with a cracked elastic interior. | When you want θp over Lp exactly. A whole frame converges less reliably. |
| Lumped hinges (IMK springs) | An elastic member between zero-length modified Ibarra–Medina–Krawinkler springs, with cyclic strength and stiffness deterioration (Haselton for RC, Lignos for steel). | The usual choice for collapse and IDA. |
The fibre mesh, materials and hinge parameters are in the Fibres and IMK groups of the model parameters.
Materials
- Strengths from the member. Each RC member takes f′c from its concrete material and fy and fyt from the rebar materials that material names. Without them, the concrete design settings apply. Steel members take fy and E from their steel material.
- Concrete: Concrete02 in the pushover, with cracking at ft = 0.33 √f′c. The core is confined by the ties (Kent–Park).
- Bars: Steel02 with 1 % hardening. Options in Member behaviour add bar buckling, fatigue and fracture.
- Expected steel strength (Member behaviour, on by default): steel fy is multiplied by Ry by grade.
- Your own values: a material’s OpenSees values (Define › Materials) and Assign › OpenSees members › Member parameters… override the defaults; a member’s own parameters win.
- Define › OpenSees › OpenSees library… lists the OpenSees materials and elements the program writes, what each models, where it is set, and whether this model uses it.
Ground motions
The commands are in Assign › Ground Motion.
| Command | What it does |
|---|---|
| Import ground motion… | PEER NGA .AT2, text with time and acceleration columns, or one column with a time step. Units g, m/s², cm/s² or mm/s². |
| Import ground motion pair… | Two horizontal components. The second is resampled to the first one’s time step. |
| Import record suite folder… | Every .AT2 file of a folder, grouped by PEER record number; files named …UP, …DWN or …-V are the vertical component. |
| Generate ground motion… | Artificial records matched to the design spectrum: one horizontal, a pair, or a pair and vertical (matched to 2/3 of the spectrum). Up to 30 records at once, on consecutive seeds. |
| Scale ground motion… | By a factor, to a PGA, to Sa at a period, or to the design spectrum (≥ 90 % over 0.2 T1 – 2.0 T1). |
| Select suite records… | Ranks the imported records by the misfit of their spectrum to the target, and sets the suite’s records. |
| Check record suite | Previews the suite’s scaling before the run. |
| Remove ground motion | Removes the shown record. |
There is no online record database: records come from files you import, for example a PEER NGA download. Artificial records are for trying the tools; use real records for design decisions.
Record suites
- ASCE 7-16 / 7-22 Chapter 16: scaling by the maximum-direction (RotD100) spectrum of each pair, the period range of §16.2.3.1, the mean drift limit of §16.4.1.2, unacceptable responses of §16.4.1.1, and at most 2.5 % damping (§16.3.5).
- KDS 41 17 00 : 2022 §7.3.4.1: SRSS spectra of each pair over 0.2–1.5 T1, the mean with 7 or more records (else the largest), and the drift limit of the study’s objective.
Damping and mass
- Mass: lumped at the nodes from the downward loads of the study’s combination. Use the seismic weight. Option: the elements’ own mass plus the other loads.
- Damping ratio (%): Rayleigh damping set at two modes of the loaded frame: the mode with most mass in the shaking direction, and the first mode at which the cumulative mass reaches 90 %.
The Damping and mass (OpenSees) group changes this:
| Property | Choices |
|---|---|
| Damping model | Rayleigh (default), stiffness-proportional, or modal damping. |
| Stiffness of the damping | Committed tangent (default), initial or current. |
| No damping on springs | On by default: zero-length springs get no stiffness-proportional damping. |
| Mass | Lumped from the loads, or Element mass + other loads. |
On isolators, the mass term is dropped and the bearings are not damped.
Devices
Select nodes, then use Assign › Devices. Devices are kept in every frame study and drawn in 3D.
| Device | Command | OpenSees element |
|---|---|---|
| Friction pendulum | Add isolators… | singleFPBearing |
| Lead-rubber bearing | Add isolators… | elastomericBearingPlasticity |
| Triple friction pendulum | Add isolators… | TripleFrictionPendulum |
| High-damping rubber bearing | Add isolators… | elastomericBearingBoucWen |
| Viscous damper (time history only) | Add damper or brace… | twoNodeLink of Viscous, F = C |v|^α |
| Buckling-restrained brace | Add damper or brace… | corotTruss of Steel02 |
- Isolators go under support nodes and replace their support. Vertical stiffness and lift-off can be set per bearing.
- Remove devices and List devices are in the same group.
- Springs and links of other kinds come from the soil, joint and member options below.
Soil springs and foundations
Supports under the footings, in the study’s Supports and devices group, replaces the fixed supports of the footings of Foundations:
| Choice | Model |
|---|---|
| Fixed | The default. |
| Soil springs | Six springs at the column base: a rigid footing on an elastic half-space (Gazetas). |
| Soil springs, footings can lift off | A bed of compression-only springs under the footing carries the vertical load and the rocking. |
| Nonlinear soil (bearing, lift-off, sliding) | A nonlinear Winkler bed: bearing capacity, lift-off and Coulomb sliding. |
| Shallow foundation (ShallowFoundationGen) | The building runs on fixed footings; then each footing runs on its own in 2D with OpenSees’ ShallowFoundationGen. |
The soil comes from Foundations › Soil, or from Soil shear modulus G. The Foundations (OpenSees) group adds radiation damping and pile groups. Site response (OpenSees) runs the record up a 1-D soil column first, with elastic, clay, sand or liquefiable layers.
Analysis settings
Each OpenSees study can carry advanced settings, in property groups. Without them the script is exactly the default.
- OpenSees analysis settings: convergence test and tolerances, the algorithms tried in turn, equation solver, numberer, constraint handler, transient integrator (Newmark, HHT, TRBDF2, explicit) and an arc-length fall-back for static pushes.
- Member behaviour (OpenSees): plastic hinges (fibre or IMK), bar buckling, low-cycle fatigue, bar fracture, degrading shear springs, RC joint springs, steel panel zones, brace buckling, shear–flexure walls (SFI-MVLEM-3D), nonlinear layered slabs and expected steel strength.
- Geometry (OpenSees): P-Δ or corotational geometry, and load steps, for nonlinear static studies.
- Outputs (OpenSees): recorders, link hysteresis, modal properties, convergence history, velocities and accelerations, and the model description. They add outputs and do not change the results.
Where a value is this program’s choice or comes from a source we could not check, the result says so.
Acceptance criteria
| Check | Source |
|---|---|
| RC beam hinges, IO / LS / CP | ASCE 41-17 Table 10-7 |
| RC column hinges | ASCE 41-17 Table 10-8; Table 10-9 for circular columns |
| Steel beams and columns | ASCE 41-17 Table 9-7.1 |
| RC wall piers | ASCE 41-17 Table 10-19 (pier rotation), or TBI 2017 / LATBSDC strain and shear gauges (Wall criteria) |
| Beam–column joint shear | ASCE 41-17 Eq. 10-4 and Table 10-12 |
| Member shear | ACI 318-19 Vc + Vs |
| Target displacement | ASCE 41-17 or FEMA 356 coefficient method; N2 method of EN 1998-1 Annex B |
| Record suite drift | ASCE 7-16 / 7-22 §16.4, or KDS 41 17 00 : 2022 |
The ASCE 41-17 numbers above were checked against the standard. Some modelling assumptions are engineering judgement, for example the wall check assumes unconfined boundaries and high shear; the program logs the status with every check. Results › Seismic › Hinge acceptance and Performance point write the checks to the Output panel.
IDA
- Default scale factors 0.25, 0.5, 0.75, 1, 1.5, 2 and 3. Every record runs all records at each level.
- Intensity levels: scale factors, or Sa(T1) of each record at T1 for the scaling (s).
- A run collapses when a storey drift passes Collapse drift (%) (default 10 %) or the analysis stops converging.
- Charts: Incremental dynamic analysis (drift against Sa(T1)) and, with two or more collapsed records, Collapse fragility (lognormal median and dispersion). Records that never collapse are left out of the fit.
See the IDA tutorial.
Collapse
Progressive collapse study
- Column removal by the alternate path method of GSA 2016 and UFC 4-023-03, on the same fibre frame as the pushover.
- Columns at automatic locations (corner, middle of the long side, middle of the short side, interior) or picked with Assign › Progressive collapse › Pick columns to remove.
- Procedure: nonlinear dynamic, or nonlinear static pushdown with the dynamic increase factor Ω_N.
- Pass: no collapse, and every hinge within the ASCE 41 CP rotation. A run that stops converging is “analysis stopped”, not a pass.
- Results › Seismic › Collapse check; Collapse in 3D in Explorer › Results › Seismic results.
Collapse simulation in a time history
Simulate the collapse removes members that fail (bar fracture, core crushing, shear) and lets the run go on until the building reaches the ground. There is no contact between falling parts. After the first removal the run is a simulation, not an acceptance check. Record suites and IDA run without it.
See the collapse tutorial.
Cyclic pushover
Analysis › Cyclic pushes the frame 0 → +a → −a and back for each drift amplitude. Defaults: amplitudes 0.25, 0.5, 1, 1.5 and 2 %, 2 cycles each (Drift amplitudes (%), Cycles per amplitude). The chart shows the hysteresis loops and the energy they dissipate. There is no idealisation or 80 % rule.
Results and charts
Every chart and check of a seismic study is in three places: Results › OpenSees › Seismic charts, Results › Seismic, and Explorer › Results › Seismic results.
| Study | Charts |
|---|---|
| Pushover | Pushover curve, Capacity spectrum (ADRS), Hinge acceptance (ASCE 41), Joint shear (ASCE 41), Wall piers (ASCE 41) |
| Time history | Time history response, Storey drifts, Record spectrum, Floor response spectra, Energy balance, hinge and joint charts |
| Record suite | Record suite spectra, Record suite drifts |
| IDA | Incremental dynamic analysis, Collapse fragility |
| Collapse | Drop at the removed column or Pushdown curve, Collapse scenarios, Collapse hinge rotations |
- Plastic hinges in 3D shows the hinges on the model, coloured by IO, LS, CP and beyond CP, at the shown step or time.
- Damage elevations draw every frame line with its hinges.
- Results › OpenSees › OpenSees outputs shows the study’s own OpenSees outputs: recorders, link hysteresis, modal participation, convergence and run notes, with CSV export.
Reports
- Seismic evaluation report: the seismic performance evaluation to KDS 41 17 00 : 2022 and ASCE 41, built from the seismic studies.
- Korean volumes: with the report language set to Korean (Explorer › Reports › Settings), Explorer › Reports offers Volume 1 · Linear analysis and seismic design and Volume 2 · Performance-based seismic design, following the Korean tables of contents.
- Calculation report (Results › Seismic): the frame studies chapter by chapter.
- Report information… (Results › Seismic): project, location, use, reviewers and conclusions, which the model cannot give.
Reports are saved as PDF, Word or HTML. See the KDS report tutorial.
Export to OpenSees
File › Export › OpenSees model (.tcl, .py) writes a study as a folder you can run in OpenSees yourself: model.tcl, model.py (OpenSeesPy), a README with the units (N, m, s, kg, Pa) and the result file format, and for a time history ground_motion.txt. Nothing runs, and OpenSees need not be installed. Every study type exports, with the same fibre sections and records as a run.
Limitations
- Steel fibre members have no residual stresses, local or lateral-torsional buckling, or connection fracture. Trusses and other shapes stay elastic.
- Not modelled in the pushover: shear and flexure–shear interaction models, axial-load-dependent column shear strength, confinement by non-rectangular ties. ASCE 41 tables for slabs and coupling beams are not applied.
- The wall check assumes unconfined boundaries, axial load ratio ≤ 0.1 and high shear.
- Progressive collapse: no tie-force method, no linear procedures, no slab membrane action, no removal of walls or of several columns at once. Ω_N and the alternate-path load combination are written from memory of UFC 4-023-03 and need checking.
- Suite element checks (ASCE 7 §16.4.2) are written from memory of the standard and need checking.
- Some IMK beam parameters and the RC joint, shear spring and brace camber shapes are this program’s choices. The result says so.
- The KDS zone factors come from KDS 17 10 00 and are not checked against it.
- There is no online record database, and no vertical floor spectra.
- A collapse simulation has no contact between falling parts.
- Whole-frame runs with fibre hinges at the ends or with bar slip may stop early.
Benchmarks
A 5-storey RC moment frame (3 × 2 bays of 6 m, storeys 4 + 4 × 3.5 m) with fibre sections was modelled identically in the OpenSees model this program writes, in MIDAS Gen NX 9.7.5 and in Perform3D 11. The other program’s result is compared with OpenSees.
| Check | MIDAS Gen NX | Perform3D 11 |
|---|---|---|
| Elastic periods | +1.5 to +2.6 % (modes 1–6) | −0.8 to +2.7 % (modes 1–3) |
| Pushover X, peak base shear | −2.6 % | +6 %, at the same displacement |
| Pushover Y | 7 to 16 % weaker up to 44 mm, where MIDAS stops converging (open) | +6 % |
| Time history, peak roof displacement | +3.1 % | +13 %, at the same time |
| Time history, peak storey-1 drift | 0.36 % against 0.46 % (open) | — |
One building, one record and one direction for the time history. The full comparisons and the open differences are on the validation page.