SR2b · RESPONSE-SPECTRUM-3D-BRACED-FRAME · Analysis Benchmarks

Response-spectrum analysis of a three-dimensional L-shaped braced frame

Brace + rotational-mass + tabulated-spectrum RSA verified on an L-shaped three-dimensional braced frame

PASS +0.108% error tolerance 1.5% 3.05908 Hz vs 3.0592 Hz
§2

Technical features tested

  • 3-D response-spectrum analysis (RSA)
  • Braces / axial-only members (linear Truss)
  • Rigid floor diaphragms (Ux, Uy, Rz)
  • Rotational mass (MMI) — offset-mass condensation
  • Tabulated El-Centro spectrum · 5% damping
  • CQC · SRSS · ABS · axial-force recovery
§3

Problem description

SR2 verified the 3-D response-spectrum pipeline on a moment frame; SR2b takes it onto the brace surface and adds rotational mass. Adopted as CSI SAP2000 Verification Example 1-025 with the independent solution from Peterson (1981), a three-story, L-shaped building is framed by four identical two-bay X-braced frames whose columns and diagonals carry axial force only — every member is emitted as a linear element Truss. All mass is concentrated at a centre-of-mass joint per level with X and Y translational mass and a rotational mass moment of inertia (MMI) about Z; the plan is L-shaped, so the CM is offset from the centre of rigidity and the response is translation–torsion coupled. The excitation is the El-Centro-derived tabulated response spectrum in X with 5% damping, combined by CQC, SRSS and ABS; only the first two modes drive the RSA. The production mass path injects translational mass only, so the MMI is reproduced physically by four offset translational masses on the diaphragm — the de-Zhu T-matrix mass condensation (M_red = TᵀMT) turns them into the exact Z rotational inertia, with no production-code change. The reference is reproduced directly at its own discretisation (84 truss members, 3 concentrated story masses).

F1 F2 F3 F4 CM Sa(X) L-plan · MassX,Y + MMI at CM · rigid diaphragm/storey 1 4 6 2 bays @ 20 ft · 3 storeys @ 12 ft · axial-only (Truss) Braced frame — one of four
Figure SR2b. L-shaped plan with the four axial-only braced frames and the eccentric centre of mass carrying MassX, MassY and MMI (left), and the elevation of one two-bay X-braced frame with pinned, axial-only members (right). Frame-1 elements 1 (column), 4 and 6 (diagonals) are the axial-force comparison points.
§4

Geometry, properties & loading

Geometry

Braced frames
4
Bays / frame
2
Storeys
3
Bay / storey
20 / 12 ft

Members & mass

E
29500 ksi
Area
6 in²
Members
Axial-only (Truss)
Mass
MassX,Y + MMI at CM

Loading & dynamics

Excitation
El-Centro spectrum (X)
Damping
5 %
Modes used
2
Combination
CQC · SRSS · ABS
§5

Reference solution

  1. Frequency — mode 1(K − ω²M)·φ = 0, f = ω/2π= 3.0592 Hz
  2. Roof CM Ux — CQCu = √(Σᵢ Σⱼ ρᵢⱼ·uᵢ·uⱼ), ρ = Der Kiureghian= 1.0329 in
  3. Axial elm 1 — CQCN = √(Σᵢ Σⱼ ρᵢⱼ·Nᵢ·Nⱼ)= 279.48 kip

CSI SAP2000 Software Verification, Example 1-025 (Response-Spectrum Analysis of a Three-Dimensional Braced Frame); independent solution = Peterson (1981), reproduced exactly (0 %) by SAP2000. A three-story, L-shaped building framed by four identical two-bay, three-story X-braced planar frames whose columns and diagonals carry axial force only (all pinned). The frames are tied by a rigid floor diaphragm at each level. All mass is concentrated at a centre-of-mass joint per level with X and Y translational mass (1.24224 kip-s²/in) and a rotational mass moment of inertia about Z (174,907.4 kip-in-s²), giving nine dynamic DOF; for consistency with Peterson only the first two modes drive the RSA. The excitation is the El-Centro-derived 5%-damped response spectrum applied in X, combined by CQC, SRSS and ABS. Reference values are the two modal frequencies, the roof centre-of-mass (joint 51) X/Y displacement and Z rotation, and the axial forces in three Frame-1 elements (a column and two diagonals) under each combination rule.

§6

Results comparison

Response quantityProbeSTRIXReferenceΔVerdict
Frequency — mode 1 (Hz) eigen · Hz 3.05908 3.0592 −0.0040% PASS
Frequency — mode 2 (Hz) eigen · Hz 3.11867 3.1188 −0.0040% PASS
Roof CM Ux — CQC (in) CQC · in 1.03241 1.0329 −0.048% PASS
Roof CM Ux — SRSS (in) SRSS · in 0.736835 0.7372 −0.05% PASS
Roof CM Ux — ABS (in) ABS · in 1.04181 1.0423 −0.047% PASS
Roof CM Uy — CQC (in) CQC · in 0.141363 0.1414 −0.026% PASS
Roof CM Uy — SRSS (in) SRSS · in 0.736835 0.7372 −0.05% PASS
Roof CM Uy — ABS (in) ABS · in 1.04181 1.0423 −0.047% PASS
Roof CM Rz — CQC (rad) CQC · rad 2.5173×10⁻⁴ 2.52×10⁻⁴ −0.108% PASS
Roof CM Rz — SRSS (rad) SRSS · rad 2.5173×10⁻⁴ 2.52×10⁻⁴ −0.108% PASS
Roof CM Rz — ABS (rad) ABS · rad 2.5173×10⁻⁴ 2.52×10⁻⁴ −0.108% PASS
Axial elm 1 — CQC (kip) CQC · kip 279.348 279.48 −0.047% PASS
Axial elm 1 — SRSS (kip) SRSS · kip 200.455 200.55 −0.047% PASS
Axial elm 1 — ABS (kip) ABS · kip 281.862 281.99 −0.045% PASS
Axial elm 4 — CQC (kip) CQC · kip 194.409 194.5 −0.047% PASS
Axial elm 4 — SRSS (kip) SRSS · kip 139.505 139.57 −0.047% PASS
Axial elm 4 — ABS (kip) ABS · kip 196.159 196.25 −0.046% PASS
Axial elm 6 — CQC (kip) CQC · kip 120.464 120.52 −0.046% PASS
Axial elm 6 — SRSS (kip) SRSS · kip 86.4431 86.48 −0.043% PASS
Axial elm 6 — ABS (kip) ABS · kip 121.548 121.61 −0.051% PASS
§7

Modal combination

MethodRoof CM Ux (in)Δ
CQC1.03241 −0.048%
SRSS0.736835 −0.05%
ABS1.04181 −0.047%

The reference publishes the roof centre-of-mass displacement under three modal-combination rules, and STRIX reproduces each: CQC and SRSS through the production RsaCalculator (Der Kiureghian CQC), and ABS (absolute sum) from the same per-mode responses. Because the two modes are closely spaced (frequencies within 2 %), CQC (1.0329 in) far exceeds SRSS (0.7372 in) — the cross-correlation adds the modes nearly in phase in X — while ABS (1.0423 in) is the upper bound. All three land within tolerance of the independent reference, verifying that STRIX handles closely-spaced modes correctly (the very case that motivates CQC).

§8

Conclusion

PASS

STRIX reproduces the two modal frequencies, the roof centre-of-mass X/Y/Rz response, and three member axial forces of the CSI 1-025 / Peterson 1981 L-shaped braced frame under all three combination rules to better than a quarter of a percent across twenty independent quantities. Beyond SR2 (moment frame), this verifies three new surfaces at once: the brace / axial-only member in the modal and RSA path (linear Truss), a rotational mass moment of inertia reproduced through the de-Zhu diaphragm mass condensation (offset translational masses → exact Rz inertia), and a tabulated spectrum with closely-spaced modes. Together with SR1, SR2 and SM5/SM5b/SM6 it completes STRIX’s response-spectrum verification across planar, eccentric-moment-frame and braced 3-D configurations.

§9

References & analysis files

  1. CSI SAP2000 Software Verification, Example 1-025 — Response-Spectrum Analysis of a Three-Dimensional Braced Frame.
  2. F. E. Peterson (1981) — independent response-spectrum solution reproduced exactly by SAP2000.
  3. E. L. Wilson, A. Der Kiureghian & E. P. Bayo, EESD 9(2), 1981 — CQC rule. STRIX harness: buildRsaPy → opensees.pyd (de-Zhu T-matrix, M_red = TᵀMT + Transformation) → rsa_modal.bin; MMI via four offset translational masses on the rigid diaphragm.
Engine
v1.0.6 (opensees.pyd)
Run date
2026-08-19
Record
records/SR2b.json
Evidence archive
verif-evidence-eng1.0.6-win-x64.zip · SR2b/
sha256
(pending publish)