diff --git a/CHANGELOG.md b/CHANGELOG.md index e5b5d88..6bf2c0b 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -117,6 +117,12 @@ All notable changes to CableDyn are recorded in this file. The format follows one generic coefficient set and the drag amplification on the in-line drag alone, against the experimenters' measured envelopes and summaries: dominant mode right in 5 of 9 and within one in all, mean in-line displacement within 17 % in 7 of 9, cross-flow amplitude −45 % to +4 %. +- Validation case L3-N (`validation/experiments/ndp_riser`): cross-flow VIV of the NDP 38 m + high-mode riser of Trim et al. (2005) at its 22 uniform-flow tow speeds, with four coefficient + sets against the measured maximum r.m.s. cross-flow strain (Thorsen et al. 2016, Fig. 8), the + dominant frequency and σ/D: the set Ulveseter et al. (2019) fitted to these tests matches strain + (r.m.s. log error 0.178), frequency and amplitude together; Andersen et al.'s (2024b) + sheared-flow average 0.186; the generic set, blind, 0.277. - Theory manual, Published coefficient sets: the VIV coefficient sets of Thorsen et al. (2016, 2017), Ulveseter et al. (2018), Andersen et al. (2024a, b) and Facchinetti et al. (2004), each with its source, its test data and the line-type columns it maps to, and Thorsen et al.'s diff --git a/README.md b/README.md index 9ee384f..5d9a19e 100644 --- a/README.md +++ b/README.md @@ -171,8 +171,9 @@ layout is described in the architecture notes in the repository. ## Validation CableDyn is verified against analytical solutions and compared with MoorDyn, -OrcaFlex, two physical experiments (Holcombe lazy-wave, Bergdahl dynamic chain), -the IEA 15 MW VolturnUS-S reference systems, and coupled OpenFAST runs. Results, +OrcaFlex, six published physical experiments (Holcombe lazy-wave, Bergdahl dynamic +chain, Chaplin riser natural frequencies and stepped-current VIV, Wang catenary-riser VIV, +NDP high-mode riser VIV), the IEA 15 MW VolturnUS-S reference systems, and coupled OpenFAST runs. Results, acceptance criteria, and reproduction commands are in the [validation record](https://cabledyn.readthedocs.io/en/latest/validation.html); data and scripts are in [`validation/`](https://cabledyn.readthedocs.io/en/latest/validation.html). diff --git a/VALIDATION.md b/VALIDATION.md index 9d8aa6e..3b523da 100644 --- a/VALIDATION.md +++ b/VALIDATION.md @@ -84,6 +84,7 @@ article's methods, are compared in | **L3-C** | Chaplin Delta Flume riser, still-water natural frequencies | measured modes 1–7 at five tensions | mean error +1.87%, mean absolute 2.28%, largest 4.88% (34 of 35; one outlier reported) | full published table, no fitted parameter | Pass | | **L3-D** | Chaplin Delta Flume riser in a stepped current, blind-trial cases 1–9, cross-flow and in-line VIV | measured dominant mode, peak cross-flow amplitude, mean in-line displacement, curvature standard deviations | mode right in 5/9, within one in 9/9; mean in-line within ±17% in 7/9 (−25% to +34%); cross-flow amplitude −45% to +4%; cross-flow curvature −39% to +26% | generic coefficient set, no fitted parameter | Pass measured | | **L3-W** | SJTU steel catenary riser in 0.2 m/s current, cross-flow VIV | measured r.m.s. strain (25 stations) and dominant frequency | frequency 0.992 vs 1.00 Hz; max r.m.s. strain +7.3%, station mean +3.6% | published coefficient set, no fitted parameter | Pass | +| **L3-N** | NDP 38 m high-mode riser in uniform flow, 0.3–2.4 m/s (22 tests), cross-flow VIV | measured maximum r.m.s. cross-flow strain, dominant frequency, spatial-mean σ/D | strain r.m.s. log error: Ulveseter et al. (2019) set 0.178, Andersen et al. (2024b) 0.186, generic set 0.277, Thorsen et al. (2016) 0.363 (their own simulation 0.187); f D/U 0.145 / 0.134 / 0.192 / 0.193 vs 0.143 measured | generic set blind; the other sets fitted by their authors to NDP tests (implementation check) | Pass measured | | **L-fig** | Along-arc curvature overlay (CableDyn / OrcaFlex / MoorDyn-F ±4×) | 3-code committed record | CableDyn peak 0.11% from OrcaFlex | <0.5% | Pass | | **L4-1** | Coupled platform + mooring | synthetic reference | bit-for-bit | — | Pass | | **L4-1b** | VolturnUS-S coupled platform + power cable | OpenFAST v5.0.0 (CompMooring 5 vs 3) | platform DOFs <0.6% (3-seed DLC-1.1); fairlead and anchor <0.001% static | <3% platform | Pass | @@ -373,6 +374,32 @@ scripts, results, and source provenance are in the test, as theirs, so the station-by-station ratio scatters. 400 elements and half the time step change the largest r.m.s. strain by 0.2% and 0.1%. +- **L3-N NDP high-mode riser, cross-flow VIV in uniform flow** + ([Trim et al., 2005](https://cabledyn.readthedocs.io/en/latest/references.html#ref-trim2005); + [Thorsen et al., 2016](https://cabledyn.readthedocs.io/en/latest/references.html#ref-thorsen2016)): + the 38 m bare riser towed horizontally at the 22 uniform-flow speeds, 0.3–2.4 m/s, with the + rig's clump-weight tension as an axial spring in series (4 kN in still water, 3.5 kN/m from the + reported 4–6 kN range). Four coefficient sets on the same model, compared with the maximum + r.m.s. cross-flow strain of each test (Thorsen et al.'s Fig. 8, read from the vector figure), the + dominant frequency and the spatial-mean σ/D (Trim et al.'s Figs. 18 and 16, read by eye from + raster figures): + - the set Ulveseter et al. (2019) fitted to these tests matches every metric: strain r.m.s. log + error 0.178 (Thorsen et al.'s own fitted simulation 0.187), f D/U 0.145 against 0.143 + measured, σ/D within 6% cross-flow and 3% in line, an implementation check of the + synchronisation, in-line and higher-harmonic forces; + - the cross-flow average Andersen et al. (2024b) fitted to the sheared-flow tests gives 0.186 at + f D/U 0.134, with σ/D 57% high; + - the generic set, blind, gives 0.277 with σ/D within 2%, at f D/U 0.192; + - Thorsen et al.'s own fitted curve and damping give 0.363 at f D/U 0.193; keeping Morison's + linear cross-flow damping moves the frequency to 0.153. + + At single speeds the response converges at 1.0 m/s for four of the five sets (half the time + step and 400 elements move the strain by at most 0.8%; 18.5% for the fifth) but not at 2.4 m/s, + where the mix of neighbouring high modes moves the strain by up to 20%. The statistics over + the 22 speeds hold: the full sweep at half the time step gives r.m.s. log errors 0.280, 0.187, + 0.184 and 0.375 for the generic, Andersen, Ulveseter and Thorsen sets, f D/U within 0.001 and + σ/D within 1.4%. + ## L4 coupled OpenFAST - **L4-1 synthetic** (`l4_platform_mooring`): 6-DOF platform moored by three taut diff --git a/doc/references.rst b/doc/references.rst index 9850119..74e5fa9 100644 --- a/doc/references.rst +++ b/doc/references.rst @@ -294,6 +294,20 @@ Verification and validation references laboratory measurements of vortex-induced vibrations of a tension riser. *Journal of Fluids and Structures* **21**\ (1), 25–40. https://doi.org/10.1016/j.jfluidstructs.2005.05.016 +.. _ref-trim2005: + +- **Trim, A. D., Braaten, H., Lie, H. & Tognarelli, M. A. (2005).** Experimental investigation + of vortex-induced vibration of long marine risers. *Journal of Fluids and Structures* + **21**\ (3), 335–361. https://doi.org/10.1016/j.jfluidstructs.2005.07.014 + +.. _ref-fontaine2011: + +- **Fontaine, E., Marcollo, H., Vandiver, K., Triantafyllou, M., Larsen, C., Tognarelli, M., + Constantinides, Y. & Oakley, O. (2011).** Reliability based factors of safety for VIV fatigue + using NDP riser high mode VIV tests. *ASME 2011 30th International Conference on Ocean, + Offshore and Arctic Engineering (OMAE2011)*, Volume 7: CFD and VIV; Offshore Geotechnics, + 511–522. https://doi.org/10.1115/OMAE2011-49820 + Comparison codes ---------------- diff --git a/validation/README.md b/validation/README.md index b8612bf..1e0d17d 100644 --- a/validation/README.md +++ b/validation/README.md @@ -16,7 +16,7 @@ redistributed. | [`RELEASE_0_1_1.md`](RELEASE_0_1_1.md) | The same for v0.1.1 | | [`RELEASE_0_1_0.md`](RELEASE_0_1_0.md) | The same for v0.1.0 | | [`bodies/`](bodies/README.md) | Rigid-body, rod and shared-anchor cases with MoorDyn-C twins, stored references and pass limits | -| [`experiments/`](experiments/README.md) | Holcombe et al. (2025) static lazy-wave, Bergdahl et al. (2016) dynamic-chain, Chaplin et al. (2005) riser natural-frequency and stepped-current VIV, and Wang et al. (2014) catenary-riser VIV comparisons | +| [`experiments/`](experiments/README.md) | Holcombe et al. (2025) static lazy-wave, Bergdahl et al. (2016) dynamic-chain, Chaplin et al. (2005) riser natural-frequency and stepped-current VIV, Wang et al. (2014) catenary-riser VIV and Trim et al. (2005) NDP high-mode riser VIV comparisons | | [`PERFORMANCE_0_1_0.md`](PERFORMANCE_0_1_0.md) | Wall time of CableDyn 0.1.0, MoorDyn-C and MoorDyn-F at equal accuracy on five mooring, cable, buoy and coupled cases | | [`performance/`](performance/README.md) | Same-host wall-clock records of the L5 cases: the resolved lazy-wave spans, the one-hour VolturnUS-S mooring and the tension-only chain | | [`PAPER_REPRODUCTION.md`](PAPER_REPRODUCTION.md) | Every CableDyn result in the journal article, re-run with the release defaults and in paper mode | diff --git a/validation/experiments/README.md b/validation/experiments/README.md index 0aa20eb..71f4091 100644 --- a/validation/experiments/README.md +++ b/validation/experiments/README.md @@ -2,7 +2,7 @@ # Experimental comparisons -This directory lets reviewers reproduce CableDyn's comparisons with five published physical +This directory lets reviewers reproduce CableDyn's comparisons with six published physical experiments. Each subdirectory gives the data source, the translation of the published parameters into a CableDyn model, the assessed quantity, the reproduction commands and the committed results. @@ -14,5 +14,6 @@ committed results. | [`chaplin/`](chaplin/README.md) | Chaplin et al. (2005), 13.12 m vertical tension riser, still-water natural frequencies | Finite-EI riser, modal analysis | | [`delta_flume/`](delta_flume/README.md) | Chaplin et al. (2005), 13.12 m vertical tension riser, stepped current, blind-trial cases 1–9 | Finite-EI riser, synchronisation VIV, cross-flow and in-line | | [`wang_scr/`](wang_scr/README.md) | Wang et al. (2014), truncated steel catenary riser in uniform current, cross-flow VIV | Finite-EI riser, synchronisation VIV | +| [`ndp_riser/`](ndp_riser/README.md) | Trim et al. (2005), NDP 38 m high-mode riser in uniform flow, cross-flow VIV with a generic and three published coefficient sets | Finite-EI riser, synchronisation VIV | -The results are summarised as cases L3-H, L3-B, L3-C, L3-D and L3-W in [`VALIDATION.md`](../../VALIDATION.md). +The results are summarised as cases L3-H, L3-B, L3-C, L3-D, L3-W and L3-N in [`VALIDATION.md`](../../VALIDATION.md). diff --git a/validation/experiments/ndp_riser/README.md b/validation/experiments/ndp_riser/README.md new file mode 100644 index 0000000..515d0a1 --- /dev/null +++ b/validation/experiments/ndp_riser/README.md @@ -0,0 +1,188 @@ + + +# NDP 38 m riser in uniform flow: high-mode cross-flow VIV + +This directory compares CableDyn's VIV with the bare-riser, uniform-flow tests of the Norwegian +Deepwater Programme's high-mode riser: Trim et al. (2005), +. A 38 m riser of 27 mm diameter was towed +horizontally at 22 speeds from 0.3 to 2.4 m/s and responded at up to 12.5 Hz cross-flow. The +same riser model runs with four coefficient sets. One is generic and not fitted to these tests. +The other three were fitted by their authors to NDP data, two of them to these very tests, so +for those the comparison checks CableDyn's implementation of the model, not its predictive +power. + +## Model and data + +- **Riser (Fontaine et al. 2011, Table 1; Thorsen et al. 2016, Table 1):** 38.00 m between + universal joints (pinned), D 27 mm, EI 598.8 N m², EA 8.19e6 N, 0.933 kg/m flooded, as tested. +- **Water:** fresh, 1000 kg/m³. The mass ratio 1.62 of Fontaine et al. is 0.933 kg/m over the + displaced water at 1006 kg/m³; 1000 kg/m³ gives 1.63 and sea water 1.59. +- **Tension (Trim et al. 2005, Sec. 2.1):** clump weights on inclined pendulum arms held the + riser at about 4–6 kN over the test speeds. The tension of each test is not published + (Ulveseter et al. 2019 give the range 4067–6285 N). The rig is modelled as an axial spring + k_ax in series with the riser, EA_eff = L/(1/k_ax + L/EA), with T0 = 4 kN in still water. + - k_ax = 3.5 kN/m comes from the reported range, not from the response: at 2.4 m/s the Morison + drag (C_D 1.2) of about 93 N/m sags a 6 kN string by wL²/(8T) ≈ 2.8 m, which takes + 8s²/(3L) ≈ 0.55 m more length, and 2 kN over 0.55 m is about 3.6 kN/m. + - The resulting mean tension of each set is in the table below. +- **Assumed:** the riser 5 m below the surface (not given; the current is uniform), in 10 m of + water. +- **Current:** uniform, normal to the riser and horizontal, so cross-flow (CF) is vertical and + in-line (IL) horizontal. +- **CableDyn:** one finite-EI line, 200 elements (0.19 m), dt 1 ms, 25 s, statistics over + t > 10 s (frequency resolution 0.067 Hz); `VIVrmsWindow` 0.5 s. +- **Analysis (`ndp_riser_viv.py`):** + - `CurvX` at every node is the CF curvature and `CurvY` the IL one: the section axis x starts + at End A as global +z projected normal to the line and is carried along it, which keeps it + within 2° of vertical on the bowed riser. Strain = κ D/2. + - The dominant CF frequency is the peak of the CF curvature spectra summed over the nodes, + searched for in f D/U = 0.05–0.30. This covers the synchronisation band of every set + (0.065–0.26) and excludes a slow swing of the riser below it and the third harmonic of the + Ulveseter et al. set above it. The IL frequency is searched for in 0.10–0.60. + - The fundamental CF strain is band-passed to 0.5–1.5 times the dominant CF frequency, as + Thorsen et al. (2016) filter the measured strain. The compared value is its largest r.m.s. + along the riser. + - σ/D is the spatial mean, over every fourth node, of the standard deviation of the + displacement band-passed in the same way, as Trim et al. reconstruct the vibration from + modes. + +### Coefficient sets + +`--sets` selects them. C_a = 1 (C_M = 2) in all. + +| Set | Source | Values | Fitted to | +| --- | --- | --- | --- | +| `b1` | The generic set of the [Delta Flume case](../delta_flume/README.md) | C_l 0.8, f̂0 0.18 ± 0.08, C_D 1.2; in-line C_x 0.85, α 0.15 (Ulveseter et al. 2018, Table 3); `VIV AMPLITUDE` C_v 0.8 with the drag amplification of Chaplin et al. (2005, Eq. 4); `VIV DAMPING` `0 0 1` (the amplification on the in-line drag only) | nothing from the NDP tests: blind | +| `and24b` | Andersen et al. (2024b), Table 3, cross-flow average | C_v,y 1.28, f̂0,y 0.131, Δf̂y 0.066, C_D 1.16; Morison drag | the 22 sheared-flow tests of this riser: semi-blind here. Their in-line set (C_v,x 1.09, f̂0,x 0.321, Δf̂x 0.161) belongs to an independent in-line synchronisation, not CableDyn's `VIV` row, and is left out | +| `uls19` | Ulveseter et al. (2019), Table 1 and Sec. 4.1 | C_D 1.2, C_v,y 0.85, f̂0 0.144 ± 0.064, in-line C_v,x 0.75, α 0.15, C_hh 1.5; Morison drag | these 22 uniform-flow tests ("a visual best fit with cross-flow strain measurements"): an implementation check | +| `th16` | Thorsen et al. (2016), Fig. 7 and Eq. 1 | `VIV AMPLITUDE` with their fitted C_v(y0/D) (`thorsen2016_fig7_cv.csv`); `VIV DAMPING` `0.31 0.89 0`, their cross-flow damping without Morison's linear term; band 0.18 ± 0.08 for the 0.10–0.26 range of their Fig. 4; C_D 1.2 in line | these 22 uniform and the 22 sheared-flow tests: an implementation check. Their model is cross-flow only | +| `th16c` | `th16` with Morison's linear cross-flow damping kept (`VIV DAMPING` `0.31 0.89 1`) | as `th16` | a variant, not a published set | + +The published values were checked against the source articles. CableDyn's synchronisation +frequency is f̂0 + Δf̂ sin(...), so `th16` matches the range of Thorsen et al.'s synchronisation +function (their Fig. 4), not its shape. + +### Measurements + +- `thorsen2016_fig8_strain.csv`: the maximum r.m.s. CF strain of each test, at the fundamental + frequency and in total, with Thorsen et al.'s own simulation using their fitted curve (their + Fig. 8; the uniform panel is compared, the sheared one is kept). The figure is vector + graphics: `extract_thorsen2016.py` maps the plotted coordinates through the plot frames, whose + ends are the tick labels 0 and 2.5 m/s and 0 and 5e-4. Each series has 22 points, one per tested + speed. `thorsen2016_fig7_cv.csv` is their fitted C_v curve from the same page (0.35 at 0, a + maximum of 1.00 at y0/D 0.46, zero from 0.67). +- `measured_trim2005_fig16.csv`: the spatial mean of σ/D, CF and IL, at all 22 speeds (Trim et + al. 2005, Fig. 16). `measured_trim2005_fig18.csv`: the peak frequency of the dominant + displacement mode, CF and IL, at five speeds (Fig. 18). The figures are raster images and their + markers overlap, so the values were read by eye on four-times enlarged crops with a guide line + at every speed and every 0.01 or 1 Hz (`digitise_trim2005.py` calibrates the axes on the + panel frames and grid lines and draws the guides). Uncertainty: ±0.01 in σ/D, about ±0.02 where + markers of neighbouring speeds overlap; ±0.3 Hz, ±0.5 Hz for CF at 2.0 m/s. The measured CF + frequency follows f D/U ≈ 0.14 and the IL one twice that. + +The two extraction scripts need the articles (not distributed), PyMuPDF and Pillow; the +comparison needs only the CSV files. + +## Reproduce + +From the repository root, with `path/to/cabledyn.exe` the release executable or a local build: + +```powershell +python validation/experiments/ndp_riser/ndp_riser_viv.py ` + path/to/cabledyn.exe validation/experiments/ndp_riser/work +``` + +The default runs the five sets at 0.5, 1.0, 1.5, 2.0 and 2.4 m/s, 25 runs in parallel, in +about two minutes on a 24-core workstation, and prints the statistics over those speeds; each +run reproduces its row of the record. The committed record is the full sweep, 110 runs in about +twelve minutes with 24 parallel runs: + +```powershell +python validation/experiments/ndp_riser/ndp_riser_viv.py ` + path/to/cabledyn.exe validation/experiments/ndp_riser/work ` + --speeds all --jobs 24 --summary validation/experiments/ndp_riser +``` + +`--sets` and `--speeds` take comma lists (quote them in PowerShell, `--speeds "1.0,2.4"`); +`--segments`, `--dt`, `--tension` and `--kax` change the mesh, the time step and the rig. A run +whose folder holds the same deck, executable and statistics window is not repeated; `--keep` +keeps the output tables (about 150 MB a run) so that the analysis can be redone without them. + +## Results + +All 22 speeds; the strain ratio is CableDyn's fundamental over the measured fundamental, and +the r.m.s. log error is √(mean (ln ratio)²). + +| Set | Fitted to these tests | Strain ratio, median (range) | r.m.s. log error | U ≥ 0.5 m/s | f D/U, CF | σ/D, CF | σ/D, IL | Mean tension (kN) | +| --- | --- | --- | --- | --- | --- | --- | --- | --- | +| Measured | | 1 | 0 | 0 | 0.143 | 0.436 | 0.149 | 4–6 | +| `b1` | no | 1.26 (0.99–1.57) | 0.277 | 0.277 | 0.192 | 0.429 | 0.041 | 4.0–10.0 | +| `and24b` | sheared tests | 1.00 (0.67–1.27) | 0.186 | 0.181 | 0.134 | 0.687 | 0.088 | 4.0–7.5 | +| `uls19` | yes | 0.89 (0.64–1.33) | **0.178** | **0.156** | **0.145** | 0.461 | **0.146** | 4.0–6.9 | +| `th16` | yes | 1.40 (0.91–1.86) | 0.363 | 0.377 | 0.193 | 0.390 | 0.067 | 4.0–6.5 | +| `th16c` | variant | 0.73 (0.51–1.06) | 0.365 | 0.376 | 0.153 | 0.307 | 0.025 | 4.0–6.4 | +| Thorsen et al.'s own simulation (their Fig. 8a) | yes | 1.12 (0.75–1.51) | 0.187 | 0.182 | | | | | + +f D/U is the mean at the five speeds of Fig. 18; σ/D the mean over the 22 speeds. +`response_comparison.csv` holds every run, `summary.json` the statistics and the SHA-256 of the +executable. + +- **Implementation check:** with the set Ulveseter et al. fitted to these tests, CableDyn's + three-dimensional model matches every metric: the strain with an r.m.s. log error of 0.178 + (0.184 at half the time step), comparable to Thorsen et al.'s own cross-flow-only fit (0.187), f + D/U 0.145 against the measured 0.143, and σ/D within 6 % cross-flow and 3 % in line, with the + higher-harmonic force (C_hh 1.5) of their set. +- **Semi-blind:** the cross-flow set Andersen et al. fitted to the sheared-flow tests predicts the + uniform-flow strain as well as Thorsen et al.'s fit (0.186 against 0.187) at f D/U 0.134, but + with σ/D 57 % high: a lower mode at a larger amplitude gives the same curvature. +- **Blind:** the generic set gives σ/D within 2 % and the strain 26 % high in the median (r.m.s. + log error 0.277), and responds at f D/U 0.19, a third above the measured frequency. Its + amplified drag raises the mean tension to 10.0 kN at 2.4 m/s, above the reported 6 kN. +- **The damping form sets the frequency:** with Thorsen et al.'s curve and band, dropping + Morison's linear cross-flow damping (`th16`, as they do) gives f D/U 0.193; keeping it + (`th16c`) 0.153, consistent with the higher modes being less damped without it. `th16` also + develops a slow cross-flow swing, at 0.6–0.8 Hz up to 1.9 m/s (raw σ/D up to 2.0, excluded from + the band-passed values), the known consequence of dropping the linear term while an in-line + force acts (theory guide, Cross-flow damping). Neither reproduces the strain of their own fit + (0.363 and 0.365 against 0.187): their curve was fitted in a cross-flow-only model run at the + tension of each test, and both differ here. +- **One metric does not discriminate:** `and24b` matches the strain with the wrong amplitude, + `b1` the amplitude with the wrong frequency, `th16c` the frequency with the wrong strain. Only + `uls19` matches strain, amplitude and frequency together. +- **In-line:** `b1` and `uls19` also swing slowly in line at 0.3–0.9 Hz, more at the higher + speeds (raw IL σ/D up to 3.1 and 2.6). The band-passed values exclude the swing; the measured + σ/D, a modal reconstruction of the vibration, could not show it, and whether the tests had + one is not reported. +- **Sensitivity to the water density:** with sea water (1025 kg/m³) the r.m.s. log errors from + 0.5 m/s are 0.299 (`b1`), 0.181 (`and24b`), 0.154 (`uls19`), 0.434 (`th16`) and 0.353 + (`th16c`). The change of `th16` (0.377 to 0.434) shows how much the mode chosen at single + speeds moves the statistic. + +### Convergence + +- **Single speeds** (1.0 and 2.4 m/s, every set; `provenance.json`): + - at 1.0 m/s, half the time step and 400 elements change the fundamental strain by at most + 0.8 % and σ/D by at most 3 % for `b1`, `and24b`, `th16` and `th16c`, at the same frequency + within one bin; `uls19` moves by +1.9 % and +18.5 % in strain and responds at 5.60 Hz instead + of 5.80 Hz with both (σ/D −2 %); + - at 2.4 m/s the response is not converged speed by speed: the strain moves by −14 % to +12 % + with half the time step and by +0.6 % to +20 % with 400 elements, and the dominant frequency + moves by more than one bin in 7 of the 10 pairs; + - 55 s runs (45 s of statistics) of `b1`, `and24b` and `th16c` at 2.4 m/s change the strain of + the 25 s runs by −0.3 %, −1.2 % and +12.5 %, and at 55 s half the time step and 400 elements + still move it by −0.7 % to +12 % and −7.8 % to +21 %: a longer window does not remove the + spread. It comes from the mix of neighbouring high modes the response settles in, which a + small change of the model moves; it is of the size of the speed-to-speed scatter of the + ratios. +- **The statistics over the 22 speeds hold:** the full sweep at half the time step gives + r.m.s. log errors of 0.280 (`b1`), 0.187 (`and24b`), 0.184 (`uls19`), 0.375 (`th16`) and + 0.376 (`th16c`), f D/U within 0.001 and σ/D within 1.4 % of the record. Differences of a few + hundredths in the r.m.s. log error between sets are within this spread. + +## Limitations + +- The tension of each test is not published; the rig spring gives the reported range, but each + run's tension differs from the test's by an unknown amount, up to the 10 kN of `b1`. +- The σ/D and frequency data are read by eye from raster figures. +- The 3D model carries in-line drag and slow swings that the 2D cross-flow model of Thorsen et + al. does not, so their fitted curve is not expected to transfer exactly. diff --git a/validation/experiments/ndp_riser/digitise_trim2005.py b/validation/experiments/ndp_riser/digitise_trim2005.py new file mode 100644 index 0000000..97b86e3 --- /dev/null +++ b/validation/experiments/ndp_riser/digitise_trim2005.py @@ -0,0 +1,133 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: Apache-2.0 +# Copyright (c) 2026 Jae Hoon Seo, SMI Lab, Inha University +"""Reading aids for the bare-riser, uniform-flow panels of Trim et al. (2005), Figs. 16 and 18. + +Both figures are raster images embedded in the article (962 px wide). The panel 'bare riser, +uniform flow' is the top-left one of each: + +- Fig. 16 (article page 21): the spatial mean of the temporal standard deviation of the + displacement over the diameter, sigma/D, cross-flow (blue diamonds) and in-line (red stars); +- Fig. 18 (page 23): the peak frequency of the dominant displacement mode [Hz]. + +The markers (about 10 px) are wider than the speed step (0.1 m/s, about 5.7 px) and overlap, so +an automatic marker search does not separate them. This script calibrates the axes on the +panel frame and its dotted grid lines (x: the 0.5 m/s grid lines and the frame's right edge at +2.5 m/s; y: the frame bottom at zero and the horizontal grid lines) and writes a four-times +enlarged crop of each panel with a guide line at every tested speed and every 0.01 (Fig. 16) or +1 Hz (Fig. 18). measured_trim2005_fig16.csv and measured_trim2005_fig18.csv were read by eye on +these crops: the centre of the marker on each speed's guide. + + python digitise_trim2005.py path/to/article.pdf [--out DIR] + +writes DIR/trim2005_fig16_guides.png and DIR/trim2005_fig18_guides.png and prints the axis +calibration. Needs PyMuPDF and Pillow. + +Source: A.D. Trim, H. Braaten, H. Lie, M.A. Tognarelli (2005), Experimental investigation of +vortex-induced vibration of long marine risers, Journal of Fluids and Structures 21(3), 335-361, +https://doi.org/10.1016/j.jfluidstructs.2005.07.014 +""" + +from __future__ import annotations + +import argparse +from pathlib import Path + +import numpy as np + +HERE = Path(__file__).resolve().parent +# figure: (zero-based page, search region x0 y0 x1 y1 [px] around the panel frame, y tick values +# from the frame bottom up, guide step in y) +FIGS = { + "fig16": (20, (100, 60, 262, 345), [0.0, 0.1, 0.2, 0.3, 0.4, 0.5], 0.01), + "fig18": (22, (60, 60, 250, 345), [0.0, 10.0, 20.0, 30.0], 1.0), +} +SPEEDS = np.round(np.arange(0.3, 2.45, 0.1), 2) # the 22 tested speeds [m/s] + + +def frame(rgb: np.ndarray, reg) -> tuple[int, int, int, int]: + """The panel frame: the dark columns and rows that span most of the region.""" + x0, y0, x1, y1 = reg + dark = rgb[y0:y1, x0:x1].sum(axis=2) < 250 + cols = np.where(dark.sum(axis=0) > 0.45 * dark.shape[0])[0] + rows = np.where(dark.sum(axis=1) > 0.3 * dark.shape[1])[0] + return ( + int(x0 + cols.min()), + int(y0 + rows.min()), + int(x0 + cols.max()), + int(y0 + rows.max()), + ) + + +def grid(rgb: np.ndarray, box, axis: int) -> list[float]: + """The dotted grey grid lines inside the frame; axis 0: vertical lines, 1: horizontal.""" + x0, y0, x1, y1 = box + sub = rgb[y0 + 3 : y1 - 2, x0 + 3 : x1 - 2].astype(int) + grey = (sub.sum(axis=2) < 450) & (np.ptp(sub, axis=2) < 60) + prof = grey.sum(axis=axis) + groups: list[list[int]] = [] + for i in np.where(prof > 0.12 * grey.shape[axis])[0]: + if groups and i - groups[-1][-1] <= 2: + groups[-1].append(int(i)) + else: + groups.append([int(i)]) + off = (x0 if axis == 0 else y0) + 3 + return [float(np.mean(g)) + off for g in groups] + + +def main() -> None: + ap = argparse.ArgumentParser(description=__doc__.split("\n\n")[0]) + ap.add_argument("pdf") + ap.add_argument("--out", default=str(HERE)) + a = ap.parse_args() + import pymupdf + from PIL import Image, ImageDraw + + doc = pymupdf.open(a.pdf) + out = Path(a.out) + out.mkdir(parents=True, exist_ok=True) + for fig, (page, reg, yt, step) in FIGS.items(): + pix = pymupdf.Pixmap(doc, doc[page].get_images(full=True)[0][0]) + if pix.n - pix.alpha > 3: + pix = pymupdf.Pixmap(pymupdf.csRGB, pix) + img = Image.frombytes("RGB", (pix.width, pix.height), pix.samples) + rgb = np.asarray(img).astype(int) + box = frame(rgb, reg) + gx = grid(rgb, box, 0) + gy = grid(rgb, box, 1) + # the vertical grid lines are at multiples of 0.5 m/s; the right frame edge is 2.5 m/s + sp = np.median(np.diff(gx)) + gx = [g for g in gx if g < box[2] - 3] + [float(box[2])] + ticks = [2.5 - 0.5 * round((box[2] - g) / sp) for g in gx] + ax, bx = np.polyfit(ticks, gx, 1) # pixel x = ax U + bx + ypix = [box[3]] + sorted((y for y in gy if y < box[3] - 5), reverse=True) + n = min(len(ypix), len(yt)) + ay, by = np.polyfit(yt[:n], ypix[:n], 1) # pixel y = ay value + by + print( + f"{fig}: frame {box}; x ticks {ticks} at {[round(g, 1) for g in gx]} px " + f"({ax:.2f} px per m/s); y ticks {yt[:n]} at {[round(y, 1) for y in ypix[:n]]} px " + f"({-ay:.2f} px per unit)" + ) + x0, y0, x1, y1 = box[0] - 7, box[1] - 10, box[2] + 7, box[3] + 8 + k = 4 + crop = img.crop((x0, y0, x1, y1)).resize( + ((x1 - x0) * k, (y1 - y0) * k), Image.NEAREST + ) + dr = ImageDraw.Draw(crop) + for u in SPEEDS: + px = (ax * u + bx - x0) * k + dr.line([(px, 0), (px, crop.height)], fill=(0, 190, 0), width=1) + dr.text((px + 1, 2), f"{u:.1f}", fill=(0, 0, 0)) + for j in range(int(round(yt[-1] / step)) + 1): + py = (ay * j * step + by - y0) * k + major = j % 10 == 0 if step < 0.1 else j % 5 == 0 + dr.line( + [(0, py), (crop.width, py)], + fill=(200, 100, 0) if major else (255, 190, 0), + width=1, + ) + crop.save(out / f"trim2005_{fig}_guides.png") + + +if __name__ == "__main__": + main() diff --git a/validation/experiments/ndp_riser/extract_thorsen2016.py b/validation/experiments/ndp_riser/extract_thorsen2016.py new file mode 100644 index 0000000..3e0327c --- /dev/null +++ b/validation/experiments/ndp_riser/extract_thorsen2016.py @@ -0,0 +1,157 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: Apache-2.0 +# Copyright (c) 2026 Jae Hoon Seo, SMI Lab, Inha University +"""Read Figs. 7 and 8 of Thorsen et al. (2016) from the vector graphics of the article's page 8. + +Fig. 7 is their excitation curve C_v(y0/D), fitted to the NDP riser tests; Fig. 8 the maximum +r.m.s. cross-flow strain of the 22 uniform-flow (a) and 22 sheared-flow (b) tests, measured at +the fundamental frequency (blue squares) and in total (red circles), with their simulation using +the fitted curve (blue line). The figures are vector paths, so the values are the plotted +coordinates mapped through the axes, not a pixel digitisation. The tick labels are drawn glyphs, +not text, so the axes come from the plot frames: + +- Fig. 7: y0/D = 0 and 1 at x = 191.7 and 381.1 pt (the frame, where the curve starts and + ends), C_v = 0 and 1.5 at y = 161.25 and 59.7 pt (the bottom and top ticks); +- Fig. 8: x from 0 to 2.5 m/s across the frame, y from 0 to 5e-4 (a) and 0 to 3.5e-4 (b) up it, + the end tick labels of each axis on the rendered page. + +The legend symbols, in the upper left of each panel, are dropped. + + python extract_thorsen2016.py path/to/article.pdf [--out DIR] + +writes DIR/thorsen2016_fig7_cv.csv (the curve sampled every 0.05 in y0/D) and +DIR/thorsen2016_fig8_strain.csv (panel, series, U, strain in units of 1e-4). Needs PyMuPDF. + +Source: M.J. Thorsen, S. Saevik, C.M. Larsen (2016), Time domain simulation of vortex-induced +vibrations in stationary and oscillating flows, Journal of Fluids and Structures 61, 1-19, +https://doi.org/10.1016/j.jfluidstructs.2015.11.006 +""" + +from __future__ import annotations + +import argparse +import csv +from collections import defaultdict +from pathlib import Path + +import numpy as np + +HERE = Path(__file__).resolve().parent +PAGE = 7 # zero-based: the article's page 8 holds Figs. 7 and 8 + + +def fig7(draws) -> np.ndarray: + """The C_v curve: the polyline with the most vertices above y = 175 pt.""" + + def fx(px: float) -> float: + return (px - 191.7) / (381.1 - 191.7) + + def fy(py: float) -> float: + return 1.5 * (161.25 - py) / (161.25 - 59.7) + + best: list[tuple[float, float]] = [] + for d in draws: + if d["rect"].y1 > 175 or d.get("color") is None or d.get("width") is None: + continue + pts: list[tuple[float, float]] = [] + for it in d["items"]: + if it[0] == "l": + if not pts: + pts.append((it[1].x, it[1].y)) + pts.append((it[2].x, it[2].y)) + if len(pts) > 20 and len(pts) > len(best): + best = pts + arr = np.array([(fx(x), fy(y)) for x, y in best]) + return arr[np.argsort(arr[:, 0])] + + +def fig8(draws) -> list[tuple[str, str, float, float]]: + # the white plot frames from top to bottom: Fig. 7, Fig. 8(a), Fig. 8(b) + boxes = sorted( + { + ( + round(d["rect"].x0, 1), + round(d["rect"].y0, 1), + round(d["rect"].x1, 1), + round(d["rect"].y1, 1), + ) + for d in draws + if d.get("fill") == (1.0, 1.0, 1.0) and d["rect"].width > 100 + }, + key=lambda b: b[1], + ) + panels = {"uniform": (boxes[1], 5.0), "sheared": (boxes[2], 3.5)} + rows = [] + for name, (box, ymax) in panels.items(): + x0, y0, x1, y1 = box + ax, bx = 2.5 / (x1 - x0), -2.5 * x0 / (x1 - x0) + ay, by = -ymax / (y1 - y0), ymax * y1 / (y1 - y0) + print(f"Fig. 8 {name}: frame {box}, y 0..{ymax}e-4") + series = defaultdict(list) + for d in draws: + r = d["rect"] + col = d.get("color") + if col is None or not (y0 - 5 < r.y0 < y1 + 5 and x0 - 5 < r.x0 < x1 + 5): + continue + cx, cy = 0.5 * (r.x0 + r.x1), 0.5 * (r.y0 + r.y1) + kinds = {it[0] for it in d["items"]} + if 2 < r.width < 8 and 2 < r.height < 8: + if col[2] > 0.8 and col[0] < 0.3 and "re" in kinds: + series["exp_fundamental"].append((ax * cx + bx, ay * cy + by)) + elif col[0] > 0.8 and col[2] < 0.3 and "c" in kinds: + series["exp_total"].append((ax * cx + bx, ay * cy + by)) + elif col[2] > 0.8 and col[0] < 0.3 and "l" in kinds and r.width > 20: + for it in d["items"]: + if it[0] == "l": + for p in (it[1], it[2]): + series["sim"].append((ax * p.x + bx, ay * p.y + by)) + for s, pts in series.items(): + # drop the legend symbols (upper left, above every data point there) + pts = [(u, v) for u, v in pts if not (u < 1.35 and v > 0.5 * ymax)] + pts = sorted({(round(u, 3), round(v, 4)) for u, v in pts}) + if s == "sim": # polyline vertices, one per test speed + pts = sorted({round(u, 1): v for u, v in pts}.items()) + rows += [(name, s, u, v) for u, v in pts] + print(f" {s}: {len(pts)} points") + return rows + + +def main() -> None: + ap = argparse.ArgumentParser(description=__doc__.split("\n\n")[0]) + ap.add_argument("pdf") + ap.add_argument("--out", default=str(HERE)) + a = ap.parse_args() + import pymupdf # only this script needs it + + draws = pymupdf.open(a.pdf)[PAGE].get_drawings() + out = Path(a.out) + out.mkdir(parents=True, exist_ok=True) + + cv = fig7(draws) + i = int(np.argmax(cv[:, 1])) + zero = cv[cv[:, 1] <= 1e-3][:, 0] + print( + f"Fig. 7: {len(cv)} vertices; Cv(0) {cv[0, 1]:.4f}, maximum {cv[i, 1]:.4f} at y0/D " + f"{cv[i, 0]:.4f}, zero from y0/D {zero.min():.4f}" + ) + grid = np.arange(0.0, 1.0001, 0.05) + table = np.clip(np.interp(grid, cv[:, 0], cv[:, 1], right=0.0), 0.0, None) + # the curve runs on along the axis, a line width above it, from its zero crossing: the + # fitted curve is zero there (their Sec. 3: C_v is positive or zero) + table[grid >= zero.min()] = 0.0 + with open(out / "thorsen2016_fig7_cv.csv", "w", newline="", encoding="utf-8") as fh: + w = csv.writer(fh, lineterminator="\n") + w.writerow(["y0_over_D", "Cv"]) + w.writerows((f"{g:.2f}", f"{c:.4f}") for g, c in zip(grid, table, strict=True)) + + rows = fig8(draws) + with open( + out / "thorsen2016_fig8_strain.csv", "w", newline="", encoding="utf-8" + ) as fh: + w = csv.writer(fh, lineterminator="\n") + w.writerow(["panel", "series", "U", "strain_1e-4"]) + w.writerows(rows) + + +if __name__ == "__main__": + main() diff --git a/validation/experiments/ndp_riser/measured_trim2005_fig16.csv b/validation/experiments/ndp_riser/measured_trim2005_fig16.csv new file mode 100644 index 0000000..e03e5a4 --- /dev/null +++ b/validation/experiments/ndp_riser/measured_trim2005_fig16.csv @@ -0,0 +1,25 @@ +# Trim et al. (2005) Fig. 16, bare riser, uniform flow: spatial mean of the temporal standard deviation of the displacement, sigma/D +# read by eye on the 4x guided crop of digitise_trim2005.py; +-0.01, about +-0.02 where markers of neighbouring speeds overlap +U,CF,IL +0.3,0.304,0.097 +0.4,0.401,0.129 +0.5,0.463,0.133 +0.6,0.395,0.142 +0.7,0.320,0.121 +0.8,0.395,0.125 +0.9,0.454,0.155 +1.0,0.415,0.136 +1.1,0.450,0.142 +1.2,0.418,0.136 +1.3,0.469,0.155 +1.4,0.452,0.155 +1.5,0.476,0.157 +1.6,0.469,0.170 +1.7,0.417,0.129 +1.8,0.433,0.133 +1.9,0.463,0.155 +2.0,0.427,0.157 +2.1,0.502,0.195 +2.2,0.484,0.183 +2.3,0.490,0.168 +2.4,0.505,0.215 diff --git a/validation/experiments/ndp_riser/measured_trim2005_fig18.csv b/validation/experiments/ndp_riser/measured_trim2005_fig18.csv new file mode 100644 index 0000000..6314870 --- /dev/null +++ b/validation/experiments/ndp_riser/measured_trim2005_fig18.csv @@ -0,0 +1,8 @@ +# Trim et al. (2005) Fig. 18, bare riser, uniform flow: peak frequency of the dominant displacement mode [Hz] +# read by eye on the 4x guided crop of digitise_trim2005.py at five speeds (the markers of the other speeds overlap); +-0.3 Hz, +-0.5 Hz for CF at 2.0 m/s (overlapping markers) +U,CF,IL +0.3,1.6,3.8 +1.0,5.6,11.1 +1.5,8.0,16.1 +2.0,9.9,20.4 +2.4,12.5,25.5 diff --git a/validation/experiments/ndp_riser/ndp_riser_viv.py b/validation/experiments/ndp_riser/ndp_riser_viv.py new file mode 100644 index 0000000..848efba --- /dev/null +++ b/validation/experiments/ndp_riser/ndp_riser_viv.py @@ -0,0 +1,570 @@ +#!/usr/bin/env python3 +# SPDX-License-Identifier: Apache-2.0 +# Copyright (c) 2026 Jae Hoon Seo, SMI Lab, Inha University +"""Cross-flow VIV of the NDP 38 m high-mode riser in uniform flow, against the measurements. + +The bare riser of the Norwegian Deepwater Programme tests (Trim et al. 2005), towed horizontally +at 0.3-2.4 m/s. Writes a deck per coefficient set and tow speed, runs them in parallel and +compares the maximum r.m.s. cross-flow strain along the riser (fundamental and total) with the +measured values of Thorsen et al. (2016, Fig. 8a), and the dominant cross-flow frequency and the +spatial mean of the cross-flow sigma/D with Trim et al. (2005, Figs. 18 and 16). + + python ndp_riser_viv.py path/to/cabledyn.exe WORKDIR [--sets b1,and24b,uls19,th16,th16c] + [--speeds 0.5,1.0,1.5,2.0,2.4 | all] [--segments 200] [--dt 0.001] [--tmax 25] + [--t0 10] [--tension 4000] [--kax 3500] [--jobs N] [--keep] [--summary DIR] + +Sources: + A.D. Trim, H. Braaten, H. Lie, M.A. Tognarelli (2005), Experimental investigation of + vortex-induced vibration of long marine risers, Journal of Fluids and Structures 21(3), + 335-361, https://doi.org/10.1016/j.jfluidstructs.2005.07.014 (the rig, Sec. 2.1; Figs. 16 + and 18) + E. Fontaine, H. Marcollo, K. Vandiver, M. Triantafyllou, C. Larsen, M. Tognarelli, + Y. Constantinides, O. Oakley (2011), Reliability based factors of safety for VIV fatigue using + NDP riser high mode VIV tests, OMAE2011, Vol. 7, 511-522, https://doi.org/10.1115/OMAE2011-49820 + (Table 1: the riser properties) + M.J. Thorsen, S. Saevik, C.M. Larsen (2016), Time domain simulation of vortex-induced + vibrations in stationary and oscillating flows, Journal of Fluids and Structures 61, 1-19, + https://doi.org/10.1016/j.jfluidstructs.2015.11.006 (Table 1, Eq. 1, Figs. 7 and 8) + J.V. Ulveseter, M.J. Thorsen, S. Saevik, C.M. Larsen (2019), Simulating fundamental and higher + harmonic VIV of slender structures, Applied Ocean Research 90, 101856, + https://doi.org/10.1016/j.apor.2019.101856 (Table 1, Sec. 4.1) + M.L. Andersen, S. Saevik, J. Wu, B.J. Leira, H. Langseth (2024b), Simulating vortex-induced + vibrations in sheared current by using an empirical time-domain model with adaptive + parameters, Applied Ocean Research 153, 104284, https://doi.org/10.1016/j.apor.2024.104284 + (Table 3) + J.R. Chaplin, P.W. Bearman, F.J. Huera Huarte, R.J. Pattenden (2005), Laboratory measurements + of vortex-induced vibrations of a vertical tension riser in a stepped current, Journal of + Fluids and Structures 21(1), 3-24, https://doi.org/10.1016/j.jfluidstructs.2005.04.010 (Eq. 4) +""" + +from __future__ import annotations + +import argparse +import csv +import hashlib +import json +import math +import os +import shutil +import subprocess +from concurrent.futures import ThreadPoolExecutor +from pathlib import Path + +import numpy as np + +HERE = Path(__file__).resolve().parent + +# The riser (Fontaine et al. 2011, Table 1; Thorsen et al. 2016, Table 1): 38.00 m between +# universal joints, OD 27 mm, EI 598.8 N m^2, EA 8.19e6 N, 0.933 kg/m flooded (as tested). +LENGTH = 38.0 # m +DIAMETER = 0.027 # m +EI = 598.8 # N m^2 +EA_RISER = 8.19e6 # N +MASS = 0.933 # kg/m +# Fresh water: the mass ratio 1.62 of Fontaine et al. (2011, Table 1) is 0.933 kg/m over the +# mass of the displaced water at 1006 kg/m^3 (1.63 at 1000, 1.59 at 1025). +RHO_W = 1000.0 # kg/m^3 +GRAVITY = 9.80665 # m/s^2 +# The rig (Trim et al. 2005, Sec. 2.1): clump weights on inclined pendulum arms set the tension, +# 4-6 kN over the test speeds (not given per test). Modelled as an axial spring in series with +# the riser, T0 in still water; the default k_ax is derived from that range (see the README). +TENSION0 = 4000.0 # N +K_AX = 3500.0 # N/m +DEPTH = 5.0 # m, the riser's depth (not given; the current is uniform) +WATER_DEPTH = 10.0 # m +VIV_RMS_WINDOW = 0.5 # s: 500 steps of the default dtM, held fixed when dt changes +SPEEDS = [round(0.1 * k, 1) for k in range(3, 25)] # the 22 uniform-flow tests [m/s] +DEFAULT_SPEEDS = "0.5,1.0,1.5,2.0,2.4" + + +def chaplin_amplification(ad: float) -> float: + """Chaplin et al. (2005, Eq. 4): Cd/Cd0 = 1 + 2.3 s + 1.7 s^2, s = (A/D)/sqrt(2).""" + s = ad / math.sqrt(2.0) + return 1.0 + 2.3 * s + 1.7 * s * s + + +def thorsen_cv() -> list[tuple[float, float]]: + """Thorsen et al. (2016, Fig. 7) C_v(y0/D), from extract_thorsen2016.py.""" + with open(HERE / "thorsen2016_fig7_cv.csv", encoding="utf-8") as fh: + return [(float(r["y0_over_D"]), float(r["Cv"])) for r in csv.DictReader(fh)] + + +# Coefficient sets: LINE TYPES Cd, Cl dF cF; VIV row (ClX, alpha, Chh); VIV AMPLITUDE table; +# VIV DAMPING (a, b, c). Each as its source gives it; "fitted" says whether the source fitted +# it to these NDP tests. +SETS: dict[str, dict] = { + "b1": { + "label": "generic set (Delta Flume case)", + "fitted": "no", + "cd": 1.2, + "cl": (0.8, 0.08, 0.18), + "viv": (0.85, 0.15, None), + "amp": "chaplin", + "damping": (0, 0, 1), + }, + "and24b": { + "label": "Andersen et al. (2024b), Table 3, cross-flow average", + "fitted": "sheared-flow tests of this riser", + "cd": 1.16, + "cl": (1.28, 0.066, 0.131), + "viv": None, + "amp": None, + "damping": None, + }, + "uls19": { + "label": "Ulveseter et al. (2019), Table 1, Chh 1.5", + "fitted": "these uniform-flow tests", + "cd": 1.2, + "cl": (0.85, 0.064, 0.144), + "viv": (0.75, 0.15, 1.5), + "amp": None, + "damping": None, + }, + "th16": { + "label": "Thorsen et al. (2016), Fig. 7 and Eq. 1", + "fitted": "these uniform and sheared-flow tests", + "cd": 1.2, + "cl": (1.0, 0.08, 0.18), # the C_v table replaces Cl + "viv": None, + "amp": "thorsen", + "damping": (0.31, 0.89, 0), + }, + "th16c": { + "label": "th16 with Morison's linear cross-flow damping kept (variant)", + "fitted": "variant of th16", + "cd": 1.2, + "cl": (1.0, 0.08, 0.18), + "viv": None, + "amp": "thorsen", + "damping": (0.31, 0.89, 1), + }, +} + + +def effective_ea(kax: float) -> float: + """The rig spring in series with the riser, as one axial stiffness of the line.""" + return LENGTH / (1.0 / kax + LENGTH / EA_RISER) + + +def deck_text( + hset: str, + u: float, + segments: int, + dt: float, + tmax: float, + tension: float, + kax: float, +) -> str: + s = SETS[hset] + ea = effective_ea(kax) + l0 = LENGTH / (1.0 + tension / ea) # unstretched length: T0 in still water + cl, df, cf = s["cl"] + viv = "" + if s["viv"]: + clx, alpha, chh = s["viv"] + cols, vals = "ClX alpha", f"{clx} {alpha}" + if chh is not None: + cols, vals = cols + " Chh", vals + f" {chh}" + viv += ( + "--------------------- VIV ----------------------------------------------\n" + f"TypeName {cols}\n(-){' (-)' * len(cols.split())}\nndp {vals}\n" + ) + if s["amp"] == "chaplin": # C_v = Cl constant, Chaplin's drag amplification + rows = "".join( + f"ndp {0.1 * k:.2f} {cl} {chaplin_amplification(0.1 * k):.5f}\n" + for k in range(16) + ) + elif s["amp"] == "thorsen": + rows = "".join(f"ndp {ad:.2f} {c:.4f} 1.0\n" for ad, c in thorsen_cv()) + if s["amp"]: + viv += ( + "--------------------- VIV AMPLITUDE ------------------------------------\n" + f"TypeName AD Cv CdAmp\n(-) (-) (-) (-)\n{rows}" + ) + if s["damping"]: + a, b, c = s["damping"] + viv += ( + "--------------------- VIV DAMPING ------------------------------------\n" + f"TypeName a b c\n(-) (-) (-) (-)\nndp {a} {b} {c}\n" + ) + nodes = segments + 1 + outs = "FairTen1\n" + "".join( + f"CurvX1N{j}\nCurvY1N{j}\n" for j in range(1, nodes + 1) + ) + outs += "".join(f"L1N{j}py\nL1N{j}pz\n" for j in range(1, nodes + 1, 4)) + return f"""--------------------- CableDyn Input File --------------------------------- +NDP 38 m riser, uniform {u} m/s, T0 {tension:g} N, k_ax {kax:g} N/m, set {hset} +----------------------- LINE TYPES ------------------------------------------ +TypeName Diam Mass EA BA EI Cd_n Cd_t Ca_n Ca_t Cl dF cF +(-) (m) (kg/m) (N) (N-s) (N-m^2) (-) (-) (-) (-) (-) (-) (-) +ndp {DIAMETER} {MASS} {ea:.6e} 0.0 {EI} {s["cd"]} 0.0 1.0 0.0 {cl} {df} {cf} +----------------------- POINTS ---------------------------------------------- +ID Attachment X Y Z M V CdA CA +(-) (-) (m) (m) (m) (kg) (m^3) (m^2) (-) +1 Coupled 0.0 0.0 {-DEPTH} 0 0 0 0 +2 Fixed {LENGTH} 0.0 {-DEPTH} 0 0 0 0 +----------------------- LINES ----------------------------------------------- +ID LineType AttachA AttachB UnstrLen NumSegs Outputs +(-) (-) (-) (-) (m) (-) (-) +1 ndp 1 2 {l0:.9f} {segments} - +{viv}---------------------- OPTIONS ---------------------------------------------- +{RHO_W} rhoW +{GRAVITY} g +{WATER_DEPTH} WtrDpth +{dt:.6f} dtM +{round(tmax / dt) * dt:.6f} TMax +1 Currents +{VIV_RMS_WINDOW} VIVrmsWindow +---------------------- OUTPUTS ----------------------------------------------- +{outs}END +------------------------- need this line -------------------------------------- +""" + + +# Search bands of the dominant frequencies, in f D/U: the cross-flow one covers the +# synchronisation band of every set (0.065-0.26) and excludes a slow swing of the riser below it +# and the third harmonic above it; the in-line one is twice that. +CF_BAND = (0.05, 0.30) +IL_BAND = (0.10, 0.60) +ANALYSIS = "1" # part of the reuse key: bump when analyse() changes + + +def spectrum_peak(x: np.ndarray, dt: float, lo: float, hi: float) -> float: + """Peak of the spectra summed over the rows, between lo and hi [Hz].""" + xm = x - x.mean(axis=-1, keepdims=True) + sp = (np.abs(np.fft.rfft(xm, axis=-1)) ** 2).sum(axis=0) + f = np.fft.rfftfreq(xm.shape[-1], dt) + sp[(f < lo) | (f > hi)] = 0.0 + return float(f[int(np.argmax(sp))]) + + +def bandpass(x: np.ndarray, dt: float, lo: float, hi: float) -> np.ndarray: + xm = x - x.mean(axis=-1, keepdims=True) + f = np.fft.rfftfreq(xm.shape[-1], dt) + sp = np.fft.rfft(xm, axis=-1) + sp[..., (f < lo) | (f > hi)] = 0.0 + return np.fft.irfft(sp, xm.shape[-1], axis=-1) + + +def analyse(out: Path, u: float, t0: float) -> dict: + """Reduce a run to the compared quantities (statistics over t > t0). + + Cross-flow is vertical (z) and in-line horizontal (y). The section axis x of a line not + restrained in torsion starts at End A as the projection of global +z normal to the line and + is carried along it, so CurvX is the cross-flow curvature (the static bow tilts it by + under 2 degrees) and CurvY the in-line one. The dominant frequencies are the peaks of the + curvature spectra summed over the nodes, inside CF_BAND and IL_BAND. The strain is + kappa D/2 at the outer surface; its fundamental part is band-passed to 0.5-1.5 times the + dominant cross-flow frequency, as Thorsen et al. (2016) filter the measured strain. sigma/D + is band-passed the same way about the dominant frequency of its direction, as the measured + modal reconstruction holds the vibration; the raw values keep any slow swing. + """ + with open(out, encoding="utf-8") as fh: + fh.readline() + head = fh.readline().split() + fh.readline() + a = np.loadtxt(fh) + t = a[:, 0] + k = t > t0 + dt = float(t[1] - t[0]) + s = u / DIAMETER # f D/U = 1 at s Hz + nodes = sorted({int(c[7:]) for c in head if c.startswith("CurvX1N")}) + cf = np.array([a[k, head.index(f"CurvX1N{j}")] for j in nodes]) + il = np.array([a[k, head.index(f"CurvY1N{j}")] for j in nodes]) + f_cf = spectrum_peak(cf, dt, CF_BAND[0] * s, CF_BAND[1] * s) + f_il = spectrum_peak(il, dt, IL_BAND[0] * s, IL_BAND[1] * s) + eps_cf = (cf - cf.mean(axis=1, keepdims=True)) * DIAMETER / 2.0 + eps_il = (il - il.mean(axis=1, keepdims=True)) * DIAMETER / 2.0 + rms_cf = np.sqrt((eps_cf**2).mean(axis=1)) + rms_cf_f = np.sqrt((bandpass(eps_cf, dt, 0.5 * f_cf, 1.5 * f_cf) ** 2).mean(axis=1)) + rms_il = np.sqrt((eps_il**2).mean(axis=1)) + pn = sorted( + {int(c[3:-2]) for c in head if c.startswith("L1N") and c.endswith("pz")} + ) + inner = pn[1:-1] # the end nodes are pinned + pz = np.array([a[k, head.index(f"L1N{j}pz")] for j in inner]) + py = np.array([a[k, head.index(f"L1N{j}py")] for j in inner]) + sd_cf = bandpass(pz, dt, 0.5 * f_cf, 1.5 * f_cf).std(axis=1) / DIAMETER + sd_il = bandpass(py, dt, 0.5 * f_il, 1.5 * f_il).std(axis=1) / DIAMETER + ten = a[k, head.index("FairTen1")] + return { + "mean_tension_N": float(ten.mean()), + "f_cf_hz": f_cf, + "f_il_hz": f_il, + "fD_over_U_cf": f_cf / s, + "f_cf_displacement_hz": spectrum_peak(pz, dt, CF_BAND[0] * s, CF_BAND[1] * s), + "f_cf_displacement_above_0p3hz": spectrum_peak(pz, dt, 0.3, np.inf), + "f_il_displacement_above_0p3hz": spectrum_peak(py, dt, 0.3, np.inf), + "frequency_resolution_hz": 1.0 / (k.sum() * dt), + "max_rms_strain_cf_fundamental": float(rms_cf_f.max()), + "max_rms_strain_cf_total": float(rms_cf.max()), + "max_rms_strain_il_total": float(rms_il.max()), + "sigma_over_D_cf_mean": float(sd_cf.mean()), + "sigma_over_D_il_mean": float(sd_il.mean()), + "sigma_over_D_cf_raw_mean": float((pz.std(axis=1) / DIAMETER).mean()), + "sigma_over_D_il_raw_mean": float((py.std(axis=1) / DIAMETER).mean()), + } + + +def run_case( + exe: Path, folder: Path, text: str, u: float, t0: float, keep: bool, sha: str +) -> dict: + """Write the deck, run it and analyse it. A folder that holds the same deck run by the + same executable is not run again: its result is reused when the statistics window and + the analysis are also the same, and its output is analysed again if it was kept.""" + folder.mkdir(parents=True, exist_ok=True) + deck, result, out = folder / "ndp.dat", folder / "result.json", folder / "ndp.out" + key = f"{sha}:{t0:g}:{ANALYSIS}" + same = deck.is_file() and deck.read_text(encoding="utf-8") == text + if same and result.is_file(): + old = json.loads(result.read_text(encoding="utf-8")) + if old.pop("reuse_key", None) == key: + return old + ran = folder / "executable_sha256.txt" + if not ( + same and out.is_file() and ran.is_file() and ran.read_text().strip() == sha + ): + deck.write_text(text, encoding="utf-8") + (folder / "current_profile.txt").write_text( + f"z ux uy uz\n(m) (m/s) (m/s) (m/s)\n{-WATER_DEPTH:.2f} 0 {u} 0\n0.0 0 {u} 0\n", + encoding="utf-8", + ) + env = dict(os.environ, OMP_NUM_THREADS="1") + proc = subprocess.run( + [str(exe), deck.name, "ndp"], + cwd=folder, + capture_output=True, + text=True, + encoding="utf-8", + errors="replace", + env=env, + check=False, + ) + (folder / "console.txt").write_text(proc.stdout + proc.stderr, encoding="utf-8") + if proc.returncode != 0: + raise SystemExit(f"{folder}: CableDyn exited with {proc.returncode}") + ran.write_text(sha + "\n", encoding="utf-8") + res = analyse(out, u, t0) + result.write_text( + json.dumps({**res, "reuse_key": key}, indent=1) + "\n", encoding="utf-8" + ) + if not keep: + out.unlink() + return res + + +def read_csv(name: str) -> list[dict]: + with open(HERE / name, encoding="utf-8") as fh: + return list(csv.DictReader(r for r in fh if not r.startswith("#"))) + + +def measured() -> dict: + """The measurements, keyed by speed rounded to 0.1 m/s.""" + m: dict = {"fig8": {}, "fig16": {}, "fig18": {}} + for r in read_csv("thorsen2016_fig8_strain.csv"): + if r["panel"] == "uniform": + key = (r["series"], round(float(r["U"]), 1)) + m["fig8"][key] = float(r["strain_1e-4"]) * 1e-4 + for r in read_csv("measured_trim2005_fig16.csv"): + m["fig16"][round(float(r["U"]), 1)] = (float(r["CF"]), float(r["IL"])) + for r in read_csv("measured_trim2005_fig18.csv"): + m["fig18"][round(float(r["U"]), 1)] = (float(r["CF"]), float(r["IL"])) + return m + + +def rms_log(ratios: list[float]) -> float | None: + return float(np.sqrt(np.mean(np.log(ratios) ** 2))) if ratios else None + + +def parse_speeds(text: str) -> list[float]: + if text == "all": + return SPEEDS + out = sorted({round(float(v), 1) for v in text.split(",")}) + if not out or any(v not in SPEEDS for v in out): + raise SystemExit( + f"--speeds: the tests are 0.3, 0.4, ..., 2.4 m/s, got {text!r}" + ) + return out + + +def main() -> None: + ap = argparse.ArgumentParser(description=__doc__.split("\n\n")[0]) + ap.add_argument("exe") + ap.add_argument("workdir") + ap.add_argument( + "--sets", default=",".join(SETS), help="comma list of " + ", ".join(SETS) + ) + ap.add_argument( + "--speeds", + default=DEFAULT_SPEEDS, + help="comma list of tow speeds [m/s], or all", + ) + ap.add_argument("--segments", type=int, default=200) + ap.add_argument("--dt", type=float, default=0.001) + ap.add_argument("--tmax", type=float, default=25.0) + ap.add_argument( + "--t0", type=float, default=10.0, help="start of the statistics [s]" + ) + ap.add_argument( + "--tension", type=float, default=TENSION0, help="T0 in still water [N]" + ) + ap.add_argument("--kax", type=float, default=K_AX, help="rig spring [N/m]") + ap.add_argument("--jobs", type=int, default=os.cpu_count() or 1) + ap.add_argument("--keep", action="store_true", help="keep the output tables") + ap.add_argument( + "--summary", + default=None, + help="directory for the comparison CSV and summary JSON", + ) + a = ap.parse_args() + found = a.exe if Path(a.exe).is_file() else shutil.which(a.exe) + if found is None: + raise SystemExit(f"CableDyn executable not found: {a.exe}") + exe = Path(found).resolve() + sets = [s.strip() for s in a.sets.split(",") if s.strip()] + if not sets or any(s not in SETS for s in sets): + raise SystemExit(f"--sets: choose from {', '.join(SETS)}, got {a.sets!r}") + speeds = parse_speeds(a.speeds) + tag = f"n{a.segments}_dt{a.dt:g}_T{a.tension:g}_k{a.kax:g}_t{a.tmax:g}" + jobs = [(s, u) for s in sets for u in speeds] + sha = hashlib.sha256(exe.read_bytes()).hexdigest() + + def one(job: tuple[str, float]) -> dict: + s, u = job + folder = Path(a.workdir).resolve() / f"{s}_u{u:.1f}_{tag}" + text = deck_text(s, u, a.segments, a.dt, a.tmax, a.tension, a.kax) + res = run_case(exe, folder, text, u, a.t0, a.keep, sha) + return {"set": s, "U_m_per_s": u, **res} + + with ThreadPoolExecutor(max_workers=max(1, min(a.jobs, len(jobs)))) as pool: + results = list(pool.map(one, jobs)) + + m = measured() + rows, per_set = [], {} + for s in sets: + rs = [r for r in results if r["set"] == s] + ratio, ratio_u5, ratio_th, ratio_th_u5, ratio_total = [], [], [], [], [] + fdu, fdu_meas, sd, sd_meas, sdi, sdi_meas = [], [], [], [], [], [] + for r in rs: + u = r["U_m_per_s"] + ef = m["fig8"][("exp_fundamental", u)] + et = m["fig8"][("exp_total", u)] + th = m["fig8"][("sim", u)] + q = r["max_rms_strain_cf_fundamental"] / ef + ratio.append(q) + ratio_th.append(th / ef) + ratio_total.append(r["max_rms_strain_cf_total"] / et) + if u >= 0.5: + ratio_u5.append(q) + ratio_th_u5.append(th / ef) + row = { + "set": s, + "U_m_per_s": u, + "mean_tension_N": round(r["mean_tension_N"], 0), + "strain_fund_measured": f"{ef:.4e}", + "strain_fund_cabledyn": f"{r['max_rms_strain_cf_fundamental']:.4e}", + "strain_fund_ratio": round(q, 3), + "strain_fund_thorsen2016_sim": f"{th:.4e}", + "strain_total_measured": f"{et:.4e}", + "strain_total_cabledyn": f"{r['max_rms_strain_cf_total']:.4e}", + "f_cf_cabledyn_hz": round(r["f_cf_hz"], 3), + "f_il_cabledyn_hz": round(r["f_il_hz"], 3), + "fD_U_cf_cabledyn": round(r["fD_over_U_cf"], 4), + "f_cf_measured_hz": "", + "f_il_measured_hz": "", + "sigma_D_cf_cabledyn": round(r["sigma_over_D_cf_mean"], 3), + "sigma_D_il_cabledyn": round(r["sigma_over_D_il_mean"], 3), + "sigma_D_cf_measured": "", + "sigma_D_il_measured": "", + } + if u in m["fig18"]: + fc, fi = m["fig18"][u] + row["f_cf_measured_hz"], row["f_il_measured_hz"] = fc, fi + fdu.append(r["fD_over_U_cf"]) + fdu_meas.append(fc * DIAMETER / u) + if u in m["fig16"]: + c, i = m["fig16"][u] + row["sigma_D_cf_measured"], row["sigma_D_il_measured"] = c, i + sd.append(r["sigma_over_D_cf_mean"]) + sd_meas.append(c) + sdi.append(r["sigma_over_D_il_mean"]) + sdi_meas.append(i) + rows.append(row) + per_set[s] = { + "label": SETS[s]["label"], + "fitted_to": SETS[s]["fitted"], + "speeds": [r["U_m_per_s"] for r in rs], + "strain_fund_ratio_median": float(np.median(ratio)), + "strain_fund_ratio_range": [float(min(ratio)), float(max(ratio))], + "strain_fund_rms_log_error": rms_log(ratio), + "strain_fund_rms_log_error_U_ge_0p5": rms_log(ratio_u5), + "thorsen2016_sim_rms_log_error_same_speeds": rms_log(ratio_th), + "thorsen2016_sim_rms_log_error_U_ge_0p5": rms_log(ratio_th_u5), + "strain_total_ratio_median": float(np.median(ratio_total)), + "strain_total_rms_log_error": rms_log(ratio_total), + "fD_U_cf_mean_at_fig18_speeds": float(np.mean(fdu)) if fdu else None, + "fD_U_cf_measured_mean_at_fig18_speeds": ( + float(np.mean(fdu_meas)) if fdu else None + ), + "fD_U_cf_mean_all_speeds": float(np.mean([r["fD_over_U_cf"] for r in rs])), + "sigma_D_cf_mean": float(np.mean(sd)) if sd else None, + "sigma_D_cf_measured_mean": float(np.mean(sd_meas)) if sd else None, + "sigma_D_il_mean": float(np.mean(sdi)) if sdi else None, + "sigma_D_il_measured_mean": float(np.mean(sdi_meas)) if sdi else None, + "mean_tension_N_range": [ + round(min(r["mean_tension_N"] for r in rs)), + round(max(r["mean_tension_N"] for r in rs)), + ], + } + ps = per_set[s] + print( + f"{s}: strain fund/measured median {ps['strain_fund_ratio_median']:.2f} " + f"({ps['strain_fund_ratio_range'][0]:.2f}-{ps['strain_fund_ratio_range'][1]:.2f}), " + f"rms log {ps['strain_fund_rms_log_error']:.3f}" + + ( + f" (U >= 0.5: {ps['strain_fund_rms_log_error_U_ge_0p5']:.3f}; Thorsen's sim " + f"{ps['thorsen2016_sim_rms_log_error_U_ge_0p5']:.3f})" + if ratio_u5 + else "" + ) + + f"; fD/U {ps['fD_U_cf_mean_all_speeds']:.3f}" + + ( + f" (Fig. 18 speeds {ps['fD_U_cf_mean_at_fig18_speeds']:.3f} vs " + f"{ps['fD_U_cf_measured_mean_at_fig18_speeds']:.3f})" + if fdu + else "" + ) + + ( + f"; sigma/D CF {ps['sigma_D_cf_mean']:.3f} vs {ps['sigma_D_cf_measured_mean']:.3f}" + if sd + else "" + ) + + f"; T {ps['mean_tension_N_range'][0]}-{ps['mean_tension_N_range'][1]} N" + ) + summary = { + "segments": a.segments, + "dt_s": a.dt, + "tmax_s": a.tmax, + "statistics_after_s": a.t0, + "tension0_N": a.tension, + "rig_spring_N_per_m": a.kax, + "effective_axial_stiffness_N": round(effective_ea(a.kax), 1), + "per_set": per_set, + "executable_sha256": sha, + } + if a.summary: + out = Path(a.summary) + out.mkdir(parents=True, exist_ok=True) + with open( + out / "response_comparison.csv", "w", newline="", encoding="utf-8" + ) as fh: + w = csv.DictWriter(fh, fieldnames=list(rows[0]), lineterminator="\n") + w.writeheader() + w.writerows(rows) + (out / "summary.json").write_text( + json.dumps(summary, indent=2) + "\n", encoding="utf-8", newline="\n" + ) + + +if __name__ == "__main__": + main() diff --git a/validation/experiments/ndp_riser/provenance.json b/validation/experiments/ndp_riser/provenance.json new file mode 100644 index 0000000..b3a109d --- /dev/null +++ b/validation/experiments/ndp_riser/provenance.json @@ -0,0 +1,308 @@ +{ + "source": "Trim, Braaten, Lie & Tognarelli (2005), Journal of Fluids and Structures 21(3), 335-361 (the rig, Sec. 2.1; Figs. 16 and 18); Fontaine, Marcollo, Vandiver, Triantafyllou, Larsen, Tognarelli, Constantinides & Oakley (2011), OMAE2011, Vol. 7, 511-522 (Table 1, the riser); Thorsen, Saevik & Larsen (2016), Journal of Fluids and Structures 61, 1-19 (Table 1, Eq. 1, Figs. 4, 7 and 8); Ulveseter, Thorsen, Saevik & Larsen (2019), Applied Ocean Research 90, 101856 (Tables 1 and 3, Sec. 4.1); Andersen, Saevik, Wu, Leira & Langseth (2024b), Applied Ocean Research 153, 104284 (Table 3); Ulveseter, Thorsen, Saevik & Larsen (2018), Marine Structures 60, 241-260 (Table 3); Chaplin, Bearman, Huera Huarte & Pattenden (2005), Journal of Fluids and Structures 21(1), 3-24 (Eq. 4)", + "doi": [ + "10.1016/j.jfluidstructs.2005.07.014", + "10.1115/OMAE2011-49820", + "10.1016/j.jfluidstructs.2015.11.006", + "10.1016/j.apor.2019.101856", + "10.1016/j.apor.2024.104284", + "10.1016/j.marstruc.2018.04.001", + "10.1016/j.jfluidstructs.2005.04.010" + ], + "bibliographic_data": "volumes, issues and pages from Crossref (api.crossref.org), 2026-11-16", + "data_used": "riser: 38.00 m between universal joints, OD 27 mm, EI 598.8 N m^2, EA 8.19e6 N, 0.933 kg/m water-filled, mass ratio 1.62 (Fontaine et al. 2011, Table 1); tension 4-6 kN over the tests (Trim et al. 2005, Sec. 2.1); the maximum r.m.s. cross-flow strain of the 22 uniform-flow tests, fundamental and total, and Thorsen et al.'s simulation (Thorsen et al. 2016, Fig. 8a; the sheared panel is extracted but not compared); Thorsen et al.'s fitted Cv(y0/D) (Fig. 7); the spatial mean sigma/D, CF and IL, at 22 speeds (Trim et al. 2005, Fig. 16, bare riser, uniform flow); the peak frequency of the dominant displacement mode, CF and IL, at 0.3, 1.0, 1.5, 2.0 and 2.4 m/s (Fig. 18)", + "extraction": "Thorsen et al. (2016) Figs. 7 and 8 are vector graphics: extract_thorsen2016.py maps the plotted coordinates through the plot frames (axis ends at the tick labels: 0-2.5 m/s and 0-5e-4 for Fig. 8a; y0/D 0-1 and Cv 0-1.5 for Fig. 7); 22 points per series. Trim et al. (2005) Figs. 16 and 18 are raster images with overlapping markers: read by eye on 4x enlarged crops with a guide at every speed and every 0.01 / 1 Hz, drawn by digitise_trim2005.py after calibrating the axes on the panel frames and grid lines; +-0.01 in sigma/D (about +-0.02 where neighbouring markers overlap), +-0.3 Hz (+-0.5 Hz for CF at 2.0 m/s)", + "assumptions": "fresh water 1000 kg/m^3 (the mass ratio 1.62 implies 1006 kg/m^3); the rig as an axial spring in series with the riser, T0 4 kN in still water, k_ax 3.5 kN/m derived from the reported 4-6 kN range with Morison drag at Cd 1.2 at 2.4 m/s (not tuned to the response); the riser 5 m below the surface in 10 m of water (not given; the current is uniform); Ca 1; VIVrmsWindow 0.5 s", + "coefficients": { + "b1": "the generic set of the Delta Flume case, nothing fitted to the NDP tests: Cl 0.8, cF 0.18, dF 0.08, Cd 1.2; VIV ClX 0.85, alpha 0.15 (Ulveseter et al. 2018, Table 3); VIV AMPLITUDE Cv 0.8 with CdAmp = 1 + 2.3 s + 1.7 s^2, s = (A/D)/sqrt(2), at A/D 0-1.5 (Chaplin et al. 2005, Eq. 4); VIV DAMPING 0 0 1", + "and24b": "Andersen et al. (2024b), Table 3, cross-flow average of the per-case optima for the 22 sheared-flow NDP tests: Cl 1.28, cF 0.131, dF 0.066, Cd 1.16, Ca 1.0; their in-line set (Cv,x 1.09, f0,x 0.321, df,x 0.161) left out", + "uls19": "Ulveseter et al. (2019), Table 1, fitted by them to these uniform-flow tests (Sec. 4.1): Cd 1.2, CM 2, Cl 0.85, cF 0.144, dF 0.064; VIV ClX 0.75, alpha 0.15, Chh 1.5", + "th16": "Thorsen et al. (2016), fitted by them to the 22 uniform and 22 sheared NDP tests: VIV AMPLITUDE with the Fig. 7 Cv curve (thorsen2016_fig7_cv.csv), VIV DAMPING 0.31 0.89 0 (Eq. 1), band 0.18 +- 0.08 for the 0.10-0.26 range of Fig. 4; Cd 1.2", + "th16c": "th16 with VIV DAMPING 0.31 0.89 1 (Morison's linear cross-flow damping kept): a variant" + }, + "CableDyn_commit": "041337e (dev) with the mid-point VIV phase update (fix/viv-coupled-order)", + "validation_date": "2026-11-16", + "executable_name": "cabledyn.exe", + "executable_sha256": "see summary.json", + "compiler": "GNU Fortran 16.2.0", + "build_type": "Release", + "platform": "Windows 11 x86-64", + "segments": 200, + "dt_s": 0.001, + "tmax_s": 25.0, + "statistics_after_s": 10.0, + "sensitivity": { + "note": "fundamental strain [-], dominant CF frequency [Hz], spatial-mean CF sigma/D and mean tension [N] at the given speeds, against response_comparison.csv; tmax_55: 55 s runs, statistics over 45 s", + "dt_0p0005": { + "b1": { + "1.0": { + "strain_fund": 0.00014618, + "f_cf_hz": 7.133, + "sigma_D_cf": 0.428, + "mean_tension_N": 4840.0 + }, + "2.4": { + "strain_fund": 0.00037938, + "f_cf_hz": 17.4, + "sigma_D_cf": 0.412, + "mean_tension_N": 9920.0 + } + }, + "and24b": { + "1.0": { + "strain_fund": 0.00012914, + "f_cf_hz": 5.333, + "sigma_D_cf": 0.631, + "mean_tension_N": 4333.0 + }, + "2.4": { + "strain_fund": 0.00025558, + "f_cf_hz": 11.4, + "sigma_D_cf": 0.69, + "mean_tension_N": 7534.0 + } + }, + "uls19": { + "1.0": { + "strain_fund": 9.1189e-05, + "f_cf_hz": 5.6, + "sigma_D_cf": 0.428, + "mean_tension_N": 4245.0 + }, + "2.4": { + "strain_fund": 0.0002669, + "f_cf_hz": 12.867, + "sigma_D_cf": 0.459, + "mean_tension_N": 6853.0 + } + }, + "th16": { + "1.0": { + "strain_fund": 0.00016422, + "f_cf_hz": 7.4, + "sigma_D_cf": 0.378, + "mean_tension_N": 4222.0 + }, + "2.4": { + "strain_fund": 0.00037693, + "f_cf_hz": 16.733, + "sigma_D_cf": 0.418, + "mean_tension_N": 6629.0 + } + }, + "th16c": { + "1.0": { + "strain_fund": 8.8411e-05, + "f_cf_hz": 6.0, + "sigma_D_cf": 0.303, + "mean_tension_N": 4191.0 + }, + "2.4": { + "strain_fund": 0.0001807, + "f_cf_hz": 12.067, + "sigma_D_cf": 0.32, + "mean_tension_N": 6405.0 + } + } + }, + "segments_400": { + "b1": { + "1.0": { + "strain_fund": 0.00014702, + "f_cf_hz": 7.133, + "sigma_D_cf": 0.424, + "mean_tension_N": 4845.0 + }, + "2.4": { + "strain_fund": 0.00040835, + "f_cf_hz": 18.333, + "sigma_D_cf": 0.399, + "mean_tension_N": 9869.0 + } + }, + "and24b": { + "1.0": { + "strain_fund": 0.00012939, + "f_cf_hz": 5.333, + "sigma_D_cf": 0.626, + "mean_tension_N": 4334.0 + }, + "2.4": { + "strain_fund": 0.00026037, + "f_cf_hz": 11.4, + "sigma_D_cf": 0.684, + "mean_tension_N": 7541.0 + } + }, + "uls19": { + "1.0": { + "strain_fund": 0.00010609, + "f_cf_hz": 5.6, + "sigma_D_cf": 0.427, + "mean_tension_N": 4238.0 + }, + "2.4": { + "strain_fund": 0.00026787, + "f_cf_hz": 12.4, + "sigma_D_cf": 0.474, + "mean_tension_N": 6935.0 + } + }, + "th16": { + "1.0": { + "strain_fund": 0.0001661, + "f_cf_hz": 7.333, + "sigma_D_cf": 0.381, + "mean_tension_N": 4225.0 + }, + "2.4": { + "strain_fund": 0.00044926, + "f_cf_hz": 17.533, + "sigma_D_cf": 0.4, + "mean_tension_N": 6565.0 + } + }, + "th16c": { + "1.0": { + "strain_fund": 8.8573e-05, + "f_cf_hz": 6.0, + "sigma_D_cf": 0.3, + "mean_tension_N": 4191.0 + }, + "2.4": { + "strain_fund": 0.00017739, + "f_cf_hz": 13.467, + "sigma_D_cf": 0.314, + "mean_tension_N": 6418.0 + } + } + }, + "tmax_55": { + "b1": { + "2.4": { + "strain_fund": 0.0003392, + "f_cf_hz": 15.578, + "sigma_D_cf": 0.423, + "mean_tension_N": 10004.0 + } + }, + "and24b": { + "2.4": { + "strain_fund": 0.00025566, + "f_cf_hz": 12.111, + "sigma_D_cf": 0.691, + "mean_tension_N": 7534.0 + } + }, + "th16c": { + "2.4": { + "strain_fund": 0.00019167, + "f_cf_hz": 12.0, + "sigma_D_cf": 0.321, + "mean_tension_N": 6389.0 + } + } + }, + "tmax_55_dt_0p0005": { + "b1": { + "2.4": { + "strain_fund": 0.00037913, + "f_cf_hz": 15.578, + "sigma_D_cf": 0.413, + "mean_tension_N": 9931.0 + } + }, + "and24b": { + "2.4": { + "strain_fund": 0.00025398, + "f_cf_hz": 11.422, + "sigma_D_cf": 0.69, + "mean_tension_N": 7533.0 + } + }, + "th16c": { + "2.4": { + "strain_fund": 0.00021533, + "f_cf_hz": 12.0, + "sigma_D_cf": 0.322, + "mean_tension_N": 6376.0 + } + } + }, + "tmax_55_segments_400": { + "b1": { + "2.4": { + "strain_fund": 0.00041136, + "f_cf_hz": 18.356, + "sigma_D_cf": 0.395, + "mean_tension_N": 9824.0 + } + }, + "and24b": { + "2.4": { + "strain_fund": 0.00025738, + "f_cf_hz": 11.422, + "sigma_D_cf": 0.685, + "mean_tension_N": 7540.0 + } + }, + "th16c": { + "2.4": { + "strain_fund": 0.00017665, + "f_cf_hz": 13.422, + "sigma_D_cf": 0.314, + "mean_tension_N": 6418.0 + } + } + }, + "full_sweep_dt_0p0005": { + "b1": { + "strain_fund_ratio_median": 1.2537, + "strain_fund_rms_log_error": 0.2802, + "strain_fund_rms_log_error_U_ge_0p5": 0.2809, + "fD_U_cf_mean_at_fig18_speeds": 0.191, + "sigma_D_cf_mean": 0.4307, + "sigma_D_il_mean": 0.0407 + }, + "and24b": { + "strain_fund_ratio_median": 1.0032, + "strain_fund_rms_log_error": 0.1868, + "strain_fund_rms_log_error_U_ge_0p5": 0.1821, + "fD_U_cf_mean_at_fig18_speeds": 0.1341, + "sigma_D_cf_mean": 0.687, + "sigma_D_il_mean": 0.0882 + }, + "uls19": { + "strain_fund_ratio_median": 0.9383, + "strain_fund_rms_log_error": 0.1842, + "strain_fund_rms_log_error_U_ge_0p5": 0.1727, + "fD_U_cf_mean_at_fig18_speeds": 0.1443, + "sigma_D_cf_mean": 0.4543, + "sigma_D_il_mean": 0.147 + }, + "th16": { + "strain_fund_ratio_median": 1.3866, + "strain_fund_rms_log_error": 0.3752, + "strain_fund_rms_log_error_U_ge_0p5": 0.3908, + "fD_U_cf_mean_at_fig18_speeds": 0.1927, + "sigma_D_cf_mean": 0.3888, + "sigma_D_il_mean": 0.0876 + }, + "th16c": { + "strain_fund_ratio_median": 0.743, + "strain_fund_rms_log_error": 0.3762, + "strain_fund_rms_log_error_U_ge_0p5": 0.3878, + "fD_U_cf_mean_at_fig18_speeds": 0.154, + "sigma_D_cf_mean": 0.3044, + "sigma_D_il_mean": 0.0256 + } + }, + "water_1025_U_ge_0p5_rms_log_error": { + "b1": 0.299, + "and24b": 0.181, + "uls19": 0.154, + "th16": 0.434, + "th16c": 0.353 + } + } +} diff --git a/validation/experiments/ndp_riser/response_comparison.csv b/validation/experiments/ndp_riser/response_comparison.csv new file mode 100644 index 0000000..215120d --- /dev/null +++ b/validation/experiments/ndp_riser/response_comparison.csv @@ -0,0 +1,111 @@ +set,U_m_per_s,mean_tension_N,strain_fund_measured,strain_fund_cabledyn,strain_fund_ratio,strain_fund_thorsen2016_sim,strain_total_measured,strain_total_cabledyn,f_cf_cabledyn_hz,f_il_cabledyn_hz,fD_U_cf_cabledyn,f_cf_measured_hz,f_il_measured_hz,sigma_D_cf_cabledyn,sigma_D_il_cabledyn,sigma_D_cf_measured,sigma_D_il_measured +b1,0.3,4017.0,1.3090e-05,1.5180e-05,1.16,1.1330e-05,1.6120e-05,1.5697e-05,2.133,4.267,0.192,1.6,3.8,0.438,0.023,0.304,0.097 +b1,0.4,4037.0,1.8130e-05,2.5739e-05,1.42,2.4420e-05,2.4680e-05,2.6388e-05,2.867,5.667,0.1935,,,0.433,0.031,0.401,0.129 +b1,0.5,4082.0,4.0290e-05,4.2118e-05,1.045,3.0210e-05,5.0610e-05,4.2880e-05,3.6,7.133,0.1944,,,0.427,0.027,0.463,0.133 +b1,0.6,4150.0,4.4320e-05,4.6559e-05,1.051,4.7080e-05,5.4380e-05,5.1423e-05,3.733,7.467,0.168,,,0.44,0.031,0.395,0.142 +b1,0.7,4260.0,5.7660e-05,8.3717e-05,1.452,6.5210e-05,7.3770e-05,8.4440e-05,5.133,10.267,0.198,,,0.42,0.027,0.32,0.121 +b1,0.8,4417.0,5.8160e-05,9.1110e-05,1.567,8.7860e-05,7.7040e-05,9.1729e-05,5.333,10.667,0.18,,,0.44,0.029,0.395,0.125 +b1,0.9,4622.0,7.9560e-05,1.2112e-04,1.522,8.5600e-05,1.0348e-04,1.2267e-04,6.2,12.4,0.186,,,0.438,0.027,0.454,0.155 +b1,1.0,4844.0,1.0575e-04,1.4678e-04,1.388,9.9440e-05,1.2312e-04,1.4772e-04,7.133,14.267,0.1926,5.6,11.1,0.428,0.031,0.415,0.136 +b1,1.1,5141.0,1.3722e-04,1.6015e-04,1.167,1.4099e-04,1.5433e-04,1.6121e-04,7.4,14.8,0.1816,,,0.452,0.023,0.45,0.142 +b1,1.2,5421.0,1.2513e-04,1.8382e-04,1.469,1.6113e-04,1.7297e-04,1.8447e-04,8.4,16.8,0.189,,,0.429,0.028,0.418,0.136 +b1,1.3,5596.0,1.2866e-04,1.8571e-04,1.443,1.5307e-04,1.8656e-04,1.8641e-04,9.267,18.6,0.1925,,,0.419,0.071,0.469,0.155 +b1,1.4,6092.0,1.8455e-04,2.2808e-04,1.236,2.0569e-04,2.1551e-04,2.2884e-04,9.8,21.133,0.189,,,0.446,0.024,0.452,0.155 +b1,1.5,6466.0,1.8077e-04,2.2950e-04,1.27,2.2734e-04,2.1123e-04,2.3127e-04,10.867,19.467,0.1956,8.0,16.1,0.453,0.026,0.476,0.157 +b1,1.6,6674.0,2.1274e-04,2.1048e-04,0.989,2.1198e-04,2.4975e-04,2.1170e-04,11.067,22.133,0.1867,,,0.43,0.115,0.469,0.17 +b1,1.7,7042.0,2.0997e-04,2.9202e-04,1.391,2.3061e-04,2.1879e-04,2.9232e-04,12.2,24.4,0.1938,,,0.401,0.032,0.417,0.129 +b1,1.8,7406.0,2.0695e-04,2.9568e-04,1.429,2.5932e-04,2.2785e-04,2.9773e-04,12.533,25.133,0.188,,,0.405,0.051,0.433,0.133 +b1,1.9,7946.0,2.4195e-04,3.0114e-04,1.245,2.8273e-04,2.8676e-04,3.0253e-04,12.267,27.133,0.1743,,,0.436,0.043,0.463,0.155 +b1,2.0,8272.0,2.4421e-04,3.6436e-04,1.492,3.0262e-04,2.7241e-04,3.6452e-04,15.067,30.2,0.2034,9.9,20.4,0.396,0.034,0.427,0.157 +b1,2.1,8760.0,2.8374e-04,3.0179e-04,1.064,3.1017e-04,4.0509e-04,3.0577e-04,13.667,28.133,0.1757,,,0.433,0.064,0.502,0.195 +b1,2.2,9143.0,2.9834e-04,2.9467e-04,0.988,3.1219e-04,3.8293e-04,2.9775e-04,14.0,27.933,0.1718,,,0.425,0.061,0.484,0.183 +b1,2.3,9539.0,2.7744e-04,3.4179e-04,1.232,3.3635e-04,3.3082e-04,3.4459e-04,16.133,32.267,0.1894,,,0.423,0.053,0.49,0.168 +b1,2.4,9981.0,3.0816e-04,3.4019e-04,1.104,3.7890e-04,4.7659e-04,3.4328e-04,15.6,32.067,0.1755,12.5,25.5,0.423,0.039,0.505,0.215 +and24b,0.3,4010.0,1.3090e-05,1.0432e-05,0.797,1.1330e-05,1.6120e-05,1.0603e-05,1.467,3.0,0.132,1.6,3.8,0.671,0.068,0.304,0.097 +and24b,0.4,4017.0,1.8130e-05,2.2822e-05,1.259,2.4420e-05,2.4680e-05,2.3700e-05,2.133,4.333,0.144,,,0.629,0.065,0.401,0.129 +and24b,0.5,4029.0,4.0290e-05,2.6969e-05,0.669,3.0210e-05,5.0610e-05,2.7808e-05,2.333,4.667,0.126,,,0.694,0.077,0.463,0.133 +and24b,0.6,4048.0,4.4320e-05,3.0147e-05,0.68,4.7080e-05,5.4380e-05,3.0634e-05,2.533,5.067,0.114,,,0.73,0.084,0.395,0.142 +and24b,0.7,4094.0,5.7660e-05,6.7629e-05,1.173,6.5210e-05,7.3770e-05,7.4047e-05,3.667,7.333,0.1414,,,0.633,0.078,0.32,0.121 +and24b,0.8,4146.0,5.8160e-05,7.4043e-05,1.273,8.7860e-05,7.7040e-05,7.9819e-05,3.867,7.733,0.1305,,,0.681,0.084,0.395,0.125 +and24b,0.9,4213.0,7.9560e-05,7.8777e-05,0.99,8.5600e-05,1.0348e-04,8.3179e-05,4.067,8.133,0.122,,,0.718,0.089,0.454,0.155 +and24b,1.0,4333.0,1.0575e-04,1.2923e-04,1.222,9.9440e-05,1.2312e-04,1.4341e-04,5.333,10.6,0.144,5.6,11.1,0.631,0.086,0.415,0.136 +and24b,1.1,4452.0,1.3722e-04,1.3572e-04,0.989,1.4099e-04,1.5433e-04,1.5234e-04,5.533,11.067,0.1358,,,0.668,0.089,0.45,0.142 +and24b,1.2,4586.0,1.2513e-04,1.4183e-04,1.133,1.6113e-04,1.7297e-04,1.5743e-04,5.733,11.467,0.129,,,0.699,0.092,0.418,0.136 +and24b,1.3,4737.0,1.2866e-04,1.4788e-04,1.149,1.5307e-04,1.8656e-04,1.6090e-04,6.0,12.0,0.1246,,,0.723,0.093,0.469,0.155 +and24b,1.4,4946.0,1.8455e-04,1.8077e-04,0.98,2.0569e-04,2.1551e-04,2.0274e-04,6.8,13.533,0.1311,,,0.697,0.093,0.452,0.155 +and24b,1.5,5164.0,1.8077e-04,2.1609e-04,1.195,2.2734e-04,2.1123e-04,2.3625e-04,7.6,15.8,0.1368,8.0,16.1,0.672,0.097,0.476,0.157 +and24b,1.6,5379.0,2.1274e-04,2.1984e-04,1.033,2.1198e-04,2.4975e-04,2.4512e-04,7.867,15.733,0.1327,,,0.691,0.101,0.469,0.17 +and24b,1.7,5605.0,2.0997e-04,1.9424e-04,0.925,2.3061e-04,2.1879e-04,2.0628e-04,8.2,16.933,0.1302,,,0.688,0.085,0.417,0.129 +and24b,1.8,5879.0,2.0695e-04,2.6359e-04,1.274,2.5932e-04,2.2785e-04,2.9328e-04,9.0,18.0,0.135,,,0.68,0.098,0.433,0.133 +and24b,1.9,6132.0,2.4195e-04,2.6145e-04,1.081,2.8273e-04,2.8676e-04,2.8067e-04,9.333,18.667,0.1326,,,0.658,0.083,0.463,0.155 +and24b,2.0,6380.0,2.4421e-04,2.2234e-04,0.91,3.0262e-04,2.7241e-04,2.4107e-04,9.6,19.867,0.1296,9.9,20.4,0.687,0.087,0.427,0.157 +and24b,2.1,6640.0,2.8374e-04,2.3506e-04,0.828,3.1017e-04,4.0509e-04,2.5299e-04,9.867,19.133,0.1269,,,0.696,0.088,0.502,0.195 +and24b,2.2,6912.0,2.9834e-04,2.7091e-04,0.908,3.1219e-04,3.8293e-04,3.0354e-04,10.133,20.267,0.1244,,,0.737,0.117,0.484,0.183 +and24b,2.3,7162.0,2.7744e-04,2.8247e-04,1.018,3.3635e-04,3.3082e-04,3.0757e-04,10.333,20.667,0.1213,,,0.746,0.097,0.49,0.168 +and24b,2.4,7535.0,3.0816e-04,2.5869e-04,0.839,3.7890e-04,4.7659e-04,2.8308e-04,11.4,23.533,0.1283,12.5,25.5,0.69,0.088,0.505,0.215 +uls19,0.3,4009.0,1.3090e-05,8.3712e-06,0.64,1.1330e-05,1.6120e-05,1.7166e-05,1.467,2.933,0.132,1.6,3.8,0.478,0.112,0.304,0.097 +uls19,0.4,4015.0,1.8130e-05,1.9791e-05,1.092,2.4420e-05,2.4680e-05,3.4424e-05,2.067,4.2,0.1395,,,0.485,0.125,0.401,0.129 +uls19,0.5,4028.0,4.0290e-05,3.3296e-05,0.826,3.0210e-05,5.0610e-05,5.2598e-05,2.8,5.6,0.1512,,,0.441,0.15,0.463,0.133 +uls19,0.6,4040.0,4.4320e-05,4.8202e-05,1.088,4.7080e-05,5.4380e-05,6.8038e-05,3.267,6.6,0.147,,,0.411,0.135,0.395,0.142 +uls19,0.7,4076.0,5.7660e-05,7.2993e-05,1.266,6.5210e-05,7.3770e-05,8.5309e-05,4.0,8.067,0.1543,,,0.463,0.129,0.32,0.121 +uls19,0.8,4108.0,5.8160e-05,7.7277e-05,1.329,8.7860e-05,7.7040e-05,9.7530e-05,4.267,8.4,0.144,,,0.459,0.138,0.395,0.125 +uls19,0.9,4172.0,7.9560e-05,8.8428e-05,1.111,8.5600e-05,1.0348e-04,1.2294e-04,4.933,9.867,0.148,,,0.416,0.154,0.454,0.155 +uls19,1.0,4245.0,1.0575e-04,8.9501e-05,0.846,9.9440e-05,1.2312e-04,1.2063e-04,5.8,11.6,0.1566,5.6,11.1,0.436,0.146,0.415,0.136 +uls19,1.1,4400.0,1.3722e-04,1.2085e-04,0.881,1.4099e-04,1.5433e-04,1.4924e-04,6.0,12.067,0.1473,,,0.508,0.145,0.45,0.142 +uls19,1.2,4500.0,1.2513e-04,1.3512e-04,1.08,1.6113e-04,1.7297e-04,1.6277e-04,6.667,13.267,0.15,,,0.496,0.143,0.418,0.136 +uls19,1.3,4627.0,1.2866e-04,1.4679e-04,1.141,1.5307e-04,1.8656e-04,1.7849e-04,6.867,13.733,0.1426,,,0.491,0.153,0.469,0.155 +uls19,1.4,4745.0,1.8455e-04,1.5619e-04,0.846,2.0569e-04,2.1551e-04,1.9707e-04,7.667,15.2,0.1479,,,0.428,0.154,0.452,0.155 +uls19,1.5,4913.0,1.8077e-04,1.5773e-04,0.873,2.2734e-04,2.1123e-04,1.9682e-04,8.0,16.0,0.144,8.0,16.1,0.474,0.154,0.476,0.157 +uls19,1.6,5115.0,2.1274e-04,1.8737e-04,0.881,2.1198e-04,2.4975e-04,2.1864e-04,8.733,17.2,0.1474,,,0.485,0.143,0.469,0.17 +uls19,1.7,5201.0,2.0997e-04,2.2019e-04,1.049,2.3061e-04,2.1879e-04,2.5745e-04,9.533,18.267,0.1514,,,0.443,0.151,0.417,0.129 +uls19,1.8,5524.0,2.0695e-04,1.9884e-04,0.961,2.5932e-04,2.2785e-04,2.5341e-04,9.4,18.733,0.141,,,0.478,0.156,0.433,0.133 +uls19,1.9,5627.0,2.4195e-04,2.1028e-04,0.869,2.8273e-04,2.8676e-04,2.4640e-04,10.2,19.933,0.1449,,,0.446,0.155,0.463,0.155 +uls19,2.0,5920.0,2.4421e-04,2.1808e-04,0.893,3.0262e-04,2.7241e-04,2.6849e-04,11.0,22.0,0.1485,9.9,20.4,0.453,0.155,0.427,0.157 +uls19,2.1,6153.0,2.8374e-04,2.5197e-04,0.888,3.1017e-04,4.0509e-04,3.0815e-04,10.867,22.067,0.1397,,,0.461,0.17,0.502,0.195 +uls19,2.2,6461.0,2.9834e-04,2.2824e-04,0.765,3.1219e-04,3.8293e-04,2.7473e-04,11.8,24.133,0.1448,,,0.46,0.149,0.484,0.183 +uls19,2.3,6591.0,2.7744e-04,2.5879e-04,0.933,3.3635e-04,3.3082e-04,2.9310e-04,12.333,24.0,0.1448,,,0.463,0.151,0.49,0.168 +uls19,2.4,6948.0,3.0816e-04,2.5200e-04,0.818,3.7890e-04,4.7659e-04,3.1634e-04,12.933,25.067,0.1455,12.5,25.5,0.461,0.149,0.505,0.215 +th16,0.3,4009.0,1.3090e-05,1.2219e-05,0.933,1.1330e-05,1.6120e-05,1.2590e-05,2.067,1.533,0.186,1.6,3.8,0.353,0.069,0.304,0.097 +th16,0.4,4013.0,1.8130e-05,2.2056e-05,1.217,2.4420e-05,2.4680e-05,2.2236e-05,2.8,2.2,0.189,,,0.365,0.081,0.401,0.129 +th16,0.5,4021.0,4.0290e-05,3.6601e-05,0.908,3.0210e-05,5.0610e-05,3.6911e-05,3.533,2.933,0.1908,,,0.362,0.089,0.463,0.133 +th16,0.6,4037.0,4.4320e-05,5.9852e-05,1.35,4.7080e-05,5.4380e-05,6.0772e-05,4.267,8.467,0.192,,,0.374,0.055,0.395,0.142 +th16,0.7,4062.0,5.7660e-05,8.1696e-05,1.417,6.5210e-05,7.3770e-05,8.2243e-05,5.0,10.6,0.1929,,,0.379,0.055,0.32,0.121 +th16,0.8,4102.0,5.8160e-05,1.0811e-04,1.859,8.7860e-05,7.7040e-05,1.0871e-04,5.733,11.533,0.1935,,,0.379,0.063,0.395,0.125 +th16,0.9,4154.0,7.9560e-05,1.3436e-04,1.689,8.5600e-05,1.0348e-04,1.3505e-04,6.533,13.8,0.196,,,0.395,0.061,0.454,0.155 +th16,1.0,4224.0,1.0575e-04,1.6547e-04,1.565,9.9440e-05,1.2312e-04,1.6613e-04,7.333,6.733,0.198,5.6,11.1,0.389,0.09,0.415,0.136 +th16,1.1,4306.0,1.3722e-04,1.5672e-04,1.142,1.4099e-04,1.5433e-04,1.5737e-04,8.2,8.867,0.2013,,,0.389,0.101,0.45,0.142 +th16,1.2,4411.0,1.2513e-04,2.2072e-04,1.764,1.6113e-04,1.7297e-04,2.2197e-04,9.067,8.4,0.204,,,0.385,0.09,0.418,0.136 +th16,1.3,4523.0,1.2866e-04,2.3862e-04,1.855,1.5307e-04,1.8656e-04,2.3973e-04,10.0,17.733,0.2077,,,0.387,0.068,0.469,0.155 +th16,1.4,4645.0,1.8455e-04,2.4232e-04,1.313,2.0569e-04,2.1551e-04,2.4381e-04,10.933,10.267,0.2109,,,0.393,0.087,0.452,0.155 +th16,1.5,4778.0,1.8077e-04,2.4964e-04,1.381,2.2734e-04,2.1123e-04,2.5145e-04,11.133,10.533,0.2004,8.0,16.1,0.401,0.088,0.476,0.157 +th16,1.6,4937.0,2.1274e-04,2.9708e-04,1.396,2.1198e-04,2.4975e-04,2.9944e-04,11.467,10.8,0.1935,,,0.396,0.078,0.469,0.17 +th16,1.7,5085.0,2.0997e-04,2.3109e-04,1.101,2.3061e-04,2.1879e-04,2.3492e-04,11.533,22.133,0.1832,,,0.407,0.048,0.417,0.129 +th16,1.8,5282.0,2.0695e-04,2.9345e-04,1.418,2.5932e-04,2.2785e-04,2.9727e-04,12.733,22.667,0.191,,,0.406,0.055,0.433,0.133 +th16,1.9,5464.0,2.4195e-04,3.0898e-04,1.277,2.8273e-04,2.8676e-04,3.1913e-04,11.8,23.267,0.1677,,,0.406,0.053,0.463,0.155 +th16,2.0,5662.0,2.4421e-04,3.7359e-04,1.53,3.0262e-04,2.7241e-04,3.7618e-04,14.8,28.067,0.1998,9.9,20.4,0.401,0.047,0.427,0.157 +th16,2.1,5881.0,2.8374e-04,3.9839e-04,1.404,3.1017e-04,4.0509e-04,4.0433e-04,14.267,26.533,0.1834,,,0.404,0.048,0.502,0.195 +th16,2.2,6087.0,2.9834e-04,4.2156e-04,1.413,3.1219e-04,3.8293e-04,4.2378e-04,16.467,23.933,0.2021,,,0.399,0.052,0.484,0.183 +th16,2.3,6310.0,2.7744e-04,3.2597e-04,1.175,3.3635e-04,3.3082e-04,3.3143e-04,15.667,28.067,0.1839,,,0.406,0.044,0.49,0.168 +th16,2.4,6532.0,3.0816e-04,4.3928e-04,1.425,3.7890e-04,4.7659e-04,4.4402e-04,15.933,28.733,0.1793,12.5,25.5,0.404,0.043,0.505,0.215 +th16c,0.3,4008.0,1.3090e-05,1.0334e-05,0.789,1.1330e-05,1.6120e-05,1.0419e-05,2.067,4.2,0.186,1.6,3.8,0.264,0.013,0.304,0.097 +th16c,0.4,4011.0,1.8130e-05,1.4470e-05,0.798,2.4420e-05,2.4680e-05,1.4527e-05,2.267,4.533,0.153,,,0.287,0.017,0.401,0.129 +th16c,0.5,4018.0,4.0290e-05,2.4439e-05,0.607,3.0210e-05,5.0610e-05,2.4754e-05,2.933,5.933,0.1584,,,0.287,0.02,0.463,0.133 +th16c,0.6,4033.0,4.4320e-05,4.5026e-05,1.016,4.7080e-05,5.4380e-05,4.5679e-05,4.2,8.4,0.189,,,0.267,0.021,0.395,0.142 +th16c,0.7,4054.0,5.7660e-05,5.0420e-05,0.874,6.5210e-05,7.3770e-05,5.1733e-05,4.333,8.733,0.1671,,,0.287,0.023,0.32,0.121 +th16c,0.8,4082.0,5.8160e-05,5.4421e-05,0.936,8.7860e-05,7.7040e-05,5.5668e-05,4.533,9.0,0.153,,,0.303,0.024,0.395,0.125 +th16c,0.9,4135.0,7.9560e-05,8.4309e-05,1.06,8.5600e-05,1.0348e-04,8.5278e-05,5.867,11.667,0.176,,,0.287,0.026,0.454,0.155 +th16c,1.0,4191.0,1.0575e-04,8.8393e-05,0.836,9.9440e-05,1.2312e-04,9.0370e-05,6.0,11.933,0.162,5.6,11.1,0.303,0.027,0.415,0.136 +th16c,1.1,4259.0,1.3722e-04,9.3237e-05,0.679,1.4099e-04,1.5433e-04,9.5519e-05,6.2,12.4,0.1522,,,0.315,0.027,0.45,0.142 +th16c,1.2,4374.0,1.2513e-04,1.3112e-04,1.048,1.6113e-04,1.7297e-04,1.3172e-04,7.6,15.2,0.171,,,0.304,0.031,0.418,0.136 +th16c,1.3,4472.0,1.2866e-04,1.2934e-04,1.005,1.5307e-04,1.8656e-04,1.3116e-04,7.6,15.267,0.1578,,,0.319,0.031,0.469,0.155 +th16c,1.4,4579.0,1.8455e-04,1.2877e-04,0.698,2.0569e-04,2.1551e-04,1.3072e-04,7.933,15.533,0.153,,,0.326,0.029,0.452,0.155 +th16c,1.5,4705.0,1.8077e-04,1.3425e-04,0.743,2.2734e-04,2.1123e-04,1.3702e-04,8.0,16.0,0.144,8.0,16.1,0.332,0.027,0.476,0.157 +th16c,1.6,4836.0,2.1274e-04,1.1102e-04,0.522,2.1198e-04,2.4975e-04,1.1214e-04,9.267,18.533,0.1564,,,0.31,0.023,0.469,0.17 +th16c,1.7,5060.0,2.0997e-04,1.7737e-04,0.845,2.3061e-04,2.1879e-04,1.7807e-04,10.067,20.0,0.1599,,,0.323,0.032,0.417,0.129 +th16c,1.8,5177.0,2.0695e-04,1.3123e-04,0.634,2.5932e-04,2.2785e-04,1.3242e-04,9.333,18.667,0.14,,,0.311,0.025,0.433,0.133 +th16c,1.9,5347.0,2.4195e-04,1.7120e-04,0.708,2.8273e-04,2.8676e-04,1.7281e-04,9.933,19.933,0.1412,,,0.318,0.024,0.463,0.155 +th16c,2.0,5552.0,2.4421e-04,1.7717e-04,0.726,3.0262e-04,2.7241e-04,1.7967e-04,10.133,20.267,0.1368,9.9,20.4,0.345,0.037,0.427,0.157 +th16c,2.1,5711.0,2.8374e-04,2.0150e-04,0.71,3.1017e-04,4.0509e-04,2.0464e-04,10.533,21.067,0.1354,,,0.319,0.023,0.502,0.195 +th16c,2.2,5959.0,2.9834e-04,1.5342e-04,0.514,3.1219e-04,3.8293e-04,1.5567e-04,12.2,22.2,0.1497,,,0.317,0.025,0.484,0.183 +th16c,2.3,6168.0,2.7744e-04,1.5234e-04,0.549,3.3635e-04,3.3082e-04,1.5523e-04,11.867,23.267,0.1393,,,0.319,0.025,0.49,0.168 +th16c,2.4,6411.0,3.0816e-04,1.7044e-04,0.553,3.7890e-04,4.7659e-04,1.7355e-04,12.267,25.267,0.138,12.5,25.5,0.317,0.027,0.505,0.215 diff --git a/validation/experiments/ndp_riser/summary.json b/validation/experiments/ndp_riser/summary.json new file mode 100644 index 0000000..1004a9e --- /dev/null +++ b/validation/experiments/ndp_riser/summary.json @@ -0,0 +1,262 @@ +{ + "segments": 200, + "dt_s": 0.001, + "tmax_s": 25.0, + "statistics_after_s": 10.0, + "tension0_N": 4000.0, + "rig_spring_N_per_m": 3500.0, + "effective_axial_stiffness_N": 130874.7, + "per_set": { + "b1": { + "label": "generic set (Delta Flume case)", + "fitted_to": "no", + "speeds": [ + 0.3, + 0.4, + 0.5, + 0.6, + 0.7, + 0.8, + 0.9, + 1.0, + 1.1, + 1.2, + 1.3, + 1.4, + 1.5, + 1.6, + 1.7, + 1.8, + 1.9, + 2.0, + 2.1, + 2.2, + 2.3, + 2.4 + ], + "strain_fund_ratio_median": 1.2571062159593067, + "strain_fund_ratio_range": [ + 0.9877067373166128, + 1.5665480591009866 + ], + "strain_fund_rms_log_error": 0.2766299415215297, + "strain_fund_rms_log_error_U_ge_0p5": 0.27737960253468885, + "thorsen2016_sim_rms_log_error_same_speeds": 0.18701650558341884, + "thorsen2016_sim_rms_log_error_U_ge_0p5": 0.1816450776805147, + "strain_total_ratio_median": 1.0584269733357312, + "strain_total_rms_log_error": 0.1766799559913615, + "fD_U_cf_mean_at_fig18_speeds": 0.19182000000000002, + "fD_U_cf_measured_mean_at_fig18_speeds": 0.14269500000000002, + "fD_U_cf_mean_all_speeds": 0.18685691678791969, + "sigma_D_cf_mean": 0.4290083586999797, + "sigma_D_cf_measured_mean": 0.4364545454545455, + "sigma_D_il_mean": 0.04052171870161913, + "sigma_D_il_measured_mean": 0.14945454545454542, + "mean_tension_N_range": [ + 4017, + 9981 + ] + }, + "and24b": { + "label": "Andersen et al. (2024b), Table 3, cross-flow average", + "fitted_to": "sheared-flow tests of this riser", + "speeds": [ + 0.3, + 0.4, + 0.5, + 0.6, + 0.7, + 0.8, + 0.9, + 1.0, + 1.1, + 1.2, + 1.3, + 1.4, + 1.5, + 1.6, + 1.7, + 1.8, + 1.9, + 2.0, + 2.1, + 2.2, + 2.3, + 2.4 + ], + "strain_fund_ratio_median": 1.0041362120271444, + "strain_fund_ratio_range": [ + 0.669381360004605, + 1.273674211040034 + ], + "strain_fund_rms_log_error": 0.18607649659184072, + "strain_fund_rms_log_error_U_ge_0p5": 0.1812834271827885, + "thorsen2016_sim_rms_log_error_same_speeds": 0.18701650558341884, + "thorsen2016_sim_rms_log_error_U_ge_0p5": 0.1816450776805147, + "strain_total_ratio_median": 0.935238467733235, + "strain_total_rms_log_error": 0.27133630525808555, + "fD_U_cf_mean_at_fig18_speeds": 0.13413, + "fD_U_cf_measured_mean_at_fig18_speeds": 0.14269500000000002, + "fD_U_cf_mean_all_speeds": 0.13055895432349438, + "sigma_D_cf_mean": 0.6872850455513722, + "sigma_D_cf_measured_mean": 0.4364545454545455, + "sigma_D_il_mean": 0.08815225700625753, + "sigma_D_il_measured_mean": 0.14945454545454542, + "mean_tension_N_range": [ + 4010, + 7535 + ] + }, + "uls19": { + "label": "Ulveseter et al. (2019), Table 1, Chh 1.5", + "fitted_to": "these uniform-flow tests", + "speeds": [ + 0.3, + 0.4, + 0.5, + 0.6, + 0.7, + 0.8, + 0.9, + 1.0, + 1.1, + 1.2, + 1.3, + 1.4, + 1.5, + 1.6, + 1.7, + 1.8, + 1.9, + 2.0, + 2.1, + 2.2, + 2.3, + 2.4 + ], + "strain_fund_ratio_median": 0.890509603830604, + "strain_fund_ratio_range": [ + 0.6395097067278822, + 1.3286994362951445 + ], + "strain_fund_rms_log_error": 0.17784402916844322, + "strain_fund_rms_log_error_U_ge_0p5": 0.15625104689068844, + "thorsen2016_sim_rms_log_error_same_speeds": 0.18701650558341884, + "thorsen2016_sim_rms_log_error_U_ge_0p5": 0.1816450776805147, + "strain_total_ratio_median": 0.9734055771641398, + "strain_total_rms_log_error": 0.18332706785625352, + "fD_U_cf_mean_at_fig18_speeds": 0.14531999999999998, + "fD_U_cf_measured_mean_at_fig18_speeds": 0.14269500000000002, + "fD_U_cf_mean_all_speeds": 0.14601728083572835, + "sigma_D_cf_mean": 0.4606471786934472, + "sigma_D_cf_measured_mean": 0.4364545454545455, + "sigma_D_il_mean": 0.14625252681604584, + "sigma_D_il_measured_mean": 0.14945454545454542, + "mean_tension_N_range": [ + 4009, + 6948 + ] + }, + "th16": { + "label": "Thorsen et al. (2016), Fig. 7 and Eq. 1", + "fitted_to": "these uniform and sheared-flow tests", + "speeds": [ + 0.3, + 0.4, + 0.5, + 0.6, + 0.7, + 0.8, + 0.9, + 1.0, + 1.1, + 1.2, + 1.3, + 1.4, + 1.5, + 1.6, + 1.7, + 1.8, + 1.9, + 2.0, + 2.1, + 2.2, + 2.3, + 2.4 + ], + "strain_fund_ratio_median": 1.4002450719245019, + "strain_fund_ratio_range": [ + 0.9084299248668841, + 1.8588629520718365 + ], + "strain_fund_rms_log_error": 0.362608870706913, + "strain_fund_rms_log_error_U_ge_0p5": 0.3774587442086474, + "thorsen2016_sim_rms_log_error_same_speeds": 0.18701650558341884, + "thorsen2016_sim_rms_log_error_U_ge_0p5": 0.1816450776805147, + "strain_total_ratio_median": 1.1162060812212764, + "strain_total_rms_log_error": 0.20064970064034177, + "fD_U_cf_mean_at_fig18_speeds": 0.19269, + "fD_U_cf_measured_mean_at_fig18_speeds": 0.14269500000000002, + "fD_U_cf_mean_all_speeds": 0.19301011480870967, + "sigma_D_cf_mean": 0.3900255003761697, + "sigma_D_cf_measured_mean": 0.4364545454545455, + "sigma_D_il_mean": 0.06650981202584938, + "sigma_D_il_measured_mean": 0.14945454545454542, + "mean_tension_N_range": [ + 4009, + 6532 + ] + }, + "th16c": { + "label": "th16 with Morison's linear cross-flow damping kept (variant)", + "fitted_to": "variant of th16", + "speeds": [ + 0.3, + 0.4, + 0.5, + 0.6, + 0.7, + 0.8, + 0.9, + 1.0, + 1.1, + 1.2, + 1.3, + 1.4, + 1.5, + 1.6, + 1.7, + 1.8, + 1.9, + 2.0, + 2.1, + 2.2, + 2.3, + 2.4 + ], + "strain_fund_ratio_median": 0.7340841919585173, + "strain_fund_ratio_range": [ + 0.5142482769213371, + 1.0596904536917338 + ], + "strain_fund_rms_log_error": 0.3652774038700901, + "strain_fund_rms_log_error_U_ge_0p5": 0.3760778263719995, + "thorsen2016_sim_rms_log_error_same_speeds": 0.18701650558341884, + "thorsen2016_sim_rms_log_error_U_ge_0p5": 0.1816450776805147, + "strain_total_ratio_median": 0.6326367996919098, + "strain_total_rms_log_error": 0.5444253236443083, + "fD_U_cf_mean_at_fig18_speeds": 0.15336, + "fD_U_cf_measured_mean_at_fig18_speeds": 0.14269500000000002, + "fD_U_cf_mean_all_speeds": 0.15542028494685356, + "sigma_D_cf_mean": 0.30719622989869017, + "sigma_D_cf_measured_mean": 0.4364545454545455, + "sigma_D_il_mean": 0.02540983564316691, + "sigma_D_il_measured_mean": 0.14945454545454542, + "mean_tension_N_range": [ + 4008, + 6411 + ] + } + }, + "executable_sha256": "6dd528518f28d78a7c9d19319939f2ce8d0590ca0103fb67c389d92ab0c20b0c" +} diff --git a/validation/experiments/ndp_riser/thorsen2016_fig7_cv.csv b/validation/experiments/ndp_riser/thorsen2016_fig7_cv.csv new file mode 100644 index 0000000..86527bb --- /dev/null +++ b/validation/experiments/ndp_riser/thorsen2016_fig7_cv.csv @@ -0,0 +1,22 @@ +y0_over_D,Cv +0.00,0.3522 +0.05,0.4835 +0.10,0.5993 +0.15,0.7012 +0.20,0.7877 +0.25,0.8597 +0.30,0.9174 +0.35,0.9591 +0.40,0.9872 +0.45,1.0008 +0.50,0.9762 +0.55,0.8233 +0.60,0.5315 +0.65,0.1019 +0.70,0.0000 +0.75,0.0000 +0.80,0.0000 +0.85,0.0000 +0.90,0.0000 +0.95,0.0000 +1.00,0.0000 diff --git a/validation/experiments/ndp_riser/thorsen2016_fig8_strain.csv b/validation/experiments/ndp_riser/thorsen2016_fig8_strain.csv new file mode 100644 index 0000000..6defc21 --- /dev/null +++ b/validation/experiments/ndp_riser/thorsen2016_fig8_strain.csv @@ -0,0 +1,133 @@ +panel,series,U,strain_1e-4 +uniform,exp_fundamental,0.299,0.1309 +uniform,exp_fundamental,0.399,0.1813 +uniform,exp_fundamental,0.499,0.4029 +uniform,exp_fundamental,0.599,0.4432 +uniform,exp_fundamental,0.699,0.5766 +uniform,exp_fundamental,0.799,0.5816 +uniform,exp_fundamental,0.899,0.7956 +uniform,exp_fundamental,0.999,1.0575 +uniform,exp_fundamental,1.099,1.3722 +uniform,exp_fundamental,1.199,1.2513 +uniform,exp_fundamental,1.299,1.2866 +uniform,exp_fundamental,1.399,1.8455 +uniform,exp_fundamental,1.499,1.8077 +uniform,exp_fundamental,1.599,2.1274 +uniform,exp_fundamental,1.699,2.0997 +uniform,exp_fundamental,1.799,2.0695 +uniform,exp_fundamental,1.899,2.4195 +uniform,exp_fundamental,1.999,2.4421 +uniform,exp_fundamental,2.099,2.8374 +uniform,exp_fundamental,2.2,2.9834 +uniform,exp_fundamental,2.299,2.7744 +uniform,exp_fundamental,2.399,3.0816 +uniform,sim,0.3,0.1133 +uniform,sim,0.4,0.2442 +uniform,sim,0.5,0.3021 +uniform,sim,0.6,0.4708 +uniform,sim,0.7,0.6521 +uniform,sim,0.8,0.8786 +uniform,sim,0.9,0.856 +uniform,sim,1.0,0.9944 +uniform,sim,1.1,1.4099 +uniform,sim,1.2,1.6113 +uniform,sim,1.3,1.5307 +uniform,sim,1.4,2.0569 +uniform,sim,1.5,2.2734 +uniform,sim,1.6,2.1198 +uniform,sim,1.7,2.3061 +uniform,sim,1.8,2.5932 +uniform,sim,1.9,2.8273 +uniform,sim,2.0,3.0262 +uniform,sim,2.1,3.1017 +uniform,sim,2.2,3.1219 +uniform,sim,2.3,3.3635 +uniform,sim,2.4,3.789 +uniform,exp_total,0.299,0.1612 +uniform,exp_total,0.399,0.2468 +uniform,exp_total,0.499,0.5061 +uniform,exp_total,0.599,0.5438 +uniform,exp_total,0.699,0.7377 +uniform,exp_total,0.799,0.7704 +uniform,exp_total,0.899,1.0348 +uniform,exp_total,0.999,1.2312 +uniform,exp_total,1.099,1.5433 +uniform,exp_total,1.199,1.7297 +uniform,exp_total,1.299,1.8656 +uniform,exp_total,1.399,2.1551 +uniform,exp_total,1.499,2.1123 +uniform,exp_total,1.599,2.4975 +uniform,exp_total,1.699,2.1879 +uniform,exp_total,1.799,2.2785 +uniform,exp_total,1.899,2.8676 +uniform,exp_total,1.999,2.7241 +uniform,exp_total,2.099,4.0509 +uniform,exp_total,2.2,3.8293 +uniform,exp_total,2.299,3.3082 +uniform,exp_total,2.399,4.7659 +sheared,exp_fundamental,0.299,0.1117 +sheared,exp_fundamental,0.399,0.1117 +sheared,exp_fundamental,0.499,0.1522 +sheared,exp_fundamental,0.599,0.228 +sheared,exp_fundamental,0.699,0.3884 +sheared,exp_fundamental,0.799,0.3602 +sheared,exp_fundamental,0.899,0.5329 +sheared,exp_fundamental,0.999,0.5876 +sheared,exp_fundamental,1.099,0.7056 +sheared,exp_fundamental,1.199,0.8237 +sheared,exp_fundamental,1.299,1.0845 +sheared,exp_fundamental,1.399,1.3207 +sheared,exp_fundamental,1.499,1.6309 +sheared,exp_fundamental,1.599,1.5533 +sheared,exp_fundamental,1.699,1.9075 +sheared,exp_fundamental,1.799,1.9516 +sheared,exp_fundamental,1.899,2.0626 +sheared,exp_fundamental,1.999,1.8212 +sheared,exp_fundamental,2.099,2.646 +sheared,exp_fundamental,2.2,2.1702 +sheared,exp_fundamental,2.299,2.5931 +sheared,exp_fundamental,2.399,2.5667 +sheared,sim,0.3,0.05 +sheared,sim,0.4,0.1029 +sheared,sim,0.5,0.1752 +sheared,sim,0.6,0.2069 +sheared,sim,0.7,0.3056 +sheared,sim,0.8,0.4184 +sheared,sim,0.9,0.4712 +sheared,sim,1.0,0.6034 +sheared,sim,1.1,0.7655 +sheared,sim,1.2,0.8272 +sheared,sim,1.3,0.9999 +sheared,sim,1.4,1.1744 +sheared,sim,1.5,1.3594 +sheared,sim,1.6,1.4529 +sheared,sim,1.7,1.6467 +sheared,sim,1.8,1.7507 +sheared,sim,1.9,1.904 +sheared,sim,2.0,2.0415 +sheared,sim,2.1,2.1948 +sheared,sim,2.2,2.3746 +sheared,sim,2.3,2.4944 +sheared,sim,2.4,2.646 +sheared,exp_total,0.299,0.1152 +sheared,exp_total,0.399,0.1135 +sheared,exp_total,0.499,0.1699 +sheared,exp_total,0.599,0.2844 +sheared,exp_total,0.699,0.4342 +sheared,exp_total,0.799,0.4113 +sheared,exp_total,0.899,0.577 +sheared,exp_total,0.999,0.6263 +sheared,exp_total,1.099,0.7514 +sheared,exp_total,1.199,0.8801 +sheared,exp_total,1.299,1.1568 +sheared,exp_total,1.399,1.3753 +sheared,exp_total,1.499,1.6802 +sheared,exp_total,1.599,1.5974 +sheared,exp_total,1.699,2.2723 +sheared,exp_total,1.799,2.0098 +sheared,exp_total,1.899,2.0996 +sheared,exp_total,1.999,1.9868 +sheared,exp_total,2.099,3.1412 +sheared,exp_total,2.2,2.2177 +sheared,exp_total,2.299,2.6495 +sheared,exp_total,2.399,2.6354