What industries does newtsim serve?

newtsim serves oil and gas (flow assurance, drilling fluids, subsea pipeline), pharmaceutical and drug discovery, energy storage (battery materials, electrolytes), semiconductor process simulation, green hydrogen (electrolyser and catalyst design), nuclear decommissioning, aerospace and composites, corrosion and materials, carbon capture, agrochemicals, mining and geotechnical, and fermentation and bioprocessing.

How is newtsim different from standard CFD or simulation software?

Standard simulation tools use empirical models at a single scale. newtsim runs the full physics chain from quantum mechanics and molecular dynamics through to plant-scale CFD and structural mechanics, GPU-native, with direct data flow between every scale and no third-party solver dependencies. Constitutive behaviour is predicted from first principles rather than empirically fitted.

What does a fixed-price simulation study include?

A newtsim fixed-price study has a defined scope, clear deliverables, fixed cost agreed upfront, and is delivered in weeks rather than quarters. Studies typically include a scoping call, physics-grounded simulation using the appropriate newtsim solvers, results analysis, and a written report with actionable recommendations.

What is physics-grounded simulation?

Physics-grounded simulation derives material properties and constitutive behaviour from lower-scale physics (molecular dynamics and quantum mechanics) rather than from empirical fitting. This gives accurate predictions for novel materials and extreme conditions outside the range of experimental data, and enables root-cause analysis at the mechanistic level.

How does newtsim approach oil and gas flow assurance simulation?

newtsim runs a continuous physics chain from molecular wax crystallisation to full pipeline transient multiphase flow, with no empirical handoffs between scales. Wax appearance temperatures, gel strength, and deposition kinetics are predicted from molecular dynamics (newtsim Bond), passed directly to mesoscale morphology (newtsim Pack), and then propagated into multiphase CFD (newtsim Stream), giving flow assurance predictions that hold outside the range of laboratory data.

Can newtsim simulate battery electrolyte behaviour and degradation?

Yes. The Bond solver runs molecular dynamics for Li-ion transport and solid-electrolyte interphase (SEI) formation; Pack handles electrode morphology at the mesoscale; these results couple upward to cell-scale electrochemical models. This gives physics-grounded predictions of electrolyte degradation, capacity fade, and thermal behaviour without relying solely on empirical ageing fits.

How does physics-grounded drug discovery simulation differ from docking software?

Molecular docking software treats the protein and ligand as rigid bodies and scores binding poses by geometry. Physics-grounded simulation uses molecular dynamics to capture conformational flexibility and explicit solvation effects that rigid-body docking misses, predicting binding free energy, ADMET properties, and metabolic activation from first principles. This is particularly valuable for targets where induced-fit binding or solvent-mediated interactions drive selectivity.

What makes GPU-native simulation faster than CPU-based CFD?

GPU parallelism maps naturally to the computational kernels underlying molecular dynamics and CFD: large numbers of independent force or flux calculations that can execute simultaneously. newtsim runs all simulation scales GPU-native, avoiding the CPU-to-GPU data transfers that bottleneck traditional solvers which run physics on CPU and only offload selected operations to the GPU. This removes the transfer overhead and allows the full multiscale chain to run without the latency penalties of heterogeneous pipelines.

Does newtsim replace existing simulation tools like ANSYS or COMSOL?

newtsim fills the multiscale gap, connecting molecular-scale physics to the engineering-scale models that tools like ANSYS operate at. It is not a replacement for FEA-only or single-scale workflows where those tools are the right fit. Where newtsim adds value is in problems where constitutive behaviour, material properties, or interfacial physics need to be derived from lower-scale simulation rather than taken from a handbook or empirical fit.

How long does a fixed-price simulation study take?

Most studies complete in 4 to 8 weeks depending on scope. The standard timeline is: scoping call and problem definition in week 1; simulation execution across the relevant physics scales in weeks 2 to 4; results analysis, validation, and report delivery in weeks 5 to 6. More complex multiscale studies with broader operating envelopes may extend to 8 weeks. Scope and timeline are agreed and fixed before work begins.

What deliverables does a simulation study produce?

A newtsim simulation study delivers: a validated multiscale model calibrated to your system; an operating envelope with physics-based safety margins; a written technical report with findings and actionable recommendations; and optionally CFD visualisations or an ML surrogate model (newtsim Neural) for fast design-space exploration by your engineering team.

Does newtsim work for green hydrogen electrolyser and catalyst design?

Yes. The Stream solver handles proton exchange membrane and alkaline electrolyser CFD: multiphase flow, bubble dynamics, and membrane transport. Bond handles catalyst surface reaction molecular dynamics, giving adsorption energies and reaction barriers from first principles rather than from empirical correlations. Neural then produces fast-to-run surrogate models for design-space exploration and stack optimisation, reducing the iteration cost for new catalyst compositions and electrode architectures.

newtsim is a physics-grounded multiscale simulation consulting firm and software platform. We deliver fixed-price simulation studies and platform licences for industrial clients, covering the full physics chain from quantum mechanics and molecular dynamics through to plant-scale CFD and structural mechanics. All solvers run GPU-native.

Does newtsim use artificial intelligence or machine learning?

Yes. newtsim Neural produces ML-trained surrogate models derived from high-fidelity physics simulations. These surrogates run in milliseconds rather than hours, enabling rapid design-space exploration and real-time control without sacrificing physical accuracy. The ML models are trained on physics simulation data, not empirical measurements, so they inherit the accuracy and extrapolation properties of the underlying physics.

How do I get started with newtsim?

Book a free technical scoping call at newtsim.com/demo. Bring a description of your engineering challenge, the physics you think are important, and any data you have. The call is 45 minutes and produces a problem definition, proposed physics chain, and outline scope that forms the basis of a fixed-price proposal.

Can newtsim simulate non-Newtonian fluids?

Yes. newtsim specialises in non-Newtonian flow: yield-stress fluids, viscoelastic materials, shear-thinning and shear-thickening systems, and complex multiphase flows. The Stream solver handles macroscale CFD for these systems; Bond provides the molecular-scale rheology predictions that replace empirical viscosity fits; and the full chain gives accurate behaviour outside the calibrated range of laboratory data.

Does newtsim work for small companies or startups?

Yes. Fixed-price studies are available to clients of any size. The fixed-price, fixed-scope model means you know the cost before committing, making it accessible for smaller organisations without large simulation teams. Many newtsim clients are startups and scale-ups who need physics-grounded analysis early in product development before investing in laboratory or pilot-scale work.

What is the difference between molecular dynamics and CFD?

Molecular dynamics (MD) simulates the motion of individual atoms and molecules to compute material properties such as viscosity, diffusivity, reaction rates, and phase behaviour, from first principles. Computational fluid dynamics (CFD) models bulk fluid flow at the engineering scale using those material properties as inputs. newtsim connects the two: MD provides accurate constitutive inputs that CFD uses for engineering-scale predictions, rather than relying on empirical correlations.

What is a surrogate model in simulation?

A surrogate model (also called a reduced-order model or emulator) is a fast mathematical approximation of a complex simulation. newtsim Neural builds ML surrogates trained on high-fidelity physics simulation data. Once trained, a surrogate evaluates in milliseconds instead of hours, enabling real-time optimisation and design-space exploration. Because surrogates are trained on physics data, they are more accurate than empirical correlations, especially outside the training range.

What deliverables does a newtsim simulation study produce?

A newtsim simulation study delivers: (1) a validated multiscale model calibrated to your system; (2) an operating envelope with physics-based safety margins; (3) a written technical report with findings and actionable recommendations; and optionally (4) CFD visualisations and an ML surrogate model for fast ongoing design-space exploration by your engineering team.

Corrosion & Materials Degradation

Corrosion & Materials Degradation · Case Study

Stress Corrosion Cracking Susceptibility Assessment for High-Strength Suspension Bridge Cables

Executive Summary

At 5:04 PM on 15 December 1967, the Silver Bridge collapsed into the Ohio River during peak rush hour. Thirty-one vehicles were crossing. Forty-six people died. The span fell in seconds — a statically determinate eyebar chain, so when one link broke, the entire structure followed. The National Transportation Safety Board investigation identified the cause as a stress-corrosion crack in eyebar 330, the northernmost link below the Ohio tower. The crack had initiated at a corrosion pit on the interior bore of the pin-hole — in a region of high triaxial stress exactly where the pin contacted the steel. Its depth at failure was approximately 2.5 mm. The steel's Charpy impact energy was 4–8 J at 0 C, placing it far below the 27 J required for fracture-critical members under current standards. With those material properties and that operating stress, a 2.5 mm crack was enough for fast fracture. Metallurgical analysis found chloride contamination from road de-icing salt accumulated in the pin-hole crevice over decades, and phosphorus grain boundary segregation promoting intergranular SCC via hydrogen-assisted decohesion. The bridge had been in service 39 years. No sub-surface inspection of any eyebar pin-hole bore had ever been conducted. The crack had likely been growing for 12–20 years.

The Silver Bridge collapse triggered the Federal-Aid Highway Act of 1968 and the National Bridge Inspection Program — mandating biennial visual inspection of all federally-funded bridges. Visual inspection. The NDE detection capability gap for sub-surface SCC cracks in fracture-critical members was not addressed for decades. In 2007, the I-35W bridge in Minneapolis collapsed from a gusset plate failure, killing 13, demonstrating that the structural vulnerability class had not been eliminated.

Had a predictive SCC simulation been commissioned on the Silver Bridge in the 1950s — at the fifteen-year mark, when the crack was likely still in the millimetre range — it would have identified the problem with specificity. The combined MD/LEFM model applied here to the equivalent geometry and environment yields a median time from 0.5 mm pit initiation to critical crack size of 16.4 years (P10: 9.1 yr, P90: 28.7 yr) — squarely within the NTSB's estimated 12–20 year growth period. It would have identified that the Stage II SCC velocity for this steel-environment combination, with the literature spanning two orders of magnitude, is in practice 3.8 × 10⁻⁹ m/s at the operating stress intensity — narrowing the management decision from "somewhere between 5 months and 41 years" to "4–17 months" for a crack of the depth found on the Henderson Memorial Bridge.

This study is commissioned by a state bridge authority operating six aging eyebar-chain suspension bridges constructed between 1928 and 1964, four using the same high-strength steel class implicated in the Silver Bridge collapse. A federal fracture-critical member review has flagged three structures. Magnetic particle inspection of the Henderson Memorial Bridge has found 14 indications on 6 eyebars, the deepest at 3.1 mm — within the uncertain K_ISCC band for this material and environment. The critical flaw size for fast fracture is 3.5 mm. The study quantifies the SCC velocity, narrows its uncertainty from two orders of magnitude to a factor of 5.7 at 90% confidence, estimates time-to-failure distributions for all 14 detected indications, and delivers an inspection protocol maintaining failure probability below 10⁻⁴ per year per eyebar. The highest-risk locations identified by the model define where newtsim livesim nodes should be placed: electrochemical noise sensors and continuous strain monitoring at the pin-hole bores of the critical eyebars, providing real-time SCC activity detection that keeps the authority ahead of sub-surface crack growth between inspection campaigns.

Scenario Background (illustrative reference case)

Operator (fictional): Tri-State Bridge Authority — responsible for six suspension bridges across the Ohio and Kanawha river valleys, West Virginia, Ohio, and Kentucky, constructed between 1928 and 1964. State-funded authority operating under federal oversight from FHWA and the National Bridge Inspection Program.

Primary structure of concern: The Henderson Memorial Bridge has a 512 m main span with a two-tower eyebar chain suspension system fabricated from high-strength heat-treated carbon steel eyebars (ASTM A517-class equivalent, estimated yield strength sigma_y of approximately 690 MPa, UTS approximately 827 MPa). Each chain consists of 12 parallel eyebars per panel point, each eyebar 25 mm thick x 450 mm wide x 6.4 m long, connected at pin-hole joints with 200 mm diameter forged steel pins. The bridge carries a four-lane divided highway and has been in continuous service for 60 years.

Material composition: The steel composition is estimated from archival Bethlehem Steel mill certificates and 2022 drill-core spectroscopy: Fe--0.42C--0.91Mn--0.19Si (wt%), with traces of P (0.018 wt%) and S (0.022 wt%). The steel was quenched and tempered to achieve the target tensile properties. Modern Charpy impact testing on drill-core samples gives CVN values of 12--18 J at 0 C — well below the 27 J minimum required for fracture-critical applications under current AASHTO specifications. This low toughness is intrinsic to the high carbon content and 1960s-era heat treatment practice, and it is the primary reason SCC is an acute structural risk for this class of bridge.

Environment: The Henderson Memorial Bridge crosses the Kanawha River in an industrial river valley with historical SO2 pollution from upstream chemical plants. The relevant corrosion environment for the lower eyebars comprises acid deposition at pH 4.1--4.8 (measured at bridge monitoring station, 2020--2024 average), chloride deposition rate of 30--80 mg/m2/day on lower eyebar faces (road de-icing salt splash from deck drainage, highest in winter months), relative humidity exceeding 80% RH for 2,800 hours/year (sufficient for sustained aqueous film formation on steel), and temperature cycling from -18 C to +38 C (annual range, with thermal stresses superimposed on dead load).

Loading: Dead load stress in the most heavily loaded eyebars (fourth panel point from tower, south chain) is sigma_DL = 415 MPa (60% sigma_y). Maximum live load stress addition under a legal load convoy is Delta-sigma_LL = 55 MPa, producing a peak total stress of 470 MPa (68% sigma_y). A fatigue spectrum recorded by a 36-channel wireless strain gauge array installed in 2015 shows 2.3 million cycles exceeding 10 MPa stress range recorded over 10 years, with the highest cycles in the 15--35 MPa range — within the fatigue crack propagation threshold for this steel at K_I >= 5 MPa-root-m, meaning fatigue and SCC are simultaneously active mechanisms in moist conditions.

Inspection findings (2024 MPI campaign): Magnetic particle inspection of all accessible eyebar faces identified 14 indications on 6 eyebars: 9 indications classified as less than 1 mm depth, 4 indications classified as 1--3 mm depth, and 1 indication at eyebar C4-S (fourth panel, south chain, pin-hole bore, inside surface) at 3.1 mm depth with axial orientation consistent with SCC morphology (branching, intergranular in visible SEM section taken from a drill plug).

Challenge

The Silver Bridge failed without warning from a crack estimated at only 2.5 mm depth. The Henderson Memorial Bridge presents an analogous risk profile, and the critical technical challenge is that the literature range for Stage II SCC velocity in high-strength steel/acidic chloride environments spans two full orders of magnitude (10⁻¹⁰ to 10⁻⁸ m/s). This uncertainty makes any empirical remaining life calculation essentially meaningless:

SCC Velocity Assumption	Time from 3.1 mm Detected Flaw to Critical Size (3.5 mm)	Management Implication
Optimistic (10⁻¹⁰ m/s)	41 years	No urgency; 5-yr inspection interval acceptable
Mid-range (10⁻⁹ m/s)	4.1 years	Immediate enhanced monitoring required
Pessimistic (10⁻⁸ m/s)	5 months	Emergency structural intervention required now

The physics-based simulation must narrow this range to the level where the required action (monitor, replace within months, or replace immediately) is unambiguous.

The eyebar pin-hole geometry creates a severe stress concentration: pin-hole radius 100 mm, eyebar bore diameter 200 mm, elastic stress concentration factor at pin-hole bore Kt = 3.2 (from FEM analysis of pin contact). Under dead load sigma_DL = 415 MPa, the net section stress at the pin-hole bore reaches 415 x 3.2 = 1,328 MPa — well above the nominal yield strength of 690 MPa. The eyebar is designed to yield locally at the pin-hole in service (the net section at the bore carries the full chain load), and this local plastic zone is exactly where the Silver Bridge crack initiated.

Fracture mechanics parameters for the assessment include a plain-strain fracture toughness K_IC estimated at 55 MPa-root-m from Charpy-CVN correlation (Barsom-Rolfe upper bound) using CVN = 15 J at 20 C. The critical crack depth for fast fracture at total stress 470 MPa is therefore a_c = (K_IC / (Y x sigma))² / pi = (55 / (1.12 x 470))² / pi = 3.5 mm. The K_ISCC literature range for ASTM A517-class steel in acidic chloride is 22--35 MPa-root-m. At operating stress 470 MPa with the 3.1 mm detected flaw, K_I = Y x sigma x sqrt(pi x a) = 1.12 x 470 x sqrt(pi x 0.0031) = 52.0 MPa-root-m — above the upper end of the K_ISCC range. Eyebar C4-S is therefore assessed as having a crack that has already entered or is about to enter Stage II SCC propagation.

Current MPI probability of detection (POD_90) — the depth at which a crack is detected with 90% probability — is 2.8 mm for the in-service configuration (from POD curve developed for similar bridge steel, surface painted, FHWA-RD-01-161). The Silver Bridge fatal crack was 2.5 mm deep — below the current POD_90. Four of the 14 indications are in the 1--3 mm range where detection is probabilistic, not certain.

Real-World Basis

The 1967 Silver Bridge Collapse: On 15 December 1967 at 5:04 PM, during peak evening rush hour, the Silver Bridge (US Route 35, connecting Point Pleasant, West Virginia, and Gallipolis, Ohio) collapsed into the Ohio River. Thirty-one vehicles were on the bridge; 46 people died (two bodies were never recovered). It was the deadliest bridge disaster in US history since the 1940 Tacoma Narrows collapse.

NTSB Investigation Findings (NTSB-HAR-71-1, 1971): The National Transportation Safety Board's investigation, supported by detailed metallurgical analysis at the National Bureau of Standards, reached the following conclusions. The initiating defect was a stress-corrosion crack in eyebar 330 — the northernmost eyebar in the Ohio subsidiary chain, first link below the top of the Ohio tower. The crack initiated at a corrosion pit on the interior surface of the eyebar pin-hole bore, in a region of high triaxial stress created by the pin contact geometry. The critical crack depth at failure was approximately 2.5 mm (0.1 inches), growing through the combination of SCC and corrosion fatigue over an estimated 12--20 years of service from first pit initiation. The steel had a Charpy impact energy of approximately 4--8 J at 0 C (very low toughness by modern standards), giving a retrospective K_IC estimate of 45--55 MPa-root-m — consistent with fast fracture from a shallow crack at the operating stress level. The eyebar chain was statically determinate: the failure of one eyebar transferred the full load to the adjacent parallel eyebar in the same panel, which immediately failed by ductile overload, initiating a progressive collapse of the entire span in seconds. No systematic sub-surface inspection had been conducted on any of the eyebar pin-hole bores during the 39 years of bridge service; visual inspection could not detect the crack at the interior bore surface.

Metallurgical Analysis: Post-collapse fractographic analysis of the failed eyebar identified a distinct boundary between the SCC-propagated zone (intergranular, flat fracture surface, oxidised) and the overload fracture zone (dimpled ductile fracture). Chloride content of 0.08 wt% was found in the corrosion products at the crack initiation site — consistent with long-term accumulation of road de-icing salt in the pin-hole crevice. Grain boundary decoration by phosphorus segregation (P = 0.018 wt% in the steel) was identified as a known promoter of intergranular SCC in high-strength steels via hydrogen-assisted grain boundary decohesion.

Regulatory Response: The Silver Bridge collapse directly triggered the Federal-Aid Highway Act of 1968, which established the National Bridge Inspection Program requiring biennial inspection of all federally-funded highway bridges. However, the act mandated visual inspection only — the NDE detection capability gap for sub-surface SCC cracks in fracture-critical members was not addressed for decades.

FHWA Economic Impact: Subsequent federal investment in bridge safety triggered by the Silver Bridge collapse exceeded $5 billion in emergency infrastructure spending in the following decade. The 2007 I-35W Mississippi River Bridge collapse in Minneapolis (13 fatalities,$ 60 million direct cost, $400 million replacement) demonstrated that fracture-critical member failures continued to represent a systematic risk 40 years later.

Simulation Approach

The simulation strategy addresses the fundamental multi-scale nature of SCC: initiation depends on atomistic hydrogen-metal interactions (nanometre scale); propagation is governed by crack-tip stress fields and hydrogen diffusion (micrometre to millimetre); structural consequence depends on bridge-level mechanics (metre to hundred-metre scale). The methodology follows the STRETCH classification for direct MD-to-LEFM coupling and instead uses the defensible and published MD-informed cohesive zone model (CZM) approach for scale bridging.

Stage 1 — Molecular Dynamics: Hydrogen Embrittlement at Crack Tip

ReaxFF MD simulations model hydrogen diffusion and accumulation at the crack-tip process zone. The Fe/H/C force field is validated against DFT crack-tip hydrogen energetics. The simulation domain is a 50 nm x 50 nm x 5 nm Fe BCC supercell with a pre-notch of depth 2 nm — large enough to capture the triaxial stress zone ahead of the crack tip where hydrogen accumulates. Applied stress intensity K_I is mapped from far-field loading to the simulation boundary.

The aqueous corrosion reaction at the crack mouth (pH 4.2, 500 ppm Cl-) generates a flux of atomic hydrogen from the hydrogen evolution reaction. The MD models the full hydrogen transport pathway: adsorption, subsurface entry, and diffusion from the crack mouth to the triaxial stress zone. Trap sites sampled include interstitial sites, dislocations, grain boundaries (modelled as a high-angle boundary 20 nm from the crack tip), and carbide/matrix interfaces.

Key outputs are the hydrogen concentration at crack tip C_H as a function of K_I (stress-enhanced diffusion, with trapping kinetics parameterised by MD trap energies), the effective surface energy gamma_eff(C_H) via the hydrogen-enhanced decohesion (HEDE) model (gamma_eff = gamma_0 - Gamma_H x C_H, where Gamma_H is the Gibbs adsorption coefficient for H at the {110} cleavage plane, computed from DFT: Gamma_H = 1.4 x 10⁻⁵ J/m2 per atom/nm2), and the bond-breaking rate as a function of C_H and applied stress, providing the microscopic basis for the Stage II SCC velocity expression.

Stage 2 — DFT Calibration

newtsim Root calculations compute hydrogen segregation energies at the grain boundary and at the Fe{110} cleavage plane. The segregation energy at the grain boundary (Delta-E_seg = -0.43 eV) and the decohesion work reduction per H monolayer (Delta-W_sep = -0.38 J/m2) directly parameterise the CZM traction-separation law. The ReaxFF force field reproduces the DFT grain boundary segregation energy to within 0.06 eV — acceptable accuracy for the subsequent MD-scale calculations.

Stage 3 — Cohesive Zone Model FEM

The CZM approach is used instead of direct MD-to-LEFM coupling because it provides a continuous, mesh-objective fracture mechanics implementation with demonstrated agreement against experimental data for hydrogen-assisted cracking in high-strength steel. The CZM traction-separation law is of the form:

$T_n = T_{max} \cdot \left(\frac{\delta_n}{\delta_0}\right) \cdot \exp\!\left(1 - \frac{\delta_n}{\delta_0}\right) \cdot \left(1 - \frac{C_H}{C_H^{crit}}\right)^n$

where T_max is the cohesive strength (reduced from the H-free value of 2.7 GPa by the HEDE factor), delta_0 is the critical separation, and the final factor captures the progressive reduction in cohesive strength with hydrogen concentration (calibrated n = 1.8 from DFT data). The CZM is implemented in newtsim Span using zero-thickness cohesive elements along the crack plane, with a 3D model of the eyebar pin-hole geometry meshed at 0.1 mm resolution at the crack tip.

Stage 4 — LEFM Crack Growth Integration

Stage II SCC velocity (constant velocity plateau) is derived from the MD-computed gamma_eff via the thermally-activated model:

$v_{SCC} = A \cdot \exp\!\left(-\frac{Q - \alpha K_I}{RT}\right)$

where Q = 68 kJ/mol is the activation energy for anodic dissolution at the crack tip (from MD Fe dissolution kinetics at pH 4.2), alpha = 0.42 m3/mol captures the stress-assisted component (from the K_I dependence of MD dissolution rate), T is temperature, and A is the pre-exponential factor calibrated to one published slow-strain-rate test result for ASTM A517 in 3.5% NaCl at pH 4: this calibration gives A = 3.2 x 10⁻³ m/s. The activation energy Q is consistent with anodic dissolution theory for Fe in acidic chloride.

Crack growth integration over the loading spectrum:

$a(t) = a_0 + \int_0^t \left[v_{SCC}(K_I(a(\tau), \sigma_{DL})) + v_{fatigue}(\Delta K(\tau))\right] d\tau$

where the fatigue crack growth contribution uses Paris law (C = 3.2 x 10⁻¹² m/cycle, m = 3.0, for the R approximately 0.8 corrosion-fatigue regime established for similar steels) and the recorded strain gauge spectrum.

Stage 5 — Probabilistic Failure Assessment

Monte Carlo simulation (n = 50,000) propagates uncertainties from five sources: initial flaw depth a_0, sampled from NDE detection probability curve (truncated exponential with POD_90 = 2.8 mm for MPI; flaws below 0.5 mm assumed undetectable); K_ISCC threshold as a Normal distribution with mean 24.5 MPa-root-m (MD-derived) and sigma = 7 MPa-root-m (plus/minus 30% from literature variability); SCC velocity model uncertainty as a Log-normal multiplicative factor with geometric mean 1.0 and GSD = 2.0 (factor of 2 uncertainty in calibration constant A); traffic loading spectrum from bootstrap resampling of the 10-year strain gauge record, reflecting inter-year variability in heavy vehicle traffic; and temperature as an annual cycle (sinusoidal, mean 10 C, amplitude 28 C), affecting the Arrhenius rate in v_SCC.

Key Predictions / Results

Narrowed SCC Velocity Range

Parameter	Literature Range	MD-Derived Value	Uncertainty (90% CI)	Reduction Factor
Stage II SCC velocity at K_I = 28 MPa-root-m	10⁻¹⁰--10⁻⁸ m/s	3.8 x 10⁻⁹ m/s	1.6--9.1 x 10⁻⁹ m/s	100x -> 5.7x
K_ISCC threshold	22--35 MPa-root-m	24.5 MPa-root-m	18--31 MPa-root-m	13 MPa range -> 13 MPa range (better centred)
Activation energy Q	50--90 kJ/mol	68 kJ/mol	61--75 kJ/mol	Constrained

Flaw Disposition Matrix — Time to Critical Crack Size (3.5 mm)

Eyebar ID	Current Depth (mm)	K_I (MPa-root-m)	P10 Time to Failure (months)	P50 Time to Failure (months)	P90 Time to Failure (months)	5-yr Failure Probability	Recommended Action
C4-S (deepest)	3.1	29.0	4.1	8.4	17.2	0.94	Replace immediately
C3-N	2.8	26.3	6.8	14.2	28.7	0.76	Replace within 3 months
C5-S	2.6	24.4	9.3	19.8	38.1	0.58	Replace within 6 months
C2-N	2.4	22.5	13.1	27.4	54.2	0.41	Replace within 12 months
C6-N	2.2	20.7	18.2	37.9	74.8	0.28	Re-inspect in 6 months (ACFM)
Indications 6--9 (1--2 mm range)	1.0--1.9	9.4--17.8	27--68	58--141	>120	0.08--0.18	Re-inspect annually (ACFM)
Indications 10--14 (< 1 mm)	< 1.0	< 9.4	>72	>156	>300	< 0.05	MPI biennial per current schedule

5-year failure probability by eyebar indication — AASHTO FCM limit and 40% replacement threshold

SCC velocity range narrowing — literature uncertainty reduced from 100x to 5.7x via MD simulation

The 5-year failure probabilities for the top four indications all exceed 0.40 — far above the AASHTO FCM criterion of 10⁻⁴ per year (0.05% per year, or approximately 0.05% cumulative over 5 months). Immediate action is unambiguous for eyebar C4-S; the next three require phased replacement within 12 months.

Silver Bridge Back-Analysis (Model Validation)

The MD/LEFM model applied to Silver Bridge geometry (eyebar 330: sigma = 446 MPa estimated from design records, K_t = 3.4 at the deeper pin-hole configuration, A7 steel composition, pH 4.5 environment) produces a simulated median time from pit initiation (a_0 = 0.5 mm) to critical crack size (a_c = 4.2 mm) of 16.4 years (P10: 9.1 yr, P90: 28.7 yr). The NTSB estimated the crack grew over 12--20 years of service from first pit formation (eyebar installed 1928, bridge opened, service estimated from corrosion product dating). The model median falls within the NTSB estimated range — a strong validation for the fundamental correctness of the SCC velocity parameterisation.

Required NDE Capability

Current MPI POD_90 = 2.8 mm is insufficient: eyebar C3-N (2.8 mm depth, P50 failure time 14 months) could be missed by the current inspection. To maintain failure probability below 10⁻⁴/year across the entire bridge, NDE must achieve POD_90 <= 1.5 mm. Alternating current field measurement (ACFM) achieves POD_90 = 1.2 mm for in-service surface cracks in steel of this thickness per FHWA ACFM trials on comparable bridge components — this is the recommended technology upgrade.

Comparison Methodology

MD-to-CZM internal validation (primary): The ReaxFF MD hydrogen segregation energies and decohesion work values directly parameterise the CZM traction-separation law. The DFT calculations validate the MD force field to within 0.06 eV at the grain boundary — the higher-fidelity DFT result anchoring the lower-fidelity ReaxFF parameters used in the CZM. The CZM-derived K_ISCC of 24.5 MPa-root-m and Stage II velocity of 3.8 x 10⁻⁹ m/s at K_I = 28 MPa-root-m are the primary simulation outputs carried forward into the LEFM crack growth integration.
Silver Bridge back-analysis: The MD/LEFM model applied to Silver Bridge geometry produces a simulated median time from pit initiation to critical crack size of 16.4 years (P10: 9.1 yr, P90: 28.7 yr). The NTSB estimated 12--20 years of growth. The model median falls within the historical estimate — the primary long-timescale validation for the full simulation chain.
Secondary confirmation — coupon testing: Drill-core SSR testing in synthetic bridge environment (pH 4.2, 500 ppm Cl-) gives measured K_ISCC of 21--28 MPa-root-m and Stage II velocity of 2.4--6.8 x 10⁻⁹ m/s at K_I = 28 MPa-root-m. The simulation-derived values (24.5 MPa-root-m and 3.8 x 10⁻⁹ m/s) fall within these experimental ranges. Published SCC velocity compilations for similar high-strength steels in 3.5% NaCl report 1.0--9.0 x 10⁻⁹ m/s at this K_I range, further confirming the simulation output. The simulation value sits at a position consistent with the lower yield strength (690 MPa) of the Henderson steel, as expected.

Simulation Caveats

Classification: STRETCH. Direct MD simulation of hydrogen embrittlement at a crack tip with LEFM coupling is computationally demanding. The simulation approach adopted here uses the MD-informed cohesive zone model (CZM) methodology rather than direct atomistic-to-LEFM coupling, for several reasons. Realistic crack-tip stress fields require simulation cells of 10⁶+ atoms to avoid artificial boundary effects, which is at the upper limit of tractable MD for useful trajectory lengths. Bridging atomistic hydrogen trapping energies to continuum fracture mechanics (LEFM or EPFM) via direct coupling remains an active research area with significant uncertainty in scale-bridging protocols. The CZM approach (MD provides H-dependent decohesion energies; FEM CZM propagates crack using these energies) is established in the hydrogen-assisted cracking literature with demonstrated agreement against experimental data for high-strength steel SCC. Finally, the LEFM crack growth integration using the Arrhenius-form v_SCC equation is calibrated against SSR coupon tests on the actual bridge material, providing a direct experimental anchor that limits reliance on pure MD extrapolation.

The 90% confidence intervals on v_SCC (factor of 5.7 at K_I = 28 MPa-root-m) remain wider than would be ideal; however, they are sufficient to make unambiguous go/no-go decisions for the critical indications (eyebars C4-S through C3-N), where even the optimistic P90 time-to-failure is shorter than any feasible inspection interval.

Deliverables

MD simulation report: Hydrogen embrittlement database — gamma_eff vs. K_I and pH across the temperature range -10 to +40 C; crack-tip C_H maps at K_I = 20, 28, and 35 MPa-root-m; H diffusion coefficient as a function of trap density (dislocation and grain boundary content); DFT-calibrated CZM traction-separation law parameters in tabulated form.
LEFM crack growth code: Python solver with probabilistic Monte Carlo wrapper (n = 50,000 default); documented for independent audit; includes the SSR coupon calibration dataset and back-analysis of Silver Bridge as worked validation examples.
Flaw disposition matrix: Each of the 14 detected indications assessed with: current K_I, predicted time to critical size (P10/P50/P90 months), 5-year failure probability, and recommended action category (replace immediately / replace within N months / re-inspect at N months / continue biennial MPI).
Inspection protocol recommendation: Specification for ACFM deployment — probe configuration, scanning pattern for eyebar pin-hole bores, reference calibration standards, required POD_90 <= 1.5 mm, and inspection frequency to maintain failure probability below 10⁻⁴/year per eyebar (calculated as: annual ACFM inspection of all eyebars with detected or suspected indications; biennial MPI of remainder).
AASHTO LRFD bridge evaluation compliance memo: Mapping of simulation results to AASHTO MBE Section 6 fracture-critical member provisions, including documentation of the Level III (refined analysis) approach used per AASHTO commentary C6.2.3.
Executive briefing deck: 15-slide summary for state DOT leadership and FHWA federal oversight, including: collapse consequence analysis (Silver Bridge precedent), simulation methodology summary, flaw disposition decisions, phased replacement cost estimate ($2.4M for 5 eyebar replacements), and residual risk quantification.

This case study is an illustrative reference scenario demonstrating newtsim's simulation methodology. All company names, personnel, and specific operational data are fictional. The incident descriptions draw on publicly documented real-world events cited in the frontmatter.