01 / Research overview

Research

Fatigue assessment, critical plane methods, welded structures, additive manufacturing and component testing at the University of Pisa.

How materials and components respond to repeated loads, studied with analytical models, finite element analysis and experiments.
FIG. R01Research overview
Research overview

Fatigue assessment

Fatigue of components under loads acting in several directions. I develop and compare critical plane methods that link local stresses and strains to fatigue life, for notched and welded components.

Methods Analytical and semi-analytical models; finite element post-processing; comparison with fatigue tests.

Further studies and technical context
FIG. R02 / Candidate-plane

FIG. R02.1SeAn-CP search paths

Three great circles instead of the whole sphere

Full-sphere maximum FSext
–
SeAn-CP maximum
–
SeAn-CP error
–
Plane families
–
  • SeAn-CP plane families: strain range, stress at t₀, stress at t₀+π
  • full-sphere maximum
  • SeAn-CP maximum
Colour: Fatemi–Socie extended parameter on every plane, relative to the full-sphere maximum. Lines: the only planes SeAn-CP evaluates. Thick-walled tube under axial force, torque and internal pressure. Drag to rotate.
How it is computed

Computed with seancp (SeAn-CP) at commit a3ade97: evaluate_fs_ext on a 3° grid, seancp_fs_ext for the paths and full_plane_scan for the reference, MCE shear amplitude, outer surface. Load cases from the Phase 3 study (seed 23109), with its four largest underpredictions; k_FS = 0.4 and S_y = 900 MPa are illustrative constants.

Data (JSON) · Generator script · SeAn-CP@a3ade97

FIG. R02.2Criteria compared

Every criterion on one load-error scale

Criterion
–
Material
–
Inputs
–
Life law
–
Material constants (* optional)
–
Assessment tests
–
Within ±10 %
–
Within ±20 %
–
Mean eL ± SD
–
Open benchmark of the multifat library, regenerated whenever the library changes. Rows: criteria, by family; dots on the right mark those that need strains or use their own life law. Columns: materials with fatigue-life data. Colour: share of independent tests predicted within ±10 % of the load, or mean load error eL. Select a cell for its statistics.
What eL measures
  • For each test, s* is the factor that, applied to the whole load history (means included), makes the calibrated criterion predict exactly the observed life: eL = ln(1/s*).
  • eL = 0 is an exact prediction. eL = +0.10 means the criterion matches the observed life only at about 0.90 × the applied load: it judges the test more severe than it was (conservative). eL = −0.10 is the non-conservative counterpart.
  • Unlike a life error, eL does not grow with the flatness of the S–N curve or with the power of the load in the damage parameter (squared for SWT and the energy criteria), so every criterion is measured on the same scale.
  • Within ±10 %: share of tests with |1/s* − 1| ≤ 0.10. Mean eL is the systematic bias; its standard deviation is the scatter.
  • Only failures not used to calibrate the criterion or its S–N curve are counted; run-outs give bounds only and are excluded. Independence from constants supplied with the data, such as fatigue limits, is not yet traced (multifat issue #14).
  • Complexity is shown beside the score, not folded into it: whether the criterion needs strains (E, ν) besides stresses, whether it uses its own life law instead of a calibrated master curve DP = A·Nᵇ, and which material constants it requires (* optional).
How it is computed

Output of multifat-pipeline --all (multiaxial-fatigue, commit 67f06ea), which calibrates and runs every registered criterion on each workbook. eL = ln(1/s*), where s* scales the load history until the calibrated criterion predicts the observed life; eL > 0 is conservative. Only failures not used for calibration are counted; run-outs and fatigue-limit-only workbooks are excluded. Test data come from the published sources listed in the repository's PROVENANCE.md.

Data (JSON) · Generator script · multifat@67f06ea

Welded structures and residual stresses

Welding leaves residual stresses before a part enters service. I study how they combine with applied loads and affect fatigue life, through simulations and measurements.

Methods Thermal–structural finite element analysis; residual stress evaluation; experimental fatigue testing.

Further studies and technical context
FIG. R03 / Welded-joint geometry schematic

FIG. R03.1ECP for sharp notches

One master curve, independent welded geometries

Control radius rc
–
Scatter index T
–
Validation in ±2σ band
–
Critical plane parameter averaged at the control radius rc versus cycles to failure. Hollow: calibration series from the literature. Filled: independent validation geometries. Band: ±2σ of the calibration fit.
How it is computed

Generated by running the sharCP calibration and prediction scripts (commit e87d01d) with the illustrative README settings, global plane strain in Mode I and a common ±2σ band. Not the calibration published in the associated article. Experimental lives of the calibration series come from the cited literature.

Data (JSON) · Generator script · sharCP@e87d01d

Additive manufacturing

Mechanical behaviour of additively manufactured structures and thermal effects of the process, from representative structure elements to analytical models of welding and metal additive manufacturing.

Methods Analytical heat-flow models; finite element analysis; experimental characterisation.

Further studies and technical context
FIG. R04 / Layered structure schematic

Materials and component testing

Static, cyclic and temperature-dependent response of materials, joints and components: polymer composites, hybrid riveted joints, automotive parts and circuit board assemblies.

Methods Static and fatigue tests; strain, vibration and temperature measurements; computational modelling.

Further studies and technical context
FIG. R05 / Axial specimen loading schematic

Research projects

05 /FABER CA23109
Previous projects

Hybrid joints for lightweight vehicle structures

2025 · PRIN PNRR 2022

Project team member

Experimental characterisation, theoretical modelling and CAE simulation of joints between dissimilar metals.

Extrusion-based additive manufacturing

2022–2024 · PRA 2022 · University of Pisa

Project co-lead

Research on process optimisation for extrusion-based additive manufacturing of metal and ceramic components.

Structural integrity in automotive applications

2022–2025 · PON 2014–2020

Research contribution with Pierburg Pump Technology Italy / Rheinmetall

Research on polymer composites, elastomeric components and PCB assemblies, combining experimental assessment with computational modelling.

International collaborations

  • Czech Technical University in Prague (Czech Republic)

    Research visit within FABER WG6 · 2026-06

  • Aarhus University (Denmark)

    Circle U. mobility · 2026-01

  • NTNU, Norwegian University of Science and Technology (Norway)

    Erasmus+ research mobility · 2024-07

  • Fraunhofer LBF, Institute for Structural Durability and System Reliability (Darmstadt, Germany)
    • Master’s thesis activity · 2017-12–2018-05
    • Post-graduation employment contract · 2018-07–2018-09
    • Doctoral research period · 2021-02–2021-06

Research collaborations

For academic or industrial collaborations, contact me.