

JANNA CARLO
Recapiti
Orario di ricevimento
(aggiornato il 10/06/2018 11:46)
Proposte di tesi
1 High Performance Constraint Preconditioning for the Stokes Equations
The numerical solution to the Stokes flow equations is a central topic in several fields of application. The linearized linear system that arises from the PDE discretization is typically saddlepoint system, a tough problem from the numerical linear algebra perspective. Constraint precondtioners are known to be a powerful tool for this kind of problems, however their effective and scalable implementation on high performance systems is still an open problem. 2 Development of scalable algorithms for subsurface simulations on HPC systems Simulating subsurface processes is required in several fields from subsurface hydrology, to geotechnical engineering, from basin simulation to reservoir engineering. The main issue related to these kinds of simulation is usually the large size and heterogeneity of the physical domain that requires discretization involving a hugh number of equations. To this aim, supercomputers are becoming available to scientist and engineers with an increasing need of designing parallel algorithms to exploit these computational resources. 3 Algebraic Multigrid preconditioning for illconditioned structural problems The numerical solution to the Cauchy indefinite equilibrium equations is a typical task in civil and mechanical engineering, as well as in many other fields, with a large number of open and commercial software available to the end user. The arising linear systems may represent a serious bottleneck in the overall simulation process that may take more than the 90% of the total time. Algebraic Multigrid (AMG) techniques represent a powerful tool for the solution of linear systems, however they have been historically designed for scalar PDEs, such as e.g. the Poisson equation. For nonscalar PDEs arising in computational mechanics, AMG performance still needs to be improved. 4 Modelling channel network dynamics in tidal landscapes (Joint thesis with Prof. Andrea D'Alpaos, Dipartimento di Geoscienze) Tidal landscapes display sticking patterns emerging from the interaction of biogeomorphic processes over a wide range of spatial and temporal scales. Among these patterns, channel networks are key features of the tidal landscape, because they exert a strong control on hydrodynamics, sediment and nutrient dynamics within tidal systems. The morphodynamic model of network evolution is based on a simplified hydrodynamic model where, by assuming a balance between water surface gradients and friction, the 2D shallow water equations is simplified to a Poisson boundary value problem, whose numerical solution is obtained through a Finite Difference discretization. The accurate modeling of this physical process requires high resolution grids with the overall computational cost becoming an unavoidable limiting factor. The use of High Performance Computers may help overcome these difficulties allowing for large scale simulations in acceptable time. Curriculum Vitae
Carlo Janna si è laureato in Ingegneria Civile il 2 ottobre 2003 presso l'Università degli Studi di Padova con votazione 109/110 e nella stessa Università ha conseguito il titolo di Dottore di Ricerca con la tesi “Modellazione numerica del comportamento meccanico delle faglie regionali per il confinamento geologico della CO2 antropica" con il Prof. Giuseppe Gambolati come relatore. Dal dicembre 2011 è Ricercatore di Analisi Numerica presso il Dip. ICEA. I principali interessi scientifici di Carlo Janna si collocano nell'ambito dello studio numericomatematico della meccanica dei mezzi porosi, con applicazioni specifiche relative al settore dell'idraulica sotterranea e dell'industria petrolifera, e nell’ambito dell’algebra lineare numerica. L'attività prevalente svolta fino ad oggi consiste nello sviluppo ed implementazione di modelli numerici FEM per la simulazione dei principali processi geomeccanici e fluidodinamici del sottosuolo. Per quanto concerne l’algebra lineare, ha studiato e approfondito le tecniche numeriche per la soluzione di sistemi lineari e la ricerca di autovalori che scaturiscono in problemi strutturali e fluidodinamici di grandi dimensioni, in particolare metodi iterativi precondizionati. Nell'ambito del calcolo sequenziale ha studiato e implementato precondizionatori ad hoc per la soluzione ottimale di determinati problemi legati alla modellazione del sottosuolo. Dal 2010 al 2012, Carlo Janna ha partecipato ai programmi di ricerca PARPSEA (PARallel Preconditioners for large Size Engineering Applications), SCALPREC (SCALable PREConditioners), OPTIDAS (OPTImization and Data ASSimilation) e SPREAD (Scalable PREconditioners for Advanced Discretizations) in ambito HPC nel corso dei quali ha sviluppato e implementato su architetture massicciamente parallele precondizionatori di nuova concezione.
Curriculum del docente in PDF: D0D3F7F9A0BF1C81F2807B2AD6515E4F.pdf Aree di ricerca
1 Analisi Numerica
2 Algebra Lineare Numerica 3 Calcolo parallelo 4 Geomeccanica 5 Modellistica Ambientale Pubblicazioni
1. V. A. Paludetto Magri, A. Franceschini, M. Ferronato, and C. Janna (2018), Multilevel Approaches for FSAI Preconditioning, Numerical Linear Algebra with Applications, available online (SJRScopus 1.104).
2. S. Ye, A. Franceschini, Y. Zhang, C. Janna, X. Gong, J. Yu and P. Teatini (2018), Earth fissure development caused by extensive aquifer exploitation. A novel modelling approach applied to the Wuxi case study, China, Water Resources Research, 54, pp. 2249–2269 (SJRScopus 2.296). 3. H. T. Honorio, C. R. Maliska, M. Ferronato, and C. Janna (2018), A stabilized elementbased finite volume method for poroelastic problems, Journal of Computational Physics, 364, pp. 49–72 (SJRScopus 2.047). 4. Franceschini, V. A. Paludetto Magri, M. Ferronato, and C. Janna (2018), A Robust Multilevel Approximate Inverse Preconditioner for Symmetric Positive Definite Matrices, SIAM Journal on Matrix Analysis and Applications, 39, pp. 123–147 (SJRScopus 1.739). 5. N. Spiezia, M. Ferronato, C. Janna and P. Teatini (2017), A twoinvariant pseudoelastic model for reservoir compaction, International Journal for Numerical and Analytical Methods in Geomechanics, 41, pp. 1870–1893 (SJRScopus 1.452). 6. Zanette, M. Ferronato, and C. Janna (2017), Enriching the finite element method with meshfree techniques in structural mechanics, International Journal for Numerical Methods in Engineering, 110, pp. 675–700 (SJRScopus 1.623). 7. R. Baggio, A. Franceschini, N. Spiezia, and C. Janna (2017), Rigid body modes deflation of the preconditioned conjugate gradient in the solution of discretized structural problems, Computers & Structures, 18, pp. 15–26 (SJRScopus 1.630). 8. Franceschini, M. Ferronato, C. Janna, and P. Teatini (2016), A novel Lagrangian approach for a stable numerical simulation of fault and fracture mechanics, Journal of Computational Physics, 314, pp. 503–521 (SJRScopus 2.047). 9. M. Bernaschi, M. Bisson, C. Fantozzi, and C. Janna (2016), A FSAI preconditioned conjugate gradient solver on GPUs, SIAM Journal on Scientific Computing, 38, pp. C53–C72 (SJRScopus 1.973). 10. C. Janna, M. Ferronato and G. Gambolati (2015), The use of supernodes in factored sparse approximate inverse preconditioning, SIAM Journal on Scientific Computing, 37, pp. C72–C94 (SJRScopus 1.973). Pubblicazioni del docente in PDF: D0D3F7F9A0BF1C81F2807B2AD6515E4F.pdf Insegnamenti dell'AA 2018/19


Università degli Studi di Padova, via 8 febbraio 2, 35122 Padova / P.IVA 00742430283 ‐ Informazioni sull'uso dei cookie 