Alessandro Mariotti, Amedeo Cesare Giovanni Bertini, Gianmarco Lunghi, Giovanni Lombardi, Marco Maganzi, Maurizio Boffadossi, Simone Fenili
2024
Cubit, Ferrari, Politecnico di Milano, Università di Pisa
Abstract
A flow-control method is applied to enhance the efficiency and flow homogeneity of three-dimensional diffusers used in open-jet wind tunnels. Suitably shaped grooves are introduced in the diffuser diverging walls. The grooves promote the relaxation of the non-slip condition along the streamlines bounding the small recirculation regions forming passively inside the grooves. That reduces momentum losses and results in a downstream boundary layer with higher momentum, which is more separation-resistant.
In this study, circular and square-section diffusers with different degrees of flow separation were examined. The grooves significantly enhanced performance in circular diffusers by reducing the extent of separation and promoting an axisymmetric and spatially uniform flow. However, negligible benefits were observed for square-section diffusers. In these cases, since flow separation originates from one of the four inclined edges of the diffuser, placing grooves along the diverging walls does not effectively reduce the separation extent. Nonetheless, the grooves become effective again in diffusers with rectangular cross sections of high aspect ratio.
Conference/Journal: Physics of Fluids 36(10), 105164 (October 2024)
CUBIT S.c.ar.l., NORBLAST S.r.l., University of Pisa
Abstract
The use of CFD and optimization cycles is becoming fundamental for saving time and reducing design costs in industrial applications where fluid dynamics plays a critical role. This work presents a CFD-based optimization methodology for sandblasting gun geometry, aimed at maximizing output velocity and media supply rate.
The optimization strategy employed a two-phase approach. First, a preliminary validation of the numerical method was performed against experimental tests on the reference configuration, achieving satisfying agreement. The optimization cycle was then executed considering only the air flow, with the best configurations subsequently verified through Discrete Element Method (DEM) simulations including material injection.
The gun geometry was parameterized using 21 design variables. The optimization workflow integrated modeFRONTIER as the optimization platform, Matlab for workflow management, CATIA V5 for parametric CAD modeling, and StarCCM+ for CFD simulations. The dual objectives were to maximize both the output velocity and the mass flow rate of the media supply tube.
The optimization process explored 4,364 geometric configurations, of which 1,421 were valid and 1,274 showed improvement over the reference design. The three best configurations—optimized for highest output velocity, highest sand supply, and best compromise between both parameters—were validated through complete DEM simulations, confirming the effectiveness of the optimization approach.
Authors: E. Grasso, G. Lombardi, M. Maganzi
Conference/Journal: International CAE Conference 2021
Catherine Ramirez, Giovanni Lombardi, Michele Farnesi, Romain Lanos
2021
Cubit, Syroco, Università di Pisa
Abstract
The Syroco Lab research and simulation team, under the leadership of Ph.D Catherine Ramirez Villalba and Ph.D Romain Lanos, is leveraging modeFRONTIER for automating simulation processes and optimizing the design of several critical components of the Syroco speedcraft. One of the most challenging aspects is to optimize the profile of the foil that anchors the speedcraft to water, and needs to work in both subcavitating and supercavitating regimes.
The major challenge in this high-performance foil design is the requirement for a hydrodynamics optimization procedure to obtain the desired performance and stability, while taking into account the feasibility of the components and controlling the computational costs. While hydrodynamic characteristics were evaluated using CFD software, the feasibility of the wing construction was evaluated with a Finite Elements Method analysis imposing custom forces from CFD results.
This procedure led to a well-defined and fast design space exploration obtaining a wing drag reduction of more than 25% with respect to a reference solution with an imposed target lift. The simulation and optimization process evaluated 4,063 designs, singling out 594 designs complying with constraints.
Conference/Journal: International CAE Conference and Exhibition, November 2021
High-level sport serves as a powerful driver for research, particularly in the fields of aerodynamics and hydrodynamics. This work presents a multi-level optimization methodology for sailboat keels, aimed at reducing hydrodynamic drag during racing conditions.
The optimization strategy consists of two complementary phases. In the first phase, a two-dimensional optimization defines the optimal fin section profile using Bezier curves parameterized with 8 design variables. This process employs modeFRONTIER as the optimization platform, Matlab for workflow management, and X-Foil for aerodynamic analysis. In the second phase, a three-dimensional optimization determines the complete fin-bulb system geometry through 21 design parameters with a constraint on bulb volume. Three-dimensional CFD simulations were performed using StarCCM+ on an HPC cluster with 1024 cores, while parametric CAD modeling was carried out in CATIA V5.
The objective function balances performance across different sailing points: F_obj = 0.4·Cd_downwind + 0.6·Cd_upwind, assigning greater weight to upwind conditions, which are typically more critical for overall racing performance.
Authors: G. Lombardi, M. Maganzi
Conference/Journal: International CAE Conference 2020
Keywords: CFD, optimization, sailboat, keel, hydrodynamics, modeFRONTIER
Giovanni Lombardi, Marco Maganzi, Pasqualetto Elena
2020
Cubit, Università di Pisa
Abstract
ADAMo (Adaptive Digital Aerodynamics for Motorcycles) is a research project guided by Piaggio S.p.A and partially funded by the Toscana region. The objective was to increase the performance, safety and comfort of a motorcycle by using active aerodynamics devices.
The paper describes the aerodynamics development, which was essentially based on intensive CFD use (STAR CCM+ code), related to the choice of the active devices and their optimization. Grids with about 100 million cells were used for preliminary evaluations and final verifications; for analysis of aerodynamic devices and optimization, grids with about 40 million cells were used.
There is an increase in lateral force and in both roll and yawing moments, resulting in a gain in lateral stability. There is also a significant decrease in turbulence. The mean turbulence value decreased by about 30% compared to the reference configuration, while the maximum value decreased more than 40% (in the zone of the driver legs). CFD results were verified in a wind tunnel, with differences less than 5% on the global values.
Conference/Journal: Proceedings of 20. Internationales Stuttgarter Symposium, Springer Vieweg
In the last years, thanks to the development of computational capabilities, the CAE methodology has become an essential tool in the design especially in the preliminary phase and configuration analysis. This happens also and above all in the fluid dynamic design by using CFD codes, which have reached high levels of reliability. In recent years the simulation times have significantly reduced and this means that the preprocessing phase has become the real bottleneck within the project. The engineering consulting companies can face trouble to define standard methodologies for the analysis, because of the wide range of application fields, and need to reduce operational costs and response times to be competitive. This kind of problems induce us to find a strategy to automate the entire design process as much as possible. This work aims at showing the use of ANSA tools for our cases, and the development of preprocessing automation methods easily adaptable to most of them.
Authors: Dr. Marco Maganzi (University of Pisa, CubitLab), Giovanni Lombardi (University of Pisa), Antonio Ercoli (University of Pisa)
Antonio Ercoli, Giacomo De Angeli, Giovanni Lombardi, Marco Maganzi
2019
Cubit, Maserati, Università di Pisa
Abstract
In automotive design, the study of the water thin layer over a car due to rain is becoming increasingly important: the challenge is to obtain a way to describe the behavior of the water over a vehicle in rainy conditions and its interactions with wipers and drainage systems, to determine potential failures of the vehicle design.
In this paper two similar numeric procedures have been realized with the software STAR CCM+ to analyze the dynamic of water thin layer starting from the impingement of the rain on the car surface and taking into account even the motion of the wipers over the windshield. Moreover, the water that flows through the drainage systems is monitored to figure out if the water could produce a malfunction of components near them.
In order to describe each status of the water, many multiphase models are used. These methodologies have been applied on a commercial vehicle model and the results have been examined and compared to each other. The analysis shows a better description of the reality for one of them, leading to the possibility of using it as a design tool in the automotive industry.
Conference/Journal: International Journal of Automotive Technology, Vol. 20, pp. 1123-1129 (2019)
Diffuser rotating stall (in both cases of a vaneless or a vaned configuration) is still one of the open questions which has never been fully understood because of the complexity of the phenomenon and the experimental difficulties to get reliable measurements in such a complex environment.
Under this perspective, Computational Fluid Dynamics (CFD) is an interesting tool to "see" the flow and provide a basic understanding of the associated physics. Several published works have shown that a simplified model of the diffuser without the upstream impeller and the downstream return channel, with realistic boundary conditions entering the diffuser, can provide a qualitative analysis of the stall onset.
This research applies CFD methods to understand the onset mechanisms of rotating stall in centrifugal compressor diffusers, comparing vaneless and vaned configurations to identify design factors that influence stability and operating range.
The flow around a wing is usually well represented by a RANS simulation in cruise conditions, because the zones with flow separations are absent or very small. But, as the angle of attack increases, because of the request of higher lift values, the separated flow zones become important, and the flow is more complex. In these conditions, a RANS approach, also with an unsteady solution, could be not satisfactory. Therefore, we decided to investigate this aspect, by comparing the results of an unsteady RANS solution with a DES solution. In the paper, the setting of both cases are described, as well as the comparison in terms of global coefficients, local distributions and field vorticity distributions. The results show that, at low angle of attack, not significant differences between the RANS and DES solutions occur. On the contrary, at higher angles of attack, the flow highlights a different behaviour, with the DES approach that seems to give a more realistic representation of the flow itself. It is important to note that the difference between RANS and DES results is so considerable that also the lift, drag and pitching moment are significantly changed. Finally, it is possible conclude that the use of a DES approach appears an improvement to evaluate the aerodynamics characteristics of a wing close to the stall conditions, a relevant problem for the aerodynamics design of an airplane.
Conference/Journal: Star Global Conference, Prague (CZ)
Diffuser rotating stall (in both cases of a vaneless or a vaned configuration) is still one of the open questions which has never been fully understood because of the complexity of the phenomenon and the experimental difficulties to get reliable measurements in such a complex environment.
Under this perspective, Computational Fluid Dynamics (CFD) is an interesting tool to "see" the flow and provide a basic understanding of the associated physics. Several published works have shown that a simplified model of the diffuser without the upstream impeller and the downstream return channel, with realistic boundary conditions entering the diffuser, can provide a qualitative analysis of the stall onset.
This research applies CFD methods to understand the onset mechanisms of rotating stall in centrifugal compressor diffusers, comparing vaneless and vaned configurations to identify design factors that influence stability and operating range.
