newsletter - enginsoft · shape with optimization ... design exploration and optimization ... [icem...

00101110101001110101100101001101110010100101010100101110101001110101100101

NewsletterSimulation Based Engineering & Sciences

Year n°4 Winter 201512

Diesel fuel efficiency takesshape with optimization

A new heart valve replacement procedure modeled with multiphysics simulation

A 3D FEM approach to evaluate the flux density and the eddy current induced on the turbogenerators

Casting simulation in heavy section ductile iron production

Analysis of a vane oil pumpmechanism failure: multibody, fluid-dynamic and validation

Interview with Technical Leader of GE Oil & Gas

Design exploration and optimizationwith modeFRONTIER at Volvo Car Corporation

3 - Newsletter EnginSoft Year 12 n°4 Flash

LASHEngineer – derived from the Latin verb ingeniare, meaning to design or create: from behind the scenes, we know that in order to fulfill the role of an engineer, a great body of knowledge needs to be developed through extensive training and industry practice to go beyond the primary function of conceptual design but to innovate great design solutions.The magnitude of the design challenges that our customers overcome on a daily bases are illustrated in this edition of the Newsletter. The diverse solutions being innovated is evidence that engineers are advancing in the means of problem solving by adopting new numerical tools available to them. This was evident in the 31st edition of the International CAE Conference, which filled me with an overwhelming sense of pride. This progression is not only an advancement in technology, but underpins a step change in improving our future and quality of life. In the article on page 29, we showcase the development of a radically new Heart Damper medical device for the ventricular chamber and its ability to prevent heart failure and lengthen life expectancy.Furthermore, the article on page 46 shows how problems are being solved within new industries taking their new products into literally new climates. Here the dangerous problem of ice throw from wind turbines has been solved by the application of FENSAP-ICE, while maximizing the energy production.At EnginSoft we are committed to developing new methods and approaches, in partnership with our customers, to meet the challenges of tomorrow. Therefore, I am excited to announce that in 2016 the MapleSoft suite of products will be added to our enterprise-level solutions portfolio in Europe, Find out more about the modelling and simulation capabilities of MapleSoft products on page 50.In addition, thank you all for your contribution and another remarkable year. EnginSoft wishes you, your colleagues and families a Merry Christmas and prosperous year ahead.

Stefano Odorizzi, Editor in chief

F

The Aerospace & Defense sector represents a mature industry with many components possessing a long development history and delivers good levels of reliability and performance. Yet the development of innovative technologies and the demand for continuous improvements cannot be delivered through isolated FEM, CFD, EM or MANUFACTURING calculations. The real challenge of Virtual Prototyping is to integrate all of these technologies in a unique design and optimization process. This is where the CAE market is moving, and where partnering with EnginSoft can make a real difference to your company. At EnginSoft, experience in multi-disciplinary simulation consultancy for the Aerospace & Defense industries, continuous investments in R&D programs and high performance computing resources (HPC) provide the reliability and expertise to support the customers’ as a key partner in product development.The knowledge and commitment of EnginSoft experts has helped enable many companies to efficiently and effectively identify suitable solutions for complex requirements. Leading-edge technologies and advanced methods are applied to specific industrial contexts in order to achieve shortened development times and improved resource utilization.Advanced computational power and skills in software management within the arena of HPC can significantly reduce both cost and time.Our novel methodologies for developing security and safety systems, multifunctional composite structures, space greenhouse environment, HVAC design and integrated manufacturing solutions are only some of the applications impletemented by leading Aerospace and Defense companies to develop strategic and competitive advantage.

010010111010100111010110010100110111001010010101010010111010100111010110010100

NewsletterSimulation Based Engineering & Sciences

The Simulation-Based approach in newHigh Technology Product Development

Optimization techniques applied to the design of Gas Turbine Blades Cooling Systems

Numerical Models forAimable Warhead

Engineering Simulation forUnmanned Vehicle Designand Development

Topological Optimizationmethods for aerospace applications

in the Aerospace&Defense Industry

Electromagnetic Issue fora Fiber/Composite Laminate

DOWNLOAD your free copy on:http://www.enginsoft.com/aerospace-newsletter

Simulation Based Engineering & Sciences

EnginSoft Newsletter

Special Issue onAerospace&Defense

Newsletter EnginSoft Year 12 n°3 - 4 ContentsContents 5 - Newsletter EnginSoft Year 12 n°4

ContentsINTERVIEW6 Interview with Eng. Marco Ruggiero, External Funding & Research Collaborations Technical Leader at GE Oil & Gas CASE HISTORIES8 Design exploration and optimization with modeFRONTIER at Volvo Car Corporation9 Influence of body stiffness on vehicle dynamics using modeFRONTIER10 An automated approach to speed up concept creation and validation based on flow optimization12 Calibration of piston FEM-models to aid the design of diesel engines13 Steering System Verification using Hardware-in-the-Loop with modeFRONTIER15 Diesel fuel efficiency takes shape with optimization17 Particleworks, an innovative particle method simulation tool, playing an important role in previously-unattainable simulation problems19 Analysis of a vane oil pump mechanism failure: Multibody, fluid-dynamic and validation22 A 3D FEM approach to evaluate the flux density and the eddy current induced on the turbogenerators frame walls and clamping bolts in order to analyze local overheating phenomena26 A new heart valve replacement procedure modeled with multiphysics simulation could eliminate the need for open-heart surgery

29 Structural analysis of an implantable cardiac device - Heart Damper - for the treatment of advanced heart failure inside a 3D finite element model of the ventricle33 Comparison of casting simulation results and experimental data in heavy section ductile iron production38 Simulation Helps European Nuclear Safeguards Activities40 Thermo-fluid dynamics model of two-phase system alloy-air inside the shot sleeve in HPDC process46 Wind energy in cold climates

SOFTWARE UPDATE50 EnginSoft to distribute and support MapleSoft products in 2016

RESEARCH52 Innovative methodologies for Robust Design Optimization with large number of uncertainties using modeFRONTIER

EVENTS55 Mentor Graphics Announces Winners of Don Miller Award for Thermo-Fluid Design Excellence during the International CAE Conference 201556 INTERNATIONAL CAE CONFERENCE Your opportunity to be part of the future 58 INTERNATIONAL CAE POSTER AWARD 2015: a great success for the fourth year

OUR ACKNOWLEDGEMENT AND THANKS TO ALL THE COMPANIES, UNIVERSITIES AND RESEARCH CENTRES THAT HAVE CONTRIBUTED TO THIS ISSUE OF OUR NEWSLETTER

Newsletter EnginSoftYear 12 n°4 - Winter 2015To receive a free copy of the next EnginSoft Newsletters, please contact our Marketing office at: [email protected]

All pictures are protected by copyright. Any reproduction of these pictures in any media and by any means is forbidden unless written authorization by EnginSoft has been obtained beforehand. ©Copyright EnginSoft Newsletter.

EnginSoft S.p.A.24126 BERGAMO c/o Parco Scientifico TecnologicoKilometro Rosso - Edificio A1, Via Stezzano 87Tel. +39 035 368711 • Fax +39 0461 97921550127 FIRENZE Via Panciatichi, 40Tel. +39 055 4376113 • Fax +39 0461 97921635129 PADOVA Via Giambellino, 7Tel. +39 049 7705311 • Fax +39 0461 97921772023 MESAGNE (BRINDISI) Via A. Murri, 2 - Z.I.Tel. +39 0831 730194 • Fax +39 0461 97922438123 TRENTO fraz. Mattarello - Via della Stazione, 27Tel. +39 0461 915391 • Fax +39 0461 97920110133 TORINO Corso Marconi, 10Tel. +39 011 6525211 • Fax +39 0461 979218

www.enginsoft.it - www.enginsoft.come-mail: [email protected]

The EnginSoft Newsletter is a quarterly magazine published by EnginSoft SpA

COMPANY INTERESTSEnginSoft GmbH - GermanyEnginSoft UK - United KingdomEnginSoft France - FranceEnginSoft Nordic - SwedenEnginSoft Turkey - TurkeyVSA-TTC3 - Germanywww.enginsoft.com

CONSORZIO TCN www.consorziotcn.it • www.improve.itCascade Technologies www.cascadetechnologies.comReactive Search www.reactive-search.comSimNumerica www.simnumerica.itM3E Mathematical Methods and Models for Engineering www.m3eweb.it

ASSOCIATION INTERESTSNAFEMS International www.nafems.it • www.nafems.orgTechNet Alliance www.technet-alliance.com

ADVERTISEMENTFor advertising opportunities, please contact our Marketing office at: [email protected]

RESPONSIBLE DIRECTORStefano Odorizzi - [email protected]

ART DIRECTORLuisa Cunico - [email protected]

PRINTING Grafiche Dal Piaz - Trento

Auto

rizza

zione

del

Trib

unal

e di

Tre

nto

n° 1

353

RS d

i dat

a 2/

4/20

08

The EnginSoft Newsletter editions contain references to the following products which are trademarks or registered trademarks of their respective owners: ANSYS, ANSYS Workbench, AUTODYN, CFX, FLUENT, FORTE’, SpaceClaim and any and all ANSYS, Inc. brand, product, service and feature names, logos and slogans are registered trademarks or trademarks of ANSYS, Inc. or its subsidiaries in the United States or other countries. [ICEM CFD is a trademark used by ANSYS, Inc. under license]. (www.ANSYS.com) - modeFRONTIER is a trademark of ESTECO Spa (www.esteco.com) - Flowmaster is a registered trademark of Mentor Graphics in the USA (www.flowmaster.com) - MAGMASOFT is a trademark of MAGMA GmbH (www.magmasoft.de) - FORGE, COLDFORM and FORGE Nxt are trademarks of Transvalor S.A. (www.transvalor.com) - LS-DYNA is a trademark of LSTC (www.lstc.com) - Cetol 6 is a trademark of Sigmetrix L.L.C. (www.sigmetrix.com) - RecurDyn™ and MBD for ANSYS is a registered trademark of FunctionBay, Inc. (www.functionbay.org) - Maplesoft are trademarks of MaplesoftTM, a a subsidiary of Cybernet Systems Co. Ltd. in Japan (www.maplesoft.com)

Interview Newsletter EnginSoft Year 12 n°4 - 6

Present in Italy since 1922, General Electric could be considered as a real Italian company with over 7,000 employees in 15 offices. Its main activities could be categorized by four macro-divisions which are well grounded in the Italian economic and industrial background, taking advantage of its workforce, knowledge and excellence, its investments have a positive repercussion on the whole system, especially where the Italian SMEs economic foundation is concerned. While investments in research are limited in Italy, GE devotes 70 million Euros every year to Research & Development in Italy. In the perspective of a growing business integration in General Electric in Italy, Sandro De Poli, has been nominated as chief for all the Italian activities, with the objective of increasing competitiveness in Italy, developing new products, offering innovative and productive solutions able to better support its customers. As a result, General Electric, performed like a great virtuous Italian company, with a turnover of over 3 billion Euros it was able to progress the business to almost 5 billion Euros.

The example of OIL&GAS – NUOVO PIGNONE This multinational has decided to locate in Florence the world headquarter for the Oil&Gas division, producing the technology for the extraction, refining and transportation of oil and gas, used in installed systems all over the world. The division was founded in 1994, when GE took over Nuovo Pignone from Eni. Since its entrance in the GE group, its turnover has increased ten times, thus reaching 10 million Euros in 2011. Within two years the number of employees has increased by

15%, reaching 4500 units at the end of 2011, thus banish the fears for possible employment issues or office disruption out of Italy, due to the taking over of the Florence company by a multinational enterprise. We have interviewed Eng. Marco Ruggiero, External Funding & Research Collaborations Technical Leader at GE Oil&Gas in relation to the ATENE (Advanced Technologies for ENergy Efficiency) Project. The ATENE project aims to improve performance for rotating and reciprocating machinery, employed in the Oil & Gas and Industrial Power Generation industries. Performance improvements, which also includes power density increase, reduction of global production and testing process environmental impact, are achieved through a branched work-plan. ATENE is structured on a network of SMEs and research laboratories distributed within Tuscany, aimed at leveraging specific local excellences. The consortium is constituted of 16 partners with 5 research centres, 1 big enterprise and 10 SMEs. The structure of the proposal is organized per technological areas and the network model generates 5 “virtual laboratories”, each one devoted to a specific Operative Objective.

General Electric: the growing multinational company thanks to simulation

Interview7 - Newsletter EnginSoft Year 12 n°4

My role is focused on External Knowledge and Funding. That means tapping external knowledge reservoirs, targeting Universities and Research Centers, and structuring ideas and partnerships to compete in the arena of public funding to develop new technologies.

How is the ATENE Project connected to the other initiatives concerning product development in General Electric? The idea behind ATENE is that of exploring 5 different research areas, selecting topics with a relatively low Technology Readiness Level and progressing them beyond the state of the art technology. All the topics revolve around the idea of enabling technologies, as opposed to new product developments. This distinction is very important and it complements nicely the typical Large Enterprise approach to research that has to be focused on bringing new products on the market.

One of the ATENE Project requirements was the setting up of a group made up of Universities and SMEs: what criteria’s have influenced the selection of the team members? Demonstrated knowledge in the research area has been the number one selection criterion. This was very closely followed by the inclination to build network connections; our idea of a good research network is that of a self-sustaining one, meaning partner collaborations and connections shouldn’t be Large Enterprise centric but driven by a common interest in the development of technologies or products. The network should be able to survive (if not thrive) even without the continuous support of the Large Enterprise.

How were the project results exploited inside the company? The development of new enabling technologies is always seen as something good and exciting. Even more so when it comes with ready to use engineering tools that incorporate knowledge and (let’s not forget) enhance productivity, freeing up budget space for more research and

innovation. This has been abundant in ATENE and has let us move beyond what we could have achieved alone.

What has been the role of the CAE tools on the project? CAE tools have played a fundamental role in the project; a large chunk of the activity has been hovering the boundary between industrial and fundamental research, which is a space better explored and understood by modeling phenomena and carrying out basic experiments. The Advanced CAE tools allowed you to perform trade offs and design of experiment procedures on a comparative scale, saving a lot of hands on time and real experimentation.

Were the CAE tools frequently used further down the project line, thus affecting the protocols of the project development? I would love to take on this one, building on the idea of model based validation. In today’s hectic market,

innovation has to move at a faster pace; we need to be experimenting sooner and more. This is where CAE tools will be instrumental. It is highly unpractical to build physical prototypes of every idea we come up with; it is mandatory to develop an excellent understanding of physics and related modeling in order to bring more of the early validation into the digital space. We also need to get over the old single tool approach and embrace the multi-objective optimization tools that are available (like your modeFRONTIER) to look at the problem from an holistic point of view and connect the various disciplines, to come up with new and integrated solutions whilst providing lots of productivity.

How has the contribution of EnginSoft influenced the quality, potentialities and capabilities of the project? EnginSoft pulled its weight and went above and beyond that, demonstrating an engineering approach to problem solving in the face of project hiccups (as with all projects, we experience bad days!). I think that the deep understanding of the physics and problems coupled with the domain knowledge in translating those into high performance models and codes has been a tremendous contribution to ATENE.

What are your wishes for the future of scientific technology, which always looks for the ideal dimension between creativity and competitiveness? An ATENE 2 of course! Joking aside, there is truth in that remark: availability of research funding tools that provide freedom to explore the design space, delivered with speed by the funding entities; government support for research organizations to build the next generation of labs…and in Tuscany!

Fig. 1 - Eng. Marco Ruggiero, External Funding & Research Collaborations Technical Leader at GE Oil&Gas

Interview with Eng. Marco Ruggiero, External Funding & Research Collaborations Technical Leader at GE Oil & Gas

Case Histories Newsletter EnginSoft Year 12 n°4 - 8

The use of design exploration through automatic CAE processes with modeFRONTIER is becoming more and more a part of an engineer’s everyday work at Volvo Car Corporation. For every discipline, from the static load calculations of the chassis components to the AC temperature calculation in the cabin there exists a CAE model which predicts the performance of future cars.

Today’s engineering tasks range from the manual adjustment of model input parameters to the more complete exploration of the, so called, design space. By automating all the steps of geometry and mesh generation, setup and post-processing in a single disciplinary analysis a clear view can be made of the whole design process. This allows the engineer to spend time learning more about how design variables interact and correlate with the product’s performance. The main aim is to connect several automated engineering processes within one single multidisciplinary design optimization (MDO).

The MDO Arena is a group of key engineers within Volvo, collated from different departments, who work on multidisciplinary problems. Some are using the tools within modeFRONTIER to analyze NVH simulations that contain up 30-50 design parameters every day. For this application, the key to success is to first determine the parameters which contribute the most to achieving the objectives. The most important parameters can then be passed to the multidisciplinary optimization platform. modeFRONTIER is used in many time limited projects that require a complex process workflow and advanced mathematical approach. For an

inexperienced engineer not used to this approach the technical support from EnginSoft is always the obvious choice to help set up the project in a time effective manner. After the process automation is complete the engineer will have a thorough understanding of how the integration of CAE tools is done in the modeFRONTIER workflow and will possess the required knowledge about applying the correct optimization strategy. The value of using automated design exploration and optimization to develop any car component has proven to be one of the greatest steps forward in increasing the technical know-how at Volvo Car Corporation. As ever higher demands to increase fuel efficiency and reduce weight is tomorrow’s world, the incorporation of a good CAE optimization strategy is a necessary step compete in today’s climate.

For more information on optimization activities in Volvo Cars:Håkan Strandberg - EnginSoft Nordic

[email protected]

Design exploration and optimization with modeFRONTIER at Volvo Car Corporation

Subjective physical measurements performed at Volvo Cars have shown that changes in body stiffness affect the vehicle dynamics in passenger cars. However, objective calculations based on data from full vehicle simulations do not predict the same behavior and cannot confirm the results of the subjective physical method. Therefore a deeper analysis of how the body stiffness influences the vehicle dynamics is needed.

The resulting study is done by comparing a standard Volvo S60 and a Volvo S60 with body reinforcement. The car’s body is modified with detachable steel bars to increase and decrease the stiffness in specific areas. The setup configuration for dampers, springs, tires, anti-roll bars, drive- train etc. are kept constant in order to isolate the effect from body stiffness.

The multi-body simulation software MSC Adams Car is used to calculate the Kinematics and Compliance. The model consists of 17 steel bars connected to different points in the car, see table below, and each bar has two

states: reinforcement on/off. A 2-level factorial Design of Experiment (DOE) is generated, but to limit the database table to a manageable size 6 groups are created out of the 17 steel bars, reducing the DOE size from 2^17 to 2^6. The groups are then based on locations and orientations in the car: Front, Rear, underbody, longitudinal, transversal, or cross mounted.

The simulation of each design consists of a standard set of Kinematic and Compliance tests: static load, parallel wheel travel and single wheel travel. These three tests are conducted to gather all Kinematic/

Case Histories9 - Newsletter EnginSoft Year 12 n°4

Fig. 1 - Adams Car model with reinforcement bar highlighted in yellow

Influence of body stiffness on vehicle dynamics using modeFRONTIER

Table 1 - Reinforcement groups based on the orientation and location of the bars

Fig. 2 - Workflow of the Design Of Experiment project


In figure 4 the velocity field of the air flow can be seen around the engine cover and bay for one design configuration. Between the engine cover and the car hood the velocity of air can be seen to increase rapidly (red colored) due to the small passage. Probe 1, 2, and 3 are the locations where the velocity should be maximized in the shape optimization of the engine cover.

Each simulation takes about 10 minutes so the designs from DOE + optimization can be obtained in 3-4 days. The result of the modeFRONTIER optimization can be seen in figure 5 where the 3 probe values are plotted. Pareto designs are colored red and located at the very up-right corner. In table 1, a design proposal is presented with the highest improvement in velocity. Not surprisingly the greatest increase in velocity can be seen in probe location 1 (Max_V) since it is positioned close to the engine cover.

Johnny Olsson, Supervisor: Johan Ölvander Linköpings universitet

Raik Orbay Volvo Car Corporation

Fig.3 - modeFRONTIER workflow with the mesh-morphing input parameters and the probe values to maximize in the OpenFOAM simulation of the engine cover

Fig. 4 - Cross section view under the car hood for a typical mid-size vehicle

Fig. 5 - Results from DOE + Optimization of velocity at probe locations 1, 2, and 3

Table 1 - Results from DOE + Optimization of velocity at probe locations 1, 2, and 3


Compliance values about spin angle, spring length, damper length etc. In table 1 the reinforcements in the car can be seen which are to be switched on/off in the statistical analysis in modeFRONTIER. Results on figure 3 shows the increase of lateral and local stiffness relative the Baseline S60 model from the 6 groups of reinforcement. The Front Transversal Reinforcement (orange) can be seen to have biggest influence on the Lateral/ Toe/Camber stiffness front.

Oskar Danielsson, Alejandro González Chalmers University of Technology

Supervisor: Matthijs Klomp Volvo Car Corporation

Fig. 3 - Influence of reinforcement compared with Volvo S60 baseline model

An automated approach to speed up concept creation and validation based on flow optimization

Volvo Car Corporation aims to decrease the vehicle production time to 20 months by 2020. It is therefore necessary to automate the process of evaluating new concept designs. In this MSc. thesis project a CFD analysis of an engine cover is automated in modeFRONTIER using the mesh-morphing methodology in ANSA. The objective of the study is to maximize the cooling air at 3 probe locations close to the engine.

A design of an existing engine cover concept is used as a baseline design. The advantages of using mesh-morphing is that mesh nodes are repositioned by the morphing tool without processing the CAD model which gives more degrees of freedom and the possibility of new concept designs which were not obvious for the Concept Engineer.

In the modeFRONTIER workflow, 9 input parameters are controlling the boxes to morph. The initial DOE is a Uniform Latin Hypercube table of 96 design. MOGA-II is chosen as the optimization algorithm, with 20 generation and 20 designs in each generation. The mesh morphing is done in ANSA, the simulation in OpenFOAM and the extraction of the probe velocity is done within the Shell node.

Fig. 1 - Morph box location on the engine cover

Fig. 2 - Illustration of morphing 6 out of 10 points in x direction


OBJECTIVE: Output variables are 15 temperature locations on the piston which fit experimental data by a single objective minimization of the Root Mean Square value. CONSTRAINTS: 13 heat flux measurements constraining the designs in the calibration process. The simulation time is 2-3 minutes, so a converged result of the piston surface temperature can be obtained in 1-2 days, see figure 3. The results from the validations show the intended engine load case span and the piston thermal analysis give accurate results in terms of the piston structure temperature field and its surface heat fluxes. This thermal analysis process enhances the way complete thermal loads are determined in a car engine as a boundary condition in combustion CFD-simulations. It also serves as an aid in the design and construction process of the piston.

Martin Gonera, Olle Sandin - Chalmers University of TechnologySupervisor: Stefan Eriksson – Volvo Car Corporation

Fig. 4 - Temperature history during calibration process on different surfaces of the piston

Steering System Verification using Hardware-in-the-Loop with modeFRONTIERThe scope of this Master’s thesis is to show how an automated method of calibration, simulation, and analysis in a Hardware-in-the-loop rig can be used to optimize a car’s steering performance. The comparison of Simulation-In-the-Loop (SIL) and Hardware-In-the-Loop (HIL) is performed by simulating a couple of driving scenarios on the different

steering system configurations. In the SIL case, CarMaker by IPG Automotive GmbH is used for simulating driving maneuvers. In the HIL case a rig is used. The main components of the rig are the Electric Power Assisted Steering (EPAS) and three actuators. Two of the actuators are linear drives that apply force on the steering rack directly

and other actuator simulates the driver by applying angle/torque from the steering robot. The complete process of calibration in INCA, simulation in CarMaker, and analysis is automated in modeFRONTIER. In order to design a test sequence a Design of Experiment is constructed. The input parameters of the calibration are: a,b, drvMod and DummySpeed. Parameter ‘a’ and ‘b’ define the active return function in Matlab. A simplified active return curve is used with 5 measurements which is fitted in the calibration process in INCA, where parameter values can be calibrated and visualized in real time. The active return curves in the ECU of the steering gear determines the steering wheel velocity [°/s] depending on the steering wheel angle. The active return functionality applies a torque opposite to the steering wheel torque in order to return the steering wheel angle to 0 degree. This functionality can be seen when the driver releases the steering wheel.

Fig. 1 – Electric Power Assisted Steering with direction of forces and torques of actuators


The current trend in the car industry is towards smaller engines with higher specific power. This requires the engine to carry higher thermal loads and for the complete temperature field to be simulated to meet these requirements.

In this Master´s thesis project, convective loads on the piston surface are calibrated to fit experimental temperature data. The temperature is measured on the piston boundaries of an Abaqus FEM-model where the

Inverse Heat Conduction Method, IHCM, is utilized to calibrate several engine load cases. The input variables are the convective load parameters of the HTC function (one HTC function for each piston surface). Since the each piston has 15 surfaces where the temperature is to be calibrated, the root mean square of all the temperature residuals is minimized using modeFRONTIER. This reduces the calibration process to a single-objective optimization, and allows the SIMPLEX algorithm to be used which is fast in convergence. The calibration is done for different engine load cases. To use the data for other engine load cases interpolation is done to produce a matrix of piston boundary HTCs for a larger number of engine load cases. INPUT: The Input variables are the convective loads parameters of the HTC-functions which are assigned to the piston surfaces.

Calibration of piston FEM-models to aid the design of diesel engines

Fig. 1 - Convective load applied on the piston surface to calculate temperature and heat flux

Fig. 2 - Workflow of the calibration project

Fig. 3 - RMS error history during the calibration process

Case Histories15 - Newsletter EnginSoft Year 12 n°4 Case Histories Newsletter EnginSoft Year 12 n°4 - 14

drvMode and DummySpeed are inputs to the CarMaker simulation and the Electronic Control Unit (ECU) of the rig. The drvMode parameter determines the amount of assisted force that is to be applied by the servo on the steering rack. A low, medium, and high level can be set depending on how much force assistance the driver needs. The DummySpeed parameter is sent to the ECU on the rig in order to ‘trick it’ into thinking the vehicle has a certain speed. This has a big influence on the car’s steering performance since the assisted force is highly dependent upon the velocity. The output variables on the workflow are angle hysteresis, steering friction, steering stiffness, and steering stiffness at zero steer. Figure 4 illustrates the definition of the output variables from a test sequence where the steering wheel torque (torque required by the driver) is measured as a function of steering wheel angle.

• Angle hysteresis [Nm] is the difference of steering wheel angle at zero between upper and lower part of hysteresis curve.

• Steering friction [Nm] is measured at zero steering angle, so this is the friction when driving straight forwards.

• Steering Stiffness [Nm/°] is a measure of how much torque is need to change the steering wheel angle. Gradient is averaged over interval ±1.5°

• Steering Stiffness at Zero steer [Nm/°] is a measure how much torque is needed to change the steering wheel angle at ±0.15°

ResultsA SOBOL DOE with 76 designs was generated for statistical analysis and shows that the DummySpeed correlates strongly with all of the objective metrics.

The Proof of Concept of this thesis makes it natural to improve the calibration of ECU parameters of the steering gear. Future optimization of ECU parameters can be done in modeFRONTIER which has a variety of optimization tools and analysis methods. By doing this the desired performance of the steering gear can be achieved in the HIL before steering gear is installed in a car. This will shorten development time tremendously since the ECU parameter will only need to be adjusted slightly when installed in the car.

Peter Karlsson, Salko Bjelevac – Linköpings UniversitetSupervisor: Matthijs Klomp – Volvo Car Corporation

Fig. 2 – Active Return function from ECU and estimated function Fig. 3 – Process flow and communication links

Fig. 4 – Measurements from the rig: Torque required to steer the wheel definition of output variables

Fig. 5 - modeFRONTIER workflow containing the active return function in Matlab, Calibration in INCA, and execution of CarMaker

Fig. 6 - Correlation Matrix chart showing the dependencies of objective metrics on input parameters of the workflow

In the debate on how best to tackle the impact of vehicles on environment, the improvement of diesel engine efficiency has emerged as a transitory but effective solution, especially for heavy-duty vehicles and passenger cars. Designers at the ISUZU Advanced Engineering Center (IAEC) have analyzed how to enhance fuel efficiency by modifying the shape of the diesel engine combustion chamber.

“The optimized chamber improved fuel consumption by 3.2% compared to its shallow dish-type counterpart.”

ChallengeTheoretical thermal efficiency affects fuel consumption in diesel engines and one way of improving it is to increase the combustion chamber compression ratio. The resulting higher in-cylinder temperature and the expansion of the impingement area between fuel spray and chamber wall, however, can cause the chamber wall to heat up and lower theoretical efficiency.

The team at IAEC looked at a new way of lowering heat loss by studying the combustion chamber shape, preventing the volumetric inefficiencies and cost and durability issues, which other methods caused. Among others, main methods in literature are the “TemperatureSwing Heat Insulation” or varying the intake air conditions and injectionstrategy.

Diesel fuel efficiency takes shape with optimization

Figure 1. Calculated heat release rates and cylinder gas temperatures of re-entrant and shallow dish-type combustion chambers for an engine with a 115 mm bore and 125 mm stroke

ISUZU Advanced Engineering Center (IAEC) enhances fuel efficiency by optimizing the combustion chamber design

Case Histories Newsletter EnginSoft Year 12 n°4 - 16 Case Histories17 - Newsletter EnginSoft Year 12 n°4

SolutionTo analyze the impact of the different chamber shapes, the team first defined the chamber outline and spray angle (A method developed at the University of Wisconsin-Madison was adopted, see for ref: Kong, S., Patel, A., Yin, Q., Klingbeil, A. et al., “Numerical Modeling of Diesel Engine Combustion and Emissions Under HCCI-Like Conditions With High EGR Levels,” SAE Technical Paper 2003-01-1087, 2003, doi:10.4271/2003-01-1087.) and adjusted it to match a given baseline compression ratio. The computational mesh was then created with CONVERGE CFD and modeFRONTIER was used to pilot the 3D-CFD simulations. “In this way, we were able identify the shapes with the maximum cumulative heat release and work, and – at the same time - the minimum heat loss” says Ing. Takashima, Chief Engineer Powertrain Product Planning at IAEC.

“modeFRONTIER helped us spot the optimal shape and further analyze the delicate trade-offs regarding the thermal balance.”

modeFRONTIER ADVANTAGES“The shape with the highest cumulative rate of heat release was analyzed in depth. We compared it to calculated heat release rates and cylinder gas temperature profiles of re-entrant-type and shallow-dish-type chambers and, later, verified it using experimental data from a single-cylinder engine. The optimized chamber improved fuel consumption by 3.2% compared to its shallow dish-type counterpart.modeFRONTIER helped us spot the optimal shape and further analyze the delicate tradeoffs regarding the thermal balance” concluded Takashima.

About IAECIsuzu Advanced Engineering Center, LTD. (IAEC) was established in 1990 for the purpose of carrying out research on future technology necessary for the product development of Isuzu vehicles. The team performs research related to commercial vehicles and diesel engines in 3 main fields — vehicle safety, environmental conservation,

and energy conservation. IAEC is committed to achieving the ideal automobile for an ideal future by forming organic relationships with engineers and researchers from R&D and educational institutions around the globe.

For more information:Caterina Moro, [email protected]

Figure 2. Degree of constant volume vs heat loss

Figure 3. Experimental heat release rates and cylinder pressures of re-entrant type, shallow dish A-type and ID280-type combustion chambers

modeFRONTIER supportSimulation-based engineering and Multidisciplinary Analysis and Optimization (MDO) are evolving as we speak, which means that an engineer’s formal education is no longer adequate to keep up with these changes, engineers need to be constantly retrained on the new technologies and methods as they become applicable to their industries. EnginSoft supports and promotes the modeFRONTIER software training in Europe with a network of specialized technicians and with a broad offering of educational and training services. You can start from a basic course, which is intended for new modeFRONTIER users, and move on towards an in-depth course on the use of specific simulation software packages and their features. If required a customized course can be created on demand for your specific organization’s needs. For more information contact: [email protected]

One of the particle methods for fluid dynamics simulations, MPS (Moving Particle Simulation) has begun to be used as an innovative and effective computational simulation technique for design and development. In contrast to one of the major particle methods SPH (Smoothed Particle Hydrodynamics), which is mainly used to model compressible flow and adopts explicit scheme for time integration, MPS is formulated to treat incompressible flow and uses semi-implicit scheme for time integration. In general, as most of engineering problems concerning fluid like materials in the manufacturing industries can be treated as incompressible, it is thought that MPS is suitable to treat these problems. In addition, the semi-implicit time integration scheme in MPS has an advantage in computational cost for longer manufacturing process. For these reasons, MPS is used in various fields of industries including automotive, power transmission, chemical and pharmaceutical, food & beverage, medical, and civil & environmental engineering. In this article, we introduce a real application from one of the leading automotive companies in Japan which is using the MPS based simulation software Particleworks.

Oil flow simulation in a reciprocating engine at Honda R&D It is important to predict oil flow in a reciprocating engine in design task. In this work the design of the structure of the breather chamber was investigated using Particleworks. The engine and the breather system are shown in Fig.1. The breather chamber is used to separate oil from blow-by gas. Efficient ability of oil separation of the breather chamber is required. Air-resistance (drag force of the air) in the breather chamber is considered by adding the drag term to Eq.(2). In addition, the drag term was defined considering the size of the particles as in Eq.(3). As the size of the real

oil mist involved in the blow-by gas is very small (1 m to 10 m), the coarse graining model technique was used to reduce computational cost. The MPS particles representing the oil mist were defined as the coarse graining model with the size of approximately 250 m to 500 m. The drag force was computed using grid based CFD software prior to the Particleworks simulation and defined as spatial function of external force in the breather chamber in the Particleworks simulation.

The governing equations for incompressible flow are the continuity and the Navier-Stokes equations:

where, ; density, u; velocity, P; pressure, ; diffusion coefficient, and g; gravity.

Particleworks, an innovative particle method simulation tool, playing an important role in previously-unattainable simulation problems


where, CD; coefficient of drag, air; mass density of air, ul; liquid velocity, ug; gas velocity, S; liquid body surface area, Dl ; particle size of MPS, dl; particle size of oil mist, and ml; oil mist mass.Two types of the breather chambers, type 1 and type 2, shown in Fig.2 were examined. The difference of these two chambers are 1) distance between the inlet hole and the collision plate of 30 mm for type 1 and 15 mm for type 2 and 2) the diameter of inlet of 15 mm for type 1 and 6 mm for type 2.The results of the simulation and the experiment of breather chamber type 1 and type 2 are shown in Fig.3 and 4 respectively. Type 1 can capture more oil mist on the collision plate and the oil flow toward the oil drainage holes at the bottom of the chamber can be seen clearly in Fig.3. While in the chamber type 2 the oil mist scatters under the collision plate and the oil flow to the oil drainage holes is not formed in Fig.4. According to the result, the breather chamber type 1 showed superior ability of separation of oil mist from the blow-by gas than the breather chamber type 2 and the simulation using MPS could reproduce the tendency of the real oil flow of these chamber types in the experiments.

ConclusionsMPS based numerical computational technology has already been used in the front-line design and development process as introduced in this article. MPS will be a more sophisticated technology to supplement the conventional numerical methods, e.g., FEM and Multi Body Dynamics for complex flow problems and fluid structure interaction problems in the near future.

AcknowledgementsThe author and Prometech Software wish to thank Mr. Haga of Honda R&D Co., Ltd. for the permission and opportunity to introduce their important research projects using Particleworks.

Source of Image at the top of the article:http://www.honda.co.uk/cars/new/civic-type-r-2015/overview.html

More detail can be obtained from “Honda R&D Technical Review Vol.26 No.2, 2014” if you have interest in this topic.Sunao Tokura, Prometech Software Inc.,

EnginSoft promotes and distributes Particleworks in Europe, the CFD software is based on an advanced numerical method known as the Moving Particle Simulation (MPS) method. The software is produced by Prometech Sofware Inc., Japan.

For more information please contact: Massimo Galbiati, EnginSoft - [email protected]

Fig.1 - Engine and breather system

Fig.3 - Oil flow in breather chamber type 1

Fig.2 - Breather chamber types compared in the simulation

Fig.4 - Oil flow in breather chamber type 2

This article presents an activity carried out by the Calculation & Simulation and Testing departments in Pierburg. The subject of the analysis is a vane oil pump for engine lubrication. This product represents a new generation oil pump in the automotive industry due to the possibility to reduce the displacement at high engine speed for fuel saving consumption. The aim of the article is to present the validation of a simulation model for the description of damaging behavior occurred in an aggressive durability test. This paper describes the evolution of both rigid body model and fluid model used for the calculation of internal loads which best fit the experimental evidences.

Product presentationVariable displacement oil pumps contribute significantly to the fuel saving capability in the automotive industry. In comparison to conventional pumps, they have the possibility to optimize the oil flow according to engine demand, with a significant reduction of power absorption. The oil flow rate of a vane pump, such as that of a general volumetric pump, depends on its actual displacement, i.e. the difference between the maximum and minimum trapped volumes. This difference is a function of

the pump eccentricity that is defined as the distance between the rotor axis (which is fixed) and the control ring axis. That’s why, in order to obtain the displacement variation in this kind of pumps, the control ring is made to slide or rotate into the housing. The variable displacement oil pump considered in this study is driven by engine crankshaft (Figure 1) so the rotor is in axis with the crankshaft.

Analysis of a vane oil pump mechanism failure: Multibody, fluid-dynamic and validation

Fig. 1 - 3D model of vane oil pump Fig 2 - Internal components


CFD modelA 1-D CFD model of the pump has been created (Figure 6). The model represents the pump system with the typical strategy of 1-D lumped parameters approach (representations of fluid volumes, pipes, orifices). The parts used for experimental test were measured and the same clearances were used in order to simulate exactly the pump that was tested. The model was built representing the geometry of each rotor chamber connecting the inlet and the delivery side of the pump and the most important leakages among rotor chambers and surrounding volumes. In all simulations, a proper model of oil aeration has been applied. Therefore the value of Bunsen coefficient was provided, time constant for passage free air – dissolved air into oil (tF D), time constant for the opposite passage (tD F) and the amount of free air initially present in oil. The values of parameters mostly affecting simulation results are given in Table 1.

Experimental testsExperimental test on bench is carried out under controlled conditions. Speed, pressure, flow-rate, temperature are set according to product requirements and are sensor-monitored. In this test, the control ring of the pump is physically locked to the maximum eccentricity position to accelerate the wear process: this has been recognized as the most aggressive condition for wear. Eccentricity is defined as the distance between rotor axis and control ring axis and it is directly proportional to the pump displacement. Working conditions for durability tests are: oil SAE5W30 at 120°C, 6500 rpm pump speed.

ResultsThe contact force between vane and control ring has been calculated for a set of simulation models and for a set of physical parameters. Figure 7 shows the evolution of the resulting contact force according to model 1 up to 4. In each simulation the contact force has been compared to a worn profile of the internal track of the control ring. Model 4 with 7% aeration and 10-5s time-constant free-to-dissolved is the combination that best explains the wear occurring in the durability test. The Polar plot in Figure 8 shows the comparison between the calculated contact force and the worn profile of the internal track of the control ring. The blue line shows the wear track measured by means of a profilometer in Pierburg metrological department. The green line represents the polar plot of the contact force calculated in model 4

with 7% aeration level and time-constant free-to-dissolved (tF D ) 10-5 s. A good agreement is found between the contact force and the worn shape of internal track of control ring.

ConclusionsIn this activity, different simulation models are shown for the validation of the contact force between the vane and the control ring that best fits the wear observed on the control ring after the durability test. A good synchronization between calculated contact force and wear signs is finally found. In the end this calculation has been fundamental to explain the root cause of the failure and to operate the best selection of the material of the components.

Andrea Barbetti, Matteo Gasperini, Fabio Guglielmo, Nicola Potenza, Raffaele Squarcini

Calculation & Simulation, Testing, R&D - Pierburg

Fig. 5 - Multibody model Fig. 6 - 1-D CFD model

Fig. 7 - Resulting contact force

Fig. 8 - Comparison

ProcedureThe analysis is focused on the contact between vane and control ring which is one of the most critical tribological couple in this kind of mechanisms. Calculations are based on a multidisciplinary approach. In fact the contact force between vane and control ring is calculated by means of multibody and CFD combined models. Physical parameters have been tuned until model validation is obtained, accordingly to experimental durability test. The Scheme in Figure 3 shows the workflow of the entire activity. The aim of this study is to calculate the contact force between the vane and the control ring that best explains the wear observed on the internal track of the control ring after durability testing, since the contact force is not easy to evaluate at the beginning, before some key hydraulic values are defined (like presence of dissolved air in oil and time-constant free-to-dissolved). Through this method the failure can be explained and countermeasure can be taken. Starting from a simplified multibody model and a simplified angular history of pressure, the target force is obtained and compared to wear signs. If they are not in accord, further iterative refinements of the model (multibody and/or fluid parameters) are needed until the target is matched.

In this activity, 3 refinement iteration loops were necessary to correspond with the experimental evidence. The main characteristics of the used models are summarized in Table 1. The first model contains only the side pressure effect on the vanes and uses a nominal shape of pressure signal. Side pressure depends on the pressure in upstream and downstream chambers (Figure 2),that are alternatively equal to 0 when the chamber is connected to the inlet and equal to delivery pressure when the chamber is connected to the delivery side. The second model is equal to the first, with the addition of the radial pressure effect on the top and the bottom of the vanes. The pressure on the top of the vane reply the pressure distribution into the gap between the vane and the control ring. The pressure on the bottom of the vane reply the pressure in the internal side of rotor. The third model is equal to the second but the nominal pressure signal is substituted by the pressure signal calculated via CFD simulation. Different aeration levels are simulated (0.5, 4 and 7%). The fourth model is equal to the third but the aeration level is fixed and

a range of time constant for passage free air – dissolved air into oil (tF D ) are simulated. The contact force is compared to the worn profile of the control ring as appear in the end of endurance test. It is a consistent approach since, according to Archard formulation for adhesive wear, the worn volume is proportional to the contact force.

Multibody modelA multibody model of the pump has been created. The model consists of several rigid bodies representing the internal components (see Figure 5). Inertia properties of the bodies have been obtained on the basis of 3D CAD drawings. Internal components are mounted with average clearances according to their tolerance range. The rotor motion has been imposed. The vane is able to slide radially during rotor revolution in a sort of “prismatic joint” and is subjected to mechanical and oil pressure loads. Mechanical interaction between rigid bodies has been guaranteed by solid contact elements; they consist on equivalent spring-damper elements which laws are based on overlap volume due to rigid bodies penetration. The hydraulic pressure effect has been modelled by force elements applied on each vane as a function of the rotor angular position (Figure 4). Both side effect and radial effect are modelled as in Figure 2. Friction effects are included in the model.

Fig. 3 - Workflow

Table 1 - Simulation models

Fig. 4 - Chamber pressure


The flux density induced on the turbogenerator frame walls is a topic of particular interest for turbogenerator manufacturers, as it produces significant hot spots on frame components, with particular overheating phenomena in proximity of the clamping bolts. To analyze this phenomena by means of a 3D FEM approach, specific skills in 3D FEM modelling and appropriate knowledge of the electromagnetic phenomena occurring in turbogenerators are required. In this kind of analysis, the complexity of the FEM modelling is slightly increased by the necessity to include in the model some components such as:

• the frame walls;• the clamping bolts;• the involutes of the stator bars;• the machine terminals;• the stator windings connections;

in addition to the main active parts normally considered in a 3D FEM model (i.e. the stator stack, the rotor bodies, the stator and rotor windings). For each of these components, a tedious job of CAD modelling has been carried out in order to obtain the simplified final model here presented, which has been considered a good compromise between the model accuracy and the calculation power needed. The implemented model, can be used to support both the designing during the development of new generators and to plan action to reduce the hot spots on the generator frame

in case they will be detected during the normal operation of the machine. For confidentiality, the obtained numerical results will be reported by adopting the per unit measurement system and, in the figures, many manufacturing details regarding the machine under investigation will be hidden.

The 3D FEM electromagnetic model of the turbogenerator end-regionIn the fig. 1ab, a view of the 3D FEM model of the turbogenerator end-region can be seen; the steps followed to implement the model are here below listed:

• importing the original 3D CAD model in ANSYS Design Modeler;

• simplifying the original 3D CAD model by means of the Design Modeler tools;

• exporting the simplified 3D CAD model from Design Modeler to ANSYS Maxwell.

A 3D FEM approach to evaluate the flux density and the eddy current induced on the turbogenerators frame walls and clamping bolts in order to analyze local overheating phenomena


In particular, the 3D CAD simplification process, has been important to faithfully approximate the real geometry of the machine by means of a CAD model greatly simplified. In fact, the meshed model in fig. 1b, composed of two million elements, has required around 60 GB to be run, which is not a large calculation power requirement today. Surely the mesh sizing reported in fig. 1b can be improved as the eddy currents calculation requires mesh elements smaller than 20÷25% of the skin depth that, for the frame material at 50 Hz, proves to be roughly a few millimeters. The method used to improve the model in order to avoid an excessive calculation power, was to refine the mesh step by step, by repeating the calculation reducing at every step the mesh elements only on the frame part where the flux density has proven higher. This under the reasonable hypothesis that the eddy currents induced on frame walls have a negligible reaction effect on the main field distribution. The calculation has been run at rated load, so before running it, the calculation of the excitation current and the load angle has been necessary to assign the input dataset for the simulation. These calculations have been done by means of the proprietary algorithm described in the Newsletter nr.4 of 2014.

The 3D FEM electromagnetic modelling of clamping bolts and gaskets As mentioned in the abstract, the frame components more exposed to the overheating phenomena due to the eddy currents circulation are the clamping bolts. In fact, being a rubber gasket that is usually applied between each frame part, we can say that these elements create an electrical connection between the frame components, thus permitting to the eddy currents induced on the frame to pass between different frame parts through the bolts. This phenomena must be analyzed with care, considering the dimensions of the frame walls (in the range of a few square meters) and the frame material properties μr=100 and FE=3÷4 [MS], it proves that just small values of leakage flux induced on the frame, it creates a relevant distribution of eddy current on it. The 3D FEM model here presented, takes in account both the clamping bolts and the rubber gasket as reported in the next figures 2ac. The rubber gasket has

been modelled by applying an insulating boundary condition to the frame surfaces in contact between each other (fig.2b), while the bolts have been considered as cylinder bodies connecting two frame parts (fig. 2c).

The calculation of the magnetic induction induced on the turbogenerator frame partsThe first calculation executed by means of the FEM model, aims to evaluate the flux density induced on the frame parts. This negative phenomena is well known to the turbogenerator manufacturers as for this kind of electrical machines, the leakage flux from the stator windings, cannot be considered a neglected effect. In addition, a significant contribution to the flux density induced on the frame, is given by the current flowing through the connection bars located in proximity to the end-region of the machine to connect the machine terminals to the end of the stator windings. A flux density distribution is reported in the next pictures 3ab, which shows the flux density induced on the frame, looking at it from the right and left sides. By looking at the pictures, the shielding effect should be evident, which is mainly due by the higher magnetic permeability of the steel in respect to the surrounding air. As a consequence of the shielding effect, to measure the magnetic induction induced on the frame components, the probes must be installed on the inner surfaces of the frame and not on the outer ones. This is also the reason why the adopting of non magnetic steel, to build the frame components, should be considered with care as the shielding effect, for non magnetic material such as the stainless steel, is really negligible. In fact, with the reduction of the shielding effect, it will be possible to have outside of the machine casing significant values of magnetic field, going against the EMC regulations.

Fig. 1a - The turbogenerator end-region simplified CAD model

Fig. 1b - The 3D mesh of the turbogenerator frame

Fig. 2a - The 3D FEM model turbogenerator end-region

Fig. 2b - The insulating boundary condition on the frame surfaces

Fig. 2c - The 3D model of the clamping bolt


In the Multiphysics model, the output dataset of the electromagnetic calculation such as the Joule losses dissipated by the induced eddy currents on frame, will be the input dataset for the thermal analysis. This load transferring is done within the thermal simulation environment by a “mapping” of the CAD model (the two models share the same CAD file, but not the same mesh). Normally, the joule losses transferring from the electromagnetic to the thermal FEM

model is quite accurate if the user is careful about assigning both to the FEM environment, similar mesh properties on the same components. The thermal calculation is run by applying to the border of each body its own coefficient of convective thermal exchange as it is shown in fig.8a thus avoiding to run a fluid-dynamic simulation. This simplified approach is possible when many experimental data have been collected during the years by the manufactures. The final temperature distribution obtained by means of the Multiphysics model is reported in the fig.8b, so the object to calculate overheating phenomena of the generator frame has been achieved.

Ing. Michele Raciti, Ing. Roberto BiondiAnsaldo Energia

The Company Ansaldo Energia is Italy’s largest supplier, installer and service provider for power generation plants and components and one of the world’s leading players in the sector.It is a full-cycle, integrated operator, with the capabilities to build turnkey power plants on green field sites using its own technology and its own independent design, production, construction, commissioning and service resources.Founded in Genoa in 1853, Ansaldo established itself right

from the outset as Italy’s leading mechanical engineering company, first working in the railway sector and then manufacturing marine engines until the start of the twentieth century, when it entered the power generation sector and built its first electrotechnical production facility in Cornigliano, where the company made electrical equipment, dynamos and other products.www.ansaldoenergia.it

The eddy currents induced on clamping bolts, “conduction” and “induced” phenomena The eddy currents in the bolts can be caused either by the currents which pass through the frame walls by crossing the bolts (“conductive currents”) or by the magnetic induction which directly affects the bolts, thus creating an eddy current circulation within the bolts (“induced currents”). To separate these two contributions which occur in the bolts at the same time, two versions of the same model have been implemented. In the first one (fig.4a) the bolts are in electric contact with the frame, while in the second one (fig.4b) they are completely insulated from the frame. So, in this last version, the eddy currents circulating in the bolts will be caused only by the magnetic induction which directly acts on them, while the eddy currents induced on the frame wall cannot flow through the bolts because of the insulating boundary condition applied on the bolts surfaces.

In the fig. 4ab, the calculation results of the Joule Losses dissipated by the bolts for each version of the model are reported. In fig. 4a, higher values of Joule Losses dissipated by the bolts can be seen, respect to the second case in fig. 4b where a significant reduction

of the joule losses in bolts can be seen. This confirms that the joule losses in bolts, is mainly caused by the “conductive” currents and not by the “induced” currents. So, more in general, we can conclude that by insulating the clamping bolts of the machine casing, will eliminate efficiently the problem of the hot spots created on the frame in proximity of the clamping bolts. In addition, by insulating the bolts, a reduction of roughly 10% of the total joule losses dissipated by the whole frame has been calculated. This beneficial result is mainly due by the insulating applying between frame parts, which obstructs the currents flowing. Moving now on the eddy currents distribution calculated on the frame walls, in fig. 5ab a view of the eddy currents induced on frame at any time instant is reported. Looking at the pictures, it should be evident why in some of the clamping bolts, hot spots phenomena can be detected. In fact, as it can be seen, the currents on frame tend to assume a vortex shape, creating extended current loops, which will find an easy closing path through the clamping bolts or every connection element between frame parts.

The stray losses dissipated by the frameThe implemented model can be used to calculate the stray losses dissipated by the frame due to the eddy currents induced on it. This result can be easily achieved, by means of the “field calculators tool” embedded within ANSYS Maxwell that, between the available post processing fields, it has the Joule losses dissipated by each part of the model. The result of the Joule losses calculation in conjunction with the time, in the graph in fig.6 is reported.

In terms of global efficiency of the machine, the Joule losses dissipated by the machine frame is absolutely negligible, but considering the number of the machines installed in the world and the working hours of each machine during one year, even a few kW of saving losses means an enormous saving of energy. Therefore, the phenomena of the eddy currents induced on frame, it should be analyzed with care. In addition to the efficiency considerations above mentioned, the calculation of the Joule Losses dissipated by the frame must be taken in account to verify the rise of temperature of the generator frame. To implement this check, the 3D FEM electromagnetic model has been imported within the ANSYS Workbench to create a 3D FEM Multiphysics model suitable to calculate the temperature reached by the frame components. In the next fig. 7, the interface of the electromagnetic & thermal model imported within ANSYS Workbench is reported; as it can be seen the electromagnetic & thermal model share the same CAD model, that has been transferred between the two model by means of Design Modeler.

Fig. 3a – Flux density induced on frame walls, right side viewed

Fig. 3b - Flux density induced on frame walls, left side viewed

Fig. 4a - Clamping bolts in electric contact with the frame

Fig.4b - Clamping bolts insulated by the frame

Fig. 5a - Eddy currents induced on frame, left side viewed

Fig. 5b - Eddy currents induced on frame, right side viewed

Fig. 6 – The Joule losses vs time dissipated by the frame

Fig. 7 - The ANSYS Workbench interface of the 3D FEM Multiphysics model for electromagnetic & thermal calculation

Fig. 8a – The 3D FEM thermal model of the frame

EnginSoft’s Experience in Electromagnetic analysesEnginSoft developed a wide experience in the analysis and simulation for electronics and electromagnetics. EnginSoft experience derives both from industrial and academics previous background and from current daily cooperation and support provided to industrial partners. A wide range of topics have been investigated in the latest 10 years, from low frequency to high frequency, with the aid of numerical tools for virtual prototyping.Explored applications fields cover industries including Defence and Aerospace, Oil&Gas, Automotive, Energy, Communications and Civil Engineer. In these fields, at low frequency, a high level of specialization was achieved in the design and analysis of electric mo-tors and generator, as well as electromechanical devices actuators (solenoid, injectors, field sensors,…). High frequency themes are investigated in the antenna and propagation field, for the analysis of radiating systems and their platform/system integration, microwave components and devices (resonators, filters).Customized simulation procedures and systems are currently settled to transversally investigate complex scenarios, such for EMI and EMC investigation, multiphysic analsysis and scripting toolkit development. In particular an automated procedure for the investigation of electromagnetic interference between high voltage lines and metallic pipes was developed in order to verify regulation fulfillment.For more information: Emiliano D’Alessandro, EnginSoft - [email protected]

Fig. 8b – The calculated temperature on the frame

Answers needed to improve new surgical methodTAVR has been performed only a relatively small number of times, so there are many unanswered questions. What are the forces exerted by the blood and aortic wall on the stent, and how long will the stent last under these loads? Is the friction between the stent and aortic wall sufficient to hold the stent and valve in the proper position over a long period of time? Answers to these and other questions could lead to the design of improved stents and help surgeons make more informed decisions on which type of surgery to use for specific patients.

There is no way to accurately measure forces on an implanted stent, so manufacturers of TAVR stents are simulating the process of implanting a stent and valve to better understand the method and estimate the forces on the implanted stent. This is a very complex analysis problem. The first challenge is modeling the highly nonlinear material properties of the shape-memory alloy (SMA) Nitinol™, which is commonly used for TAVR stents. Nitinol is an alloy of approximately 50 percent nickel and 50 percent titanium with a high biocompatibility and corrosion resistance. The most important characteristic of this shape-memory alloy is its super-elasticity, which allows self-expansion of the stent after release from a catheter. Simulation needs to include folding the stent prior to surgery (crimping) as well as releasing the stent against the aortic wall when it reaches its resting position in the aorta. An even greater challenge is the need for two-way coupled fluid–structure interaction, which shows forces on the stent that result from the relationship between flowing blood and the aortic wall.Mesh morphing and remeshing in the fluid domain is required because of large displacements of the replacement valve.

First successful TAVR multiphysics simulationCADFEM engineers overcame these challenges and produced what they believe to be the first successful simulation of a TAVR procedure that accounts for the impact of flowing blood on the stent after expansion. They used ANSYS Fluent computational fluid dynamics (CFD) software to simulate blood flow because its remeshing capabilities make it possible to accurately model the large displacement of the heart valve during simulation.

The engineers employed ANSYS Mechanical to model the stent and heart valve because the software can accurately model the memory alloy and orthotropic properties of the tissue valve. The orthotropic model accounts for the fact that the valve is stiff when pulled but bends easily. Both simulation tools run in the ANSYS Workbench environment, in which it is relatively simple to unite fluid and structural models in a two-way coupled transient simulation using system coupling. The transient fluid–structure interaction simulation was run from 0 to 0.3 seconds. CADFEM engineers used a constant-temperature super-elastic model for the material properties of the shape-memory alloy because they didn’t have enough data to model the effects of temperature changes. The model undergoes a phase change as it is crimped to its compressed state. Solid shell elements, which can model thin geometry more efficiently, made up the structural model.

A rigid model of the aortic wall was used in this initial analysis to save modeling and computation time. Engineers employed a bonded contact to connect the replacement valve to the stent; they used frictional contacts at the interface between the stent and the aortic wall. They applied the non-Newtonian Carreau model to predict variation of blood viscosity as a function of its shear rate. The boundary condition for the fluid model is utilized as a function for the mass flow rate of blood that models the heart’s pumping action. Blood flow causes the valve to open and close.


Fig. 1 – Deformation of stent due to blood flow as predicted by multiphysics simulation

Fig. 2 - Fluid–structure interaction tracks displacement of the heart valve


Aortic valve stenosis, a narrowing of the aortic valve, is the most common type of heart valve disease. It affects about 2 percent of adults aged 65 or older. Symptoms of this chronic progressive disease include chest pain, difficulty breathing, and fainting; in some cases, congestive heart failure can occur if the valve is not replaced.Surgical aortic valve replacement, which involves open-heart surgery with a heart–lung machine, has been the definitive treatment for aortic valve stenosis for over 40 years. The surgical team replaces the aortic valve with either a mechanical valve or a tissue valve taken from a human donor or animal. The operative mortality of aortic valve replacement in low-risk patients younger than 70 years is around 2 percent.

Long-term survival following aortic valve replacement is similar to that of patients of a similar age who do not have the condition.The number of elderly patients with aortic valve stenosis is increasing. These patients are often high-risk candidates for

traditional aortic valve replacement. A recent study reported an operative mortality rate of 24 percent for patients who are 90 years and older after open-heart surgery — so there is a need for a less-invasive aortic valve replacement technique.Transcatheter aortic valve replacement (TAVR) (also called transcatheter aortic valve implantation or TAVI) is a relatively new approach to traditional treatment. For this procedure, a tissue valve attached to an expandable stent is inserted into an artery near the groin and delivered via a catheter into position in the aorta. The stent is then expanded against the aortic wall to hold the existing valve open and secure the replacement valve in the proper position. This method eliminates the need for open-heart surgery.

A new heart valve replacement procedure modeledwith multiphysics simulation could eliminate the need for open-heart surgery

Change of Heart

Case Histories Newsletter EnginSoft Year 12 n°4 - 28 Case Histories29 - Newsletter EnginSoft Year 12 n°4

Simulation useful for surgeonsand stent manufacturersSimulation results showed the contact status and pressure at the stent–aortic wall interface. The results will be useful in evaluating the ability of proposed stent designs to lock the valve firmly in place. Simulation makes it possible to design the stent to avoid exerting too much stress on any part of the aortic wall. Von Mises stress–strain curves were generated for specific points on the stent model, and these curves could be used to predict fatigue life of the stent using a fatigue analysis model. The simulation also generated the time evolution of forces at joints between the valve and the stent. The model predicts a patient’s blood pressure following surgery, which is another critical factor in stent design. The systolic blood pressure in this case was about 175 mm Hg.CADFEM produced the simulation for Admedes Schuessler GmbH, the leading global provider of finished Nitinol self-expandable components to the medical device industry and a manufacturer of TAVR stents. This pilot study proved the feasibility of accurately modeling TAVR surgery. It provided results that can be used to optimize stent design for patients under different conditions, such as various amounts of hardening of the aortic wall. The next step is to incorporate a flexible aorta using either the ANSYS Mechanical anisotropic hyperelasticity model, which models fiber reinforcements in an elastomer-like matrix typical of living tissue, a userdefined model or simply a visco-elastic model.Finally, these results demonstrate the potential of multiphysics simulation to drive improvements in stent design and surgical procedures by providing insights that otherwise could be gleaned only through the experience of operating on human patients.

By Joël Grognuz, Team Leader Multiphysics,CADFEM (Suisse) AG, Switzerland

This work was made by CADFEM (Suisse) AG.Founded in 1985, CADFEM provides everything that is required

for the success of the simulation from a single source: First-Class software and complete, ready-to-use systems; comprehensive

services; the latest knowledge. CADFEM is the ANSYS Competence Center FEM in Central Europe.

www.cadfem.net

For more information in Biomedical CAE applications:Alessandra Pelosi, EnginSoft

[email protected]

Fig.4 - Stress prediction for selected point on the stent helps determine fatigue life of stent

Fig.5 -Simulation predicted the patient’s blood pressure after valve replacement. The blood pressure reading was high because the aortic wall was modeled without flexibility

Fig.3 - Simulation predicts contact status and pressure between stent and aortic wall

Fig.6 - A real expanded stent on which pericardium leaflets will be mounted

A structural analysis of a medical device (Heart Dumper) integrated inside a 3D finite element model of a ventricle is presented. Heart Damper is conceived as an implantable cardiac device, aiming at supporting heart pumping activity, reduced due to heart failure. Twenty five million people all over the world are affected by heart failure and their number is estimated to increase. Currently available treatments comprise pharmacological therapy, invasive device implantation or, in case of patients who are less than 65 years of age, cardiac transplantation. In this context, Heart Damper represents an innovative and promising solution in treating this disease.

The aim of the present study is to develop a computational model capable of simulating the interaction between the device and the ventricular chamber, so as to verify its efficacy in increasing ejection fraction and lowering cardiac work. First, a computational model has been developed in order to simulate physiological and insufficient behavior of the left ventricle, both during systole and diastole cycles. Volume and pressure values have been obtained from literature and have been used to calibrate the computational models for a normal and insufficient heart. Then an FE model of the Heart Damper was created and bonded to the ventricle. The performance of the insufficient ventricle alone or in presence of the device have been obtained. The results of the computational analysis show an increase in cardiac output and ejection fraction as well as an improvement in the ventricular efficiency.

IntroductionHeart failure (HF) is a chronic, progressive condition in which the heart muscle loses its contractile ability, becoming unable to pump sufficiently enough to maintain a blood flow that meet the needs of the body. An advanced condition can lead to the progressive failure of organs (e.g. brain, kidney, lung), leading to Multi Organ Failure (MOF) and finally to death. Incidences of HF is rapidly increasing; twenty five million people all over the world are affected by this disease and their number is estimated to increase. Currently available treatments comprise pharmacological therapy, invasive device implantation or, in the case of patients who are less than 65 years of age, cardiac transplantation. Nonetheless, existing pharmacological strategies fail to compensate for a weakening heart in the long term. Ventricular assist devices (VADs) or artificial heart (TAH) are highly invasive and high cost treatments. Heart transplantation has significant drawbacks, such as ineligibility over

Structural analysis of an implantable cardiac device - Heart Damper - for the treatment of advanced heart failure inside a 3D finite element model of the ventricle


constraints have also been introduced to take into account the presence of the entire ventricle during the simulation. The LV pressure versus time curve in Fig. 2 represents the applied loads to the internal walls of the ventricle.

Ventricle walls propertiesVentricle walls have been represented as muscle fibers (myofibers) embedded within a matrix (myocardium tissue). The myocardium tissue was modeled as a hyperelastic material following the 1-order Ogden model with the following values of parameters: μ=0.11MPa, =11.77, time-dependent bulk modulus. The myofibers are characterized by a time-dependent Young modulus, having a resistant section of 3mm2. Since it is not possible to set a time-dependent Young modulus in ANSYS, it has been set as temperature-dependent, with thermal expansion coefficient near to zero (1e-9).

In order to find the laws of variation of bulk and Young moduli which give the best fit with the healthy and pathological variation of the internal volume of the ventricle (Fig. 3), an optimization process has been performed. To this end an APDL-based routine for the implementation of a PID controller has been developed (Fig. 4), where the ventricle is the actuator and the internal volume of the ventricle is the target value.

Heart Damper modeling within the insufficient ventricleIn order to reduce the computational load, the Heart Damper device (Fig. 5A) has been simplified, keeping its main features (Fig. 5B). The wires constituting the nitinol frame have been modeled as monodimensional beam elements (BEAM188), with a Young modulus of 75000MPa, a density of 6.45g/cm3 and a Poisson ratio of 0.3 (austenitic state).

The membrane surrounding the frame is made of silicone and the following values have been set for the material: Young modulus=50MPa, density=1.12g/cm3 and Poisson ratio=0.44. As to the constraint between the device and the cardiac wall (bonded contact), the following assumptions have been made:

• The constraint determines the isolation of the portion of the ventricle underneath the membrane;

• The volume underneath the membrane is filled with an uncompressible fluid (HSFLD242, Fig. 5C).

Cyclic symmetry constraints have also been introduced (Fig. 6A) and the LV pressure has been applied as shown in Fig. 6B.

Both for the insufficient ventricle alone and for the insufficient ventricle with the Heart Damper device, the laws of variation of Young and bulk moduli have been found using the same procedure as for the healthy ventricle. In particular, the model with the device, new internal volumes have been calculated considering the partitioning performed by the Heart Damper and new values for cardiac output and end systolic volume have been obtained setting the ejection fraction to 40%.

Fig. 4 - PID controller schematic, where: yp=LV internal cavity volume from literature (set point); y=calculated LV internal cavity (measured output); E= Young modulus (variable); K=Bulk modulus (variable)

Fig. 5 - A) Heart Damper device; B) Mesh of the simplified model; C) Mesh of the uncompressible fluid

Fig. 6 - A) Cyclic symmetry constraints applied to Heart Damper device; B) Applied pressure

65 years, inability to meet demand and post-surgery complications. In this context, Heart Damper (HD) represents an innovative and promising solution in treating this disease. Heart Damper (patent EP2579909) is an implantable cardiovascular device, intraventricular, definitive and doesn’t require external power supply. It consists of a circular-shaped membrane stretched on a radial metal structure, moving in a paradoxical way towards the valvular plane, exploiting heart residual energy. Its continuous closing/opening cycles are induced by the heart systolic pulsation. It aims to guarantee a low hospitalization index, improving the quality and lengthening the life of people affected by this disease. The objective of the present study was to realize a finite element model to verify the effectiveness of the HD device in increasing left ventricular efficiency, compromised by heart failure. Computational methods have encountered an increasing interest and a rapid evolution in the recent decades, due to their potential to significantly reduce time and costs of device design. Furthermore they are becoming accepted tools during the certification phase, also by regulatory bodies. Different models have been implemented in order to try to accurately reproduce the behavior of the heart muscle during the heart cycle. In this study the integration of the model of an active ventricle, reproducing both the behavior of healthy and insufficient heart, with the Heart Damper device is shown.

MethodsGeometry, constraints and boundary conditionsA three-dimensional FE model representing the human left ventricle was developed using ANSYS. The ventricle geometry was simplified and represented by an ellipsoid truncated at two-thirds of the major axis. A parametric approach was adopted in order to adapt the model both to physiological and insufficient ventricles characterized by different parameters (long and short axes, wall thickness, etc.). Starting from the complete model (Fig.1A) a reduced model (Fig.1B), considering a twelfth of the ventricle, has been created in order to lower the computational load during the calibration and

optimization phase. The model was discretized with hexahedral elements (SOLID185, Fig. 1C), with reinforcing elements (REINF264) oriented circumferentially, representing the muscle fibres (Fig. 1D). The internal volume of the ventricular cavity was modelled as nearly-incompressible fluid trough fluid dynamic elements (HSFLD242, Fig. 1E). To prevent rigid body motion of the model, degrees of freedom for all nodes at the base were suppressed in the longitudinal direction (UY=0). Cyclic symmetry

Fig. 1 - A) Geometry of the left ventricle; B) Reduced model; C) Mesh; D) Reinforcing elements; E) Mesh of the internal volume

Fig. 2 - Variation of left ventricle pressure versus time throughout one cardiac cycle

Fig. 3 - Healthy (A) and insufficient (B) ventricle volume variation throughout the cardiac cycle


ResultsResults of the assessment of Young and bulk moduli, compared with those of the healthy (red line) and insufficient heart (grey line), are shown in Fig. 9. Values of end systolic volume, end diastolic volume, cardiac output and ejection fraction for each model are presented in Table 1.

Results show that the positioning of the HD inside the insufficient ventricle determines: • The cardiac output (CO) increases by 15%; • An improvement in the ventricular efficiency, as the graph of

the Young modulus shows (Fig. 9): a clear improvement in the performance of the insufficient ventricle can be noticed after the positioning of the HD device.

ConclusionsThe simulations reveal that introducing HD inside an insufficient ventricle lead to a significant increase in both the cardiac output and the ejection fraction (40-50% of initial values). These results confirm the validity of the preliminary hypothesis and prove the effectiveness of the research: it is indeed reasonable to assert that the Heart Damper is the first implantable totally mechanical device ever designed. So far, it seems to be the only device without external power supply, able to treat with positive spillovers patients affected by heart failure, being the incidence of this disease continuously increasing. Furthermore, the enhancement of the cardiac efficiency (as shown by Young modulus curve), due to the positioning of the device, shows that the therapeutic effect of HD could be far beyond expectation, since it directly affects the cause of the pathologic condition (loss of efficiency). The positive outcome leads to a new scenario, from a medical, scientific and economic side. The prospect that HD entails an improvement of the cardiac efficiency

can modify the possibility to implant HD, both considering the number of patients to be treated and the time of the implantation, with the possibility to operate when the first significant symptoms occurs and not only at a late stage of the disease (IV class of NYHA). From the point of view of the patient, this chance entails the regression of the pathological condition, with the consequent lengthening of the life expectancy and the improvement of its quality.

Caterina Turrisi, Roberto Parravicini - Eucardia Michele Camposaragna, Sergio Sarti - EnginSoft

Daniele Galavotti - RanD

Fig. 9 - Comparison of Young modulus (A) and Bulk modulus (B) for the healthy (red line), insufficient (grey line) and insufficient with HD (blue line) ventricles

Table 1 - Values of ESV (end systolic volume), EDV (end diastolic volume), CO (cardiac output) and EF (ejection fraction) for a healthy, insufficient and insufficient + Heart Damper ventricle

Challenge and Solutions in Biomedical Industry with SimulationThe medical device industry is continually pushed to develop new devices that improve the quality of medical care and the way treatments are delivered. This growth is driven primarily by the quality of life demands of an aging population. The resulting escalating costs call attention to the need for process understanding, rapid innovation and solution efficiency.

Combining design, manufacturing and performance testing in an integrated engineering simulation environment enables large-scale optimization of a candidate device, from development to production to regulatory testing. Simulation offers the additional benefit of reduced experimentation and animal testing required to evaluate each design.

Today, casting simulation represents a helpful and effective tool for designers to investigate, in advance, the influence of the casting process on material strength. Moreover, this sharing of knowledge between design and manufacturing engineering, usually called Concurrent Engineering, plays a key role in the design of heavy ductile iron casting components, due to their large distribution on microstructure and mechanical properties inside the item itself. Because of the measurement’s complexity, only few studies have investigated so far in comparing simulated results and experimental data in the field of long solidification time. In this study, some thermocouples are placed inside a casting to record the cooling curves and several tensile specimens which were core drilled in the area with different cooling conditions. The experimental data were then compared with the results obtained with a solidification and cooling simulation. The comparison shows a good agreement between the experimental and calculated cooling curves, and the mechanical properties.

IntroductionToday, ductile iron is one of the most used materials in critical engineering applications, such as wind turbines, gas and steam turbines, nuclear waste storages, big engine blocks and hydraulic presses, due to its excellent mechanical properties and castability. For heavy and thick items, from few to hundreds of tons, it is critical to provide design engineers with consistent and reliable data on mechanical properties and microstructure, as well as how they change inside the item itself. Typically, casting design is based on average properties from international standards, but, it is well known, they are not homogenous inside the casting due

to different manufacturing processing parameters, for example, local solidification and segregation path. It is also acknowledged that the mechanical properties of ductile iron are strictly related to the microstructure. The factors that influence the mechanical properties include chemical composition of the matrix, graphite nodules shape and size, ferrite to perlite ratio, dimension of the ferrite grain and pearlite lamellae spacing. A lot of effort has been made to correlate manufacturing casting parameters to the microstructure and the microstructure to the mechanical properties in order to predict and to map them on the casting. Unfortunately, there is limited research focused on the prediction of microstructure and mechanical properties in large items, where lengthy solidification time and huge segregation have a strong impact and the established prediction model begins to incur problems. Moreover these studies are focused on specific geometries and it’s not easy to transfers the results onto generic commercial items. In this study we used a generic commercial ductile iron casting to obtain experimental data of the cooling curves, microstructure and mechanical properties inside the castings. A solidification simulation was performed using MAGMAsoft 5.3 and MAGMAiron and the simulated results were compared to the experimental ones.

Comparison of casting simulation results and experimental data in heavy section ductile iron production


are shown: the continuous lines represent the cooling curve as recorded (red Temperature vs. Time, blue Cooling Rate vs. Time), while the dotted lines represent the curve where the data recording failed; the green dashed lines represent the LOESS smoothed cooling rate curves.

Mechanical Properties and Microstructure AnalysisAfter the shake out and fettling the appendix were cut and 7 core drilled specimens were also cut out using a CNC machine tool (Fig. 5). Their position was selected to get solidification time between 1 to 6 hours. Before cutting the appendix, it has been checked using

UT to avoid developing some traces of refractory coating. Moreover their surfaces were investigated using PT to check the presence of porosities. From the drilled specimens, 14 tensile specimens were extracted with a diameter Ø 10mm (Fig. 6). The specimens have been tested in a Zwick/Roell Z250 to check the tensile properties (Rp0.2, Rm, and A%).After the mechanical tests 7 of them have been cut along the main axis, polished and checked with an optical microscope LEICA DM 6000 M. The graphite nodules have been checked using the image analysis software Fiji. For each specimen 800 mm2 were investigated and the nodule count was recorded. After this the specimens were etched with Nital 2% to check the ferrite to pearlite ratio (Fig. 7).

Simulation Set-UpA solidification and cooling simulation was set up using MAGMAsoft 5.3 and MAGMAiron. We decided to skip the filling simulation due to the dimension of the item: since the solidification takes several hours the filling has a negligible impact on it. The mesh counted a total of 2 741 706 elements, 598 542 of which metal cells (Fig. 8). The material used were the GJS400 presents in the MAGMAsoft database with the chemistry changed according to Tab. 1, the chillers and the sand were respectively GJL300 and furan resin bonded according to MAGMAsoft material database.

During the pre-processing some “virtual thermocouples” were placed in the same positions as the real thermocouples’ tips and in the centre of the tensile test specimens in order to record the simulated cooling curves and to get the simulated mechanical properties and microstructures results (Fig. 9).

Results and DiscussionCooling Curves ComparisonFor the simulated cooling curve of the thermocouples T2 and T4 the first derivative was calculated as if it was made for the real component. For each thermocouple both the cooling curves and its first derivative were compared (Fig. 10). In the solidification field the simulated results concurred with the experimental results. Only at the beginning, there are some mismatches due to the thermal inertia of the thermocouple coating as discussed previously. Looking at the solidification time we can find that the predicted and actual corresponded perfectly (Table 2). As mentioned before only Thermocouple T4 was measured after the solidification, but still the simulated curve matches well the experimental data.

Fig. 5 - The appendix during the specimens core drilling

Fig. 6 - The tensile specimen geometry

Fig. 7- Microstructure specimen not-etched (left) and Nital 2% etched (right)

Fig. 8 - Mesh of the casting

Experimental Set-UpTo acquire experimental data on the temperature field during the solidification process and to check the mechanical properties and microstructures analysis inside the casting, an experiment was set up using a commercial ductile iron casting. The material investigated is ferritic ductile iron EN-GJS-400-18 with the

chemical composition showed in Table 1, produced in a furanic resin bonded sand. An appendix added was directly attached to the casting (weight ~ 12000 kg), (Fig. 1). The appendix, was introduced during the sand filling, with a polystyrene pattern removed before assembling the mould box cope and drag, this weighed ~1050 kg with the

following dimension 1200mm x 400mm x 320mm (Fig. 2). The appendix was added in order to introduce some thermocouples to record the cooling curve in the middle of the thickness 320mm and to core drill some mechanical specimens. During the moulding 8 thermocouples, type-K Inconel sheath

0.5÷1mm, were located in different areas of the appendix, with their tips on its centre (Fig. 3). The thermocouples were coated with different refractory material in order to withstand up to the shake out of the item. The behavior of the coatings have been investigated in a previous experiment using smaller items and comparing the effects with non-coated thermocouples: the results shown that after an initial transitory interval, due to the heat resistivity of the

coating, the cooling curve measured by coated and non-coated thermocouples were almost identical. Since the solidification time in the experimental item is much longer than the initial transitory we assumed the effect of the coating negligible. The data logger to record the cooling curves was an Agilent 34970A.

Cooling CurvesThe cooling curves were recorded from the pouring up to the shake out time. Unfortunately only 2 thermocouples out of 8 gave us reliable results in the solidification field, due to the high aggressive environment they were located. Only one thermocouple recorded data after the solidification down to about 800°C. The measured cooling curves were exported in Microsoft Office Excel, and the first derivative were calculated. The first derivative have been smoothed using a LOESS algorithm to evaluate the end of solidification: in which the minimum of the first derivative was considered the end of solidification and the corresponding time was assumed as the solidification time. In Fig.4 the measured cooling curves

Fig. 1 - Geometry of the casting (gray) and the appendix (red)

Table 1 - Chemistry composition

Fig. 2 - Geometry of the casting (gray) and the appendix (red)

Fig. 3 - Coated thermocouples during the mould box assembly

Fig. 4 - Thermocouple T2: cooling curve and cooling rate (left); Thermocouple T4: cooling curve and cooling rate (right)


ConclusionsIn this article we present a comparison between simulated and measured cooling curves, mechanical properties and microstructure characteristics in a commercial heavy section ductile iron casting.The cooling curves measured and calculated in the centre of a section 320mm thick correlate well as does the cooling rate. For the solidification time there is no difference between the measured and calculated values. Concerning the temperature field and its development over the time we can assume MAGMAsoft can predict it correctly for large ductile iron items.

Concerning the mechanical properties the Rp0.2 and Rm values are in agreement if we compare the minimum values supplied by MAGMAiron. The A% shows some differences probably due to the stress-strain cast iron behaviour and the need to set-up the melt treatment properly in the simulation step.For the microstructure the P% is predicted correctly for all the specimens, while the Nc is subjected to some variation: also in this case we consider them due to a not perfect set-up of the melt treatment.

AcknowledgementsThe Authors want to thank “Fonderia Ariotti SpA” and the “Interdepartmental Center for Industrial Research on Advanced Applications in Mechanical Engineering and Materials Technology CIRI-MAM” of the University of Bologna for the support in the experimental steps.

Giacomo Bertuzzi, SACMI ImolaGiampietro Scarpa, EnginSoft

The CompanySACMI is an international group manufacturing machines and complete plants for the Ceramics, Packaging (including Beverage and Closures&Containers), Food industries and Automation - markets in which it is a recognized worldwide leader. Its strength lies in the application of innovative technology, the outstanding position of the Group on international markets and its commitment to research and development and providing customers with top-flight quality and service.

Fig. 12 - Measured vs. calculated A%

Table 4 -Max. and mean percentage error for A% (min., mean and max.) and best fit

Fig. 13 - strain hardening phenomena and the effect on A%

Figure 14: measured vs. calculated Nc.

Mechanical Properties and Microstructure Analysis ComparisonThe mechanical properties and microstructure have been predicted using MAGMAiron: we simulated 0.2% Proof Strength (Rp0.2), tensile Strength (Rm) and Elongation (A%). MAGMAiron supplies for each of the mentioned properties, three values: minimum, medium and maximum. In Fig. 11 the measured vs. calculated values of Rp0.2 and of Rm are respectively shown. For both the two properties we considered only the minimum value since it matches the measured value the best, while both mean and maximum values were skipped since they are not aligned to the usual values we see in standard production. In Table 3 a summary of the maximum and mean percentage errors are reported for both the two values.

For A%, instead, all the three supplied values (min., mean, max.) were considered since there was not a clear trend in choosing one of them (Fig. 12). In Table 4

a summary is presented: in this case we added a value “Best Fit” where we calculated the max., and mean percentage error choosing for each comparison the closer of the three.

The difference between the calculated and measured values can be due to set-up of the melt treatment in the simulation software that can be tuned to get a better prediction of the microstructure and in the strain hardening behavior of the material. In Fig. 13 an example is shown: since the Rm values are in the area of the flattening of the stress strain curve a small change in the Rm can produce a huge effect on the A%.

Concerning the microstructure we considered the Pearlite content (P%) and the Nodule Count (Nc). The P% calculated was less than 3% for all the specimens in good agreement with the measured values that showed only few percent of Pearlite. The Nc (Fig.

14) is in good agreement for the specimens of the “Int.” series, while for the “Ext.” series the deviations are bigger. As mentioned for A%, probably, the differences are related to the set-up of the melting treatment in the simulation pre-processing. Further investigation in this field are needed in the future to try to improve the prediction.

Fig. 9 - “Virtual thermocouples” location: tensile specimens “Int.” (a), tensile specimens “Ext.” (b), actual thermocouples’ tips (c)

Fig. 10 - Measured vs. calculated cooling curve and cooling rate thermocouple T2 (left) and thermocouple T4 (right)

Fig. 9 - “Virtual thermocouples” location: tensile specimens “Int.” (a), tensile specimens “Ext.” (b), actual thermocouples’ tips (c)

Fig. 11 - Measured vs. calculated Rp0.2 (left) and Rm (right)

Table 3 - max. and mean percentage error for Rp0.2 (left); max. and mean percentage error Rm (Right)


Energy Demand and Nuclear PowerThe increasing world energy demand pushes towards the search of sources alternative to coal and gas. The continuous increase of uranium production and demand (Figure 1) indicates that nuclear power is seen as a valuable source. Indeed, China, India, South Korea and Russia remain committed to it even after the Fukushima accident and the global uranium demand forecast indicates a long-term growth.

Uranium EnrichmentIn the world, today, there are about 500 commercial nuclear power reactors operating or under construction, most of which require uranium enriched in the U-235 fissile isotope for their fuel. One of the most widely used technology for enriching uranium is represented by centrifugation of gaseous uranium fluoride. In gas centrifuge enrichment plants, hundreds or even thousands of centrifuges are arranged in cascades. Each cascade is made up by stages containing a certain number of centrifuges (Figure 2). Recent news about Iranian nuclear program has clearly shown how uranium enrichment is a sensitive technology from a non-proliferation point of view because it can be used for producing atomic weapons. For this reason uranium enrichment activities need to be subject to tight international control.

Nuclear SafeguardsMost countries participate in international initiatives designed to limit the proliferation of nuclear weapons. Nuclear safeguards are measures to verify that states do not use nuclear materials to develop

weapons and that they respect their obligations under international non-proliferation treaties. The European Union has set up a system of nuclear safeguards under the Euratom Treaty. In this framework, The Nuclear Security Unit of the Institute for Transuranium Elements at JRC Ispra provides research, technology, instruments,

Simulation Helps European Nuclear Safeguards Activities


technical services and training to the inspectors of the International Atomic Energy Agency (IAEA). The Non Proliferation analyses of Gas Centrifuge Enrichment Plants make regular use of advanced numerical modelling techniques supported and/or validated with data acquired during field inspections. By adopting this approach, normal and off normal conditions can be tested at an early stage improving the odds of a timely detection of eventual misuses or diversions of Nuclear Materials.

Numerical Simulation of Gas Centrifuge Enrichment PlantsThe numerical simulation of Gas Centrifuge Enrichment Plants presents many important challenges:• Fluid properties: uranium hexafluoride is a heavy gas, having

a density about 10 times larger than air;• Flow conditions: the system works at low pressure (around

500 Pa) and with extremely small flow rates, in the order of micrograms per seconds;

• System complexity: plants contains hundreds or thousands of centrifuges;

• Physical complexity: the isotope separation process takes place into centrifuges spinning at hypersonic velocities.

A system level fluid-dynamic approach was implemented using the advanced software tool Flowmaster. In order to model the complexity of the system, a bunch of custom components were implemented into the library. The most important one is the component capable to model a single centrifuge or a single stage of the cascade by providing the separative power of the centrifuge as a function of gas flow rate and the number of centrifuges in the stage.The preliminary simulations performed on simplified network models (Figure 3) show that a system level approach is capable to model the main features of an uranium enrichment cascade on workstations in times ranging from few seconds to few hours depending on the length of the simulated times. In particular the model allows a reliable estimation of the cascade separation performances under different operating conditions opening the way to effective simulations of misuse and diversion scenarios.

Giovanni Mercurio - European Commission DG JRCInstitute for Transuranium Elements Nuclear Security

Alberto Deponti - EnginSoft

Institute for Transuranium Elements (JRC-ITU)The mission of the JRC Institute for Transuranium Elements (JRC-ITU) is to provide the scientific foundation for the protection of European citizens against risks associated with the handling and storage of highly radioactive material. JRC-ITU’s prime objectives are to serve as a reference centre for basic actinide research, to contribute to an effective safety and safeguards system for the nuclear fuel cycle, and to study technological and medical applications of radionuclides/actinides.JRC-ITU works very closely with national and international bodies in the nuclear field, both within the EU and beyond, as well as with the nuclear industry. In addition to playing a key role in EU policy on nuclear waste management and the safety of nuclear installations, JRC-ITU is also heavily involved in efforts to combat illicit trafficking of nuclear materials, and in developing and operating advanced detection tools to uncover clandestine nuclear activities. JRC-ITU provides the expertise and access to the necessary special handling facilities for the study of the actinide elements, which is of relevance for the issues related to nuclear power generation and the radioactive waste treatment and disposal, but also for the advancement of science in general. Another key role is in the study and production of radionuclides used in the treatment of cancer.

Fig. 1 - World uranium production and demand trends

Fig. 2 - Schematic representation of a centrifuge cascade for uranium enrichment

Fig. 3 - Network model of a 5-stage cascade implemented in Flowmaster


implement the curves that describes properties variability as a function of temperature.

Fig. 2 - Transport equation for internal energy i

Fig. 3 shows the algorithm of the modified solver. The thermal equation is solved at the end of the internal and temporal loops. The first one updates all the fields for each time step and the second one increments the temporal variable and then repeats all the steps.Using the implemented solver it has been possible to simulate the first phase of HPDC process: the details of this step will be described next. Fig. 4 contains a representative frame of a simulation that shows a half of the shot sleeve, the separation surface and temperature field.

MATHEMATICAL MODELDesign modelHPDC process are controlled by a very high number of process parameters that influence the quality of the cast parts. Among all the influencing variables two geometrical (diameter and length of the shot sleeve) and two process parameters (initial filling of the shot sleeve and first phase velocity) have been defined as relevant, based on the HPDC process knowledge: they are the inputs for the DOE. Internal diameter (D) and length (L) have been chosen to describe the geometry of the shot sleeve. All other dimensions have been parameterized as a function of these two variables. A finite volume mesh has been employed for properly describing fluid flow phenomena. Finally, only a half of the shot sleeve has been modelled and simulated (using symmetry boundary conditions on the middle plane) to save computational time (Fig. 5).The third parameter is initial filling (F): it represents the initial volume of the melt into the shot sleeve (percentage value of the entire volume) and it is strictly correlated with its initial height (Fig. 6) once the diameter and length have been defined for each single

Fig. 3 - Algorithm for the numerical model implemented into a CFD code. Parts in italic font represent changes to the original solver

Fig. 4 - 3D view of computational domain used to describe the system inside the shot sleeve. As can be seen, only a half of cylinder has been implemented. This frame shows temperature field and separation surface between air and melt alloy

Fig. 5 - View of the computational domain discretized using a finite volume the mesh


In HPDC manufacturing process, the final quality of castings is related to the injection phase: the dynamics of formation of melt waves and their reflection on the walls of the sleeve can cause defects due to air entrapment during filling. The development of numerical and mathematical models are described in the present article. The numerical model consists in the implementation of thermal equation into the open-source CFD code openFOAM™. The mathematical model consists in a Design Of Experiment (DOE) generation and execution using the previously created model. They have permit to study the correlations obtained by Response Surface Methodology (RSM) between the percentage of trapped air in function of the relevant variable parameters of the process.

IntroductionIn HPDC process, the final quality of castings is related to the first stage of injection. During this phase, the wave system in the melt due to plunger acceleration causes air entrapment, inducing air porosity into the component. This could have a detrimental effect on mechanical properties and produces internal and surface defects. To prevent these phenomena, it is important predict how the relevant process parameters are correlated to air porosity and then keep them under strictly control. In the present work, this objective has been achieved through the development of a numerical model that describes the thermo fluid dynamics behavior inside the shot sleeve. The developed solver has been used to simulate several designs with different combinations of input parameters: the designs have been generated using a mathematical model through DOE techniques.The models allow to determine the response surface that has been used to analyze the percentage of entrapped air.

Numerical ModelThe aim of this phase of the project has been the development of the solver that permits to calculate the dynamics of the two-phase system alloy-air inside the shot sleeve.The solver has been implemented in the CFD framework OpenFOAM™: it is based on the existing “interDyMFoam”, solver for two incompressible, isothermal immiscible fluids using a Volume Of Fluid (VOF) phase-fraction based interface capturing approach, with optional mesh motion. This solver takes into account most of the physical phenomena that govern the fluid motion inside the shot sleeve, its main features are (Fig. 1):• Multiphase flow, which enable the description of the two

phase system alloy-air;• Incompressible fluids, reliable also for the air phase because

the stream velocity is relative slow;• Dynamic mesh, which allow considering the movement of the

plunger.A preliminary study on the fluid-dynamics equations and discretization methods has pointed out the necessity to customize the existing solver because of one of its main features: the isothermal assumption.The solver has been modified and expanded by adding features:• The energy equation (Figure 2) has been added to the existing

Navier-Stokes transport equations. Also, it has been necessary to implement the temperature field.

• Variable thermo physical properties as a function of the temperature field. In this way the fluid properties become fields itself because they depend on temperature, time and spatial coordinates. These properties are: density ( ), thermal conductivity ( ), specific heat (cp) and kinematic viscosity ( ). The user can manage different type of aluminium alloy by

Thermo-fluid dynamics model of two-phase system alloy-air inside the shot sleeve in HPDC process

Fig. 1 - Physical phenomena taken into account for the original solver and modified ones.

Case HistoriesCase Histories 43 - Newsletter EnginSoft Year 12 n°4 Newsletter EnginSoft Year 12 n°4 - 42

in the directories structure to initialize every simulations and add them in a queue on the cluster system. The simulations require a lot of computational power: for this project, it has been allocated 64 processors divided in groups of eight (8 simulations with 8 processors for each have been executed simultaneously). The whole DOE simulation has required about of 5 months.

Application of a RSMThe aim of this work is the determination of a predictive model that enable the correlation between “entrapped air” and some relevant process parameters during the first phase of HPDC process. The “entrapped air” is the volume of air inside the fluid domain at the end of the first phase of injection. To compare results for different geometries it is not possible to adopt the absolute value of air volume (to bigger shot sleeves correspond also greater air volumes). For this reason, the “rate of trapped air” R has been introduced. It is calculated by dividing the final volume of air (Vair) respect the final total volume of the fluid domain (Vtot).

The R values are now comparable between designs. The predictive mathematical model consists in the definition of the response surface achieved through Response Surface Methodology (RSM). The next validation procedure has been adopted to find the algorithm with the best fit: the entire DOE table has been split into two distinct tables, named “training table” and “validation table”. The “validation table” contains 10% of the total number of designs. Some functions (response surfaces) have been generated, based on the “training table”, using different types of algorithms (Gaussian Processes, Kriging, Radial Basis Function, Neural Networks and Multivariate Polynomial Interpolation). The absolute, relative and normalized errors have been evaluated for each functions, based on the “validation table”. The errors are defined as the difference between the values of R derived from the simulations and the values of R obtained from the predictive models (Fig. 10). The validation procedure has highlighted that Radial Basis Function algorithm interpolated in the best way the experimental data and minimize the errors. Therefore the RBF algorithm has been adopted to re-calculate the predictive model function using the entire DOE table.

Obtained resultsThe predictive model can be utilized to study some different cases of the first phase of HPDC process. With this tool, in fact, it is possible to find the optimal value for one (or more) parameters in function of the other ones. Next, it has been reported two possible way of investigation of the response surface.

Finding of the plunger’s optimal velocity with a fixed shot sleeve geometryOne of the main target in HPDC process is the research of the optimal velocity of the plunger to minimize the entrapped air given the geometry of the shot sleeve. With this tool it is possible to explore the surface response fixing the geometry variables (D and L) and plotting the R values as a function of V and F into a 3D graph (Fig. 11).In this way, the process engineer can evaluate the optimal velocity for its initial filling value F (imposed by the volume of the cavity). For example, for F = 50 %, the model reports a minimum value for R at V = 0.6 m/s (Fig. 12).

Ideal shot sleeve choice as a function of the melt volumeIf it has to select from two shot sleeve to realize a component, it is possible to interrogate the mathematical model in order to support the choice. The constraint is on the volume of the melt

Fig. 9 - Excel spreadsheet that reports inputs and outputs for each design. Inputs derive from SOBOL algorithm of modeFRONTIER and the results originate from an OpenFOAM function and transcribed in this table with a script

Fig. 10 - Differences between R values derived from DOE simulations (blue line) and the ones obtained from RBF predictive model (orange line)

design.The fourth parameter is the first phase velocity (V). The shot profile of the first phase of injection presents a first section with reduced (and constant for all designs) speed to avoid leakage and a second section at a constant velocity V (Fig. 7): the velocity profile has been implemented in the solver trough a boundary displacement condition on the plunger wall. This in turn causes the cells to stretch to adapt the whole mesh to the updated dimensions (dynamic mesh motion without topology changes).All the remaining variables have been set as a function of the four previous input parameters. It has been necessary to define a reliable and consistent condition for the end time of each simulation with the aim to properly compare the results of entrapped air volume percentage belonging to different designs (different geometries, initial filling conditions and velocities imply different simulation times). The simulations end when the total volume of the cylinder is equal to the initial volume occupied by melt alloy (condition of full chamber).

Definition and execution of the DOEEach design that has been simulated require a reliable value for the input parameters to simulate conditions that are consistent with the real process. To avoid unfeasible designs some constraints have been imposed on combinations between input variables. These consist in:

• Geometric constraints: to avoid “squat” or “slim” shot sleeves, diameter and length combination must respect this empiric relation: -10<D-0.1L<30 [mm]

• An upper limit for the initial filling value has been imposed to 70%;

• In addition, an upper limit of the simulation time has been imposed to remove possibilities for long shot sleeve associated with slow velocity and long filling times. This limit is equal to 2.5 s.

In this way, only the relevant cases in foundry practice have been simulated. As a basis of the planning of the DOE it has been used modeFRONTIER, an integration platform for multi-objective and multi-disciplinary optimization. SOBOL algorithm has been adopted to uniformly distribute a given number of experiments in the design space (standard practice when the final target is the creation of a predictive model), with respect to the upper and lower limits for each input variable and the four described constraints (Fig. 8).The algorithm returns a table that reports, for each row, the values of the four input variables (D, L, F and V, Fig. 9). Some bash scripts have been developed from the creation to the execution of all designs for automate purpose of the entire procedure. The scripts read each row of the DOE table, write all necessary files

Fig. 6 - Initial filling of the shot sleeve

Fig. 7 - Shot profile implemented for the first phase of injection

Fig. 8 - Two example graphs that show the distribution of experiments in the design space. The graph a) gives the idea about the constraints imposed on the geometric variables


The MUSIC Project at the International CAE Conference 2015At the 2015 edition of the International CAE Conference, the MUSIC Project had the chance to show the Beta version of the MUSIC “Control&Cognitive System” (C&CS) in the Agorà Research, thanks to the support of EnginSoft and the sponsorship of Assomet Servizi srl.The C&CS, developed by EnginSoft S.p.A., will activate a quality control and cost efficiency loop in the high pressure die casting of light alloys (HDPC) and plastic injection molding (PIM) industry by introducing for the very first time a holistic approach in real time data monitoring, analysis and control of all the phases of the currently fragmented automated production lines. The two days’ event of the CAE Conference provided the ideal context and venue to discuss the increasing relevance of “simulation based engineering and sciences”, research and innovation as opportunities to become “smart factories” as required by the current market challenges.The MUSIC project was also a protagonist in the successful “Foundry session” of the conference with a presentation concerning “Analytical compu-tation of the plunger kinematic parameters affecting quality in HPDC” .Further information about the project and the abstract of the paper are available at: http://music.eucoord.com/

MUSIC Tour at EUROGUSS 2016The Consortium of the MUSIC Project invites the visitors of the EUROGUSS International Trade Fair for Die Casting Technology, Processes and Products to visit its partners and get acquainted with their research activities in the High Pressure Die Casting and Plastic Injection Moulding sectors.During the three day exhibition, the MUSIC project will also deliver insights at the International Die Casting Congress, proving once more the rele-vance and the quality of the results achieved. For further information and to download the flyer of the event go to:https://music.eucoord.com/Euroguss/body.pe

alloy that must be equal in all cases. With this value, and knowing geometry (D and L), we can calculate the filling F. It is possible to demonstrate that there is a parabolic relationship between D and F considering L as a constant. Fig. 13 shows this approach to plot the response surface.

ConclusionsThe described work has permit to develop a predictive model for air entrapment during the first phase of injection in HPDC process.A new solver has been developed in the open source framework OpenFOAM™. The code is a customization of the solver “interDyMFoam” with adding features such as temperature field, heat exchange, temperature dependent aluminium alloy properties.A DOE has been created using the SOBOL algorithm (133 total designs) considering the inputs parameters (diameter and length of the shot sleeve, initial filling and first phase velocity), their maximum and minimum value and combinations between them that represent relevant cases in foundry practice.Each design has been simulated using the customized solver. A High Performance Computing (HPC) system has been necessary to run the 133 designs due to the high computational resources required for each simulation (a total of 64 processor have been used for 20 weeks of computational time).A predictive model has been created based on the simulation results. The function has been created using response surface methodology for which radial basis function algorithm has proved to provide the best interpolation fit to numerical data.In this way, a parametric shot sleeve and variable process parameters have been related to the entrapped air percentage value in condition of full chamber. The predictive model allows process engineers to choose the best combination of process parameters to avoid air entrapment during the first phase of injection without requiring high time consuming numerical simulations.

Roberto Meneghello, University of PadovaEnrico Boesso, EnginSoft

Fig. 11 - Explore 3D graph that reports R as a function of V and F parameters. The geometry variables have been imposed to D = 70 mm and L = 500 mm

Fig. 12 - Plot of R as a function of the plunger’s speed for D = 70 mm, L = 500 mm and F = 50 %. This plot is derived with the 3D graph of Figure 11

Fig. 13 - Surface response as a function of D and F (with constant value of L and V). The red line represent the conservation of volume law for a given volume of the melt alloy

MUSIC Survey on High Pressure Die Casting Market in EuropeThe MUSIC project – MUlti-layers control&cognitive System to drive metal and plastic production line for Injected Componen-ts – is about to launch a European survey aimed at collecting useful quantitative information to better understand the quality requirements in particular for the SME companies, widely re-presentative in the High Pressure Die Casting sector.

Collaborate with us and participate in the project survey: https://music.eucoord.com/Survey/body.pe


of performance degradation due to glaze and rime ice.FENSAP-ICE has been originally developed and is widely used by the aeronautical industry for in-flight icing certification, but its capabilities to predict ice accretion and to support the design of Ice Protection Systems is also of fundamental importance for the wind turbine industry.FENSAP-ICE accuracy has been validated against experimental data using the NREL phase VI two-blade wind turbine rotor, for which a comprehensive set of experimental data is publicly available.Four different icing scenarios have been simulated (Table 1), considering glaze and rime ice at moderate wind speed, glaze ice at high speed and glaze ice with large super-cooled droplets.Table 1 shows also the typical input weather data needed to perform icing simulations.

Figure 2 shows the capability of the software to predict ice coverage, thickness and shape on the suction and pressure side of a wind turbine blade in the four icing conditions.Ice coverage and thickness depend on several factors. In rime ice conditions the temperature is -15°C and the freezing process is very fast, and ice mainly forms at the leading edge, where droplets impinge on the blade. In glaze ice conditions the temperature is closer to 0°C, the freezing process is slower and water runback occurs with a more extended ice coverage, with different ice thicknesses, shape and density. In the third case with higher wind speed, the highly separated flow produces more ice coverage, while in case four large droplets produce increased coverage and ice mass.After calculating the ice shapes, the blade and rotor performances can be compared between the clean and iced conditions.Figure 3 compares the torque values for the clean blade against the contaminated blade using the same four scenarios in Figure 2. The torque reduction ranges from 10% up to 60%

dependent upon the wind speed and icing conditions, with glaze ice having a greater detriment to rime ice.

Design of ice protection systemsThe first step is to predict how droplet impingement affects ice build-up and the resulting performance degradation, but it is also helpful in designing Ice Protection Systems (IPS).The anti-icing power requirements for an IPS depends strongly on the icing scenario for which it is designed and is largely defined by two types of unknowns: the region of the blade to be heated (coverage) and the power distribution, expressed in specific power [W/m^2].FENSAP-ICE simulates the air flow field with droplet impingement and the surface water thermodynamics, including the effect of heat transfer from electro-thermal pads or hot air. This calculates the heating requirements to keep the blade surface above freezing and to evaporate the water film that forms on the blade due to water impingement.In this way FENSAP-ICE can be used to define and optimize the coverage and power distribution of the Ice Protection Systems, where the optimum is the right trade-off between coverage and specific power, so that the total energy consumption is minimized.Cases I and II presented in Table 1 have been use to evaluate the anti-icing requirements in glaze and rime ice conditions for the NREL phase VI two-blade wind turbine rotor. The specific power along the blade span is presented in Figure 4 and Figure 5 for rime and glaze. In both cases the maximum specific power has to be applied close to the tip, with values that are higher in rime conditions.

Table 1- Icing scenarios for the simulation of ice accretion and performance degradation

Fig. 3 - Torque and power loss versus wind speed in four icing scenarios

Fig. 2 - Ice coverage on a wind turbine blade in different icing conditions (left) – iced profile at 80% span (right)


Annual energy losses, ice detection, de-icing and anti-icing of wind turbines are becoming relevant topics for wind farms operators and for turbine manufactures as installation of wind turbines grows in cold climates.Therefore the use of Computational Fluid Dynamics to simulate and visualize ice accretion is of paramount importance for several reasons:

• The prediction of ice accretion in different meteorological scenarios can help to assess the performance degradation and production losses due to ice. This kind of calculation can be extended from one single turbine to an entire wind farm in order to predict annual energy losses.

• The use of virtual prototyping can support the design and optimization of anti-icing and de-icing systems based on electro-thermal heating or hot air. The design of these systems has to achieve conflicting objectives: the main target is to prevent or remove ice from the blade, but it has to be done with the minimum energy consumption, in a uniform way and, in case of de-icing, with the minimum time.

• Understanding and visualizing when and where ice accretes on the turbines and calculating its shape, thickness and mass is vital for the design of a reliable ice detection systems. For example; the comparison of vibration modes between clean and iced blades can be used to design and tune ice detection systems that are based on these kind of measurements.

• Increased loading due to ice and how ice affects the structural behaviour and life of the turbine can be examined coupling icing simulation with Finite Element Analysis.

Finally CFD can also be used to understand the impact of ice in terms of safety, associated to ice throw, and in terms in increased noise.This article explains how Computational Fluid Dynamics and the simulation of ice accretion can be used to predict the performance degradation of a wind turbine in icing conditions, to design an anti-icing system and to evaluate the costs and benefits of an Ice Protection Systems.

Prediction of performance degradationIce accretion on wind turbines occurs when super-cooled water droplets within clouds impinge on the surface of the blades. The mechanisms of ice accretion are usually divided into two distinct categories depending on the atmospheric conditions. Rime ice occurs when the impinging water freezes instantly on impact with the blades, usually when the temperature is far below the freezing point. Glaze ice occurs in warmer conditions, when the impinging droplets partially freeze on contact while the remaining water gradually freezes while running down the surfaces of the blades.

Since the surface characteristics and shape of the accreting ice is very different in the two cases, glaze and rime ice can have very different effects on the performance of a contaminated wind turbine rotor.FENSAP-ICE, a state-of-the-art simulation tool combining Computational Fluid Dynamics with 3D ice accretion have been used in recent years by wind turbine manufactures for the prediction of ice accretion on wind turbines and for the assessment

Wind energy in cold climates

Fig. 1 - 3D ice pattern on a wind turbine blade, left: photograph of a wind turbine in North Sweden, right: simulation using FENSAP-ICE

Prediction of energy losses and Ice Protection Systems

Sezione49 - Newsletter EnginSoft Year 12 n°3 Case Histories Newsletter EnginSoft Year 12 n°4 - 48

significantly reduce the annual energy production. In turn, the relevant impact of icing events on wind farms profitability is driving the development of anti-icing and de-icing systems. Therefore the use of simulation tools is of paramount importance in calculating the impact of ice on turbine performance, to estimate the annual energy losses, to design and optimize Ice Protection Systems and finally to reduce investment risks.

EnginSoft have applied Computational Fluid Dynamics and ice simulation to real industrial cases, in particular to the design of anti-icing and de-icing systems and to the assessment of performance degradation of wind turbines in icing conditions. In some of these cases the simulation results have been validated versus experimental data from icing wind tunnels.

For more information, please contact: Susanne Glifberg, [email protected]

Or visit the website:http://www.enginsoft.it/applications/aerospace/deicing.html

Fig. 7 - comparison of production benefit and cost of IPS

Annually carhs.training publishes the SafetyCompanion. On more than 170 pages, it provides current and selected expert knowledge in automotive safety that supports engineers in their daily work. This includes tables and graphs summarizing and illustrating rules, regulations, crash test configurations and requirements in automotive safety.carhs‘ SafetyCompanion is published in three languages (English, German and Chinese) and reaches experts for automotive safety worldwide. In addition to 16,500 hardcopies, the pdf version is downloaded about 5,000 times a year. For many engineers the SafetyCompanion is the reference book at hand throughout the year.The SafetyCompanion gives companies that work in the field of active and passive safety the opportunity to place advertisements that will directly reach industry experts and decision makers. The advertisement deadline for the 2016 edition is on October 6, 2015. Advertisement format and rates are available here: https://www.carhs.de/en/media-data.html?file=files/carhs/data/media/Advertisement_SafetyCompanion_2016.pdf The SafetyCompanion can be ordered here: https://www.carhs.de/en/companion-poster/product/safetycompanion-2015-pdf.html

Safety Companion the Reference Book for Vehicle Safety Experts

Instead, Figure 5 shows the coverage of the anti-icing system on pressure, suction and leading edge of the blade. Again the maximum coverage is at the tip, where the droplet collection is higher due to the rotational effects. Different colors in Figure 5 show that the heated area can be reduced by increasing the specific power, so that water evaporation is faster.When the weather conditions are so extreme that ice prevention is not possible, wind turbines are kept in an inoperative state and de-icing is needed before re-starting the turbine. In this case, the use of simulation can support the design of hot air de-icing systems, with the aim of optimizing the hot air distribution, pressure drops, heat fluxes and the temperature uniformity on the blade. Moreover the time needed to bring the blade surface above the freezing point and to de-ice the blade can be evaluated by running a transient CFD simulation.

The value of using virtual prototyping is its ability to compare the performances and costs of different anti-icing and de-icing systems in real life conditions. This allows decisions to be made between anti-icing and de-icing or between different Ice Protection solutions, designs and materials.Simulation allows the assessment of how the IPS performance is affected by weather conditions (temperature, wind speed, liquid

water content in air) or how the de-icing efficiency is affected by factors like altitude.

Prediction of losses in long icing eventsThe results of FENSAP-ICE simulations can be scaled up from a single turbine to a whole wind farm to assess the production losses due to ice and the cost-effectiveness of Ice Protection Systems. Complex, historical weather events can be simulated on industrial-scale wind turbines. Figure 6 shows the meteorological data of a 17 hours icing event at a wind farm in the Gaspé Peninsula, Québec, Canada. The weather data has been used to simulate the event in FENSAP-ICE and to predict the production losses associated to this event.At the same time, the heating requirements to prevent ice build-up have also been calculated for this specific

event. The benefit of an Ice Protection System is realized in the production improvement, due to the fact that no turbine downtime is needed and that the performance does not deteriorate during the icing event. The cost of operating the Ice Protection System has been compared to the production improvement. The comparison is shown in Figure 7, where the green curve represents the Net Power Gain, with a positive integral value, meaning that in this specific event an anti-icing system can prevent ice formation with a positive balance between energy consumption and increased energy production.

ConclusionsThe increasing demand for renewable energy is driving the increased need for wind farm development in cold climate regions, characterized by frequent atmospheric icing events, which can

Fig. 5 - Ice Protection System coverage for a turbine blade in two icing scenarios

Fig. 4 - Heating requirements for a turbine blade to prevent ice formation in two icing scenarios

Fig. 6 - 17 hours icing event at a wind farm in the Gaspé Peninsula, Québec, Canada

Sezione51 - Newsletter EnginSoft Year 12 n°3 Software Update Newsletter EnginSoft Year 12 n°4 - 50

MapleSoft, leading provider of high-performance solutions for engineering, science and mathematics, and EnginSoft, signed a comprehensive agreement covering the distribution and support in Europe of all the MapleSoft products, including Maple, MapleSim, MapleNet and the related add-ons, connectors and libraries.Thanks to this landmark agreement, EnginSoft adds to its enterprise-level solutions portfolio the modeling and simulation capabilities of the MapleSoft products suite.

MapleMaple is the world leader in numeric and symbolic computation solutions: used in Fortune 500 companies all over the world to analyze, explore, visualize, and solve engineering math and computation problems in a single environment. Maple is used by Design Engineers and Advanced Analysts to quickly and accurately perform calculations and mathematical manipulations using live mathematical expressions. Its intuitive

interface allows the direct use of live mathematical expressions, turning mathematical concepts into solutions through powerful symbolic processing and advanced numeric computation.Maple enables the creation of intuitive user interfaces using ready-made UI components that directly integrate with the underlying algorithms and routines. These Functional Mock-Ups capture the technical knowledge behind the solution through fully-disclosed equations and documentation capabilities.

MapleSimMapleSim is the advanced system-level modeling solution based on the Maple mathematical engine and analysis environment to design and simulate multidomain systems, plants and controls in one single environment. The schematic diagram interface enables rapid model development, thanks to the Maple symbolic computation foundation providing a highly concise, numerically efficient model formulation.MapleSim enables Engineering System Designers to answer the most critical questions faced in the everyday activity “Will the system work as required?”, and “Can we make it work better?”, by developing virtual prototypes of multi-domain dynamic systems and performing functional validation, critical system-level parametric studies and optimizations. The Control System Developer will be able to leverage MapleSim to develop high-fidelity models of complex dynamic systems for plant

EnginSoft to distribute and support MapleSoft products in 2016

Fig. 1 - Simulation running in Maple

Fig. 2 - MapleSim capabilities allows for fast and accurate system modeling and simulation

characterization and integrating models through automated code generation into (typically) MATLAB®/Simulink® for controller development.

MapleNet and MapleSim ServerBoth MapleNet and MapleSim Server allowing for easy deployment and sharing of Maple documents and MapleSim simulation models. Using MapleNet, Engineering solutions can be deployed through web-publication and automatic code generation (C, Java, VB, FORTRAN…), allowing engineers and users to interact with content, perform changes, and visualize results all from within a standard web browser and with no need for a local installation of Maple or MapleSim.This approach will easily capture technical knowledge and design documentation across companies, via an intuitive interface that supports multiple styles of interaction.

Endless possibilitiesMapleSoft products suite are the reference for Physical Modeling and Simulation, Real-Time Simulation and Hardware-in-the-Loop Applications, Engineering Calculations and Design Documentation, Optimization and Analysis , Functional Mock-Up Development, Deployment and Connectivity, and much more.Current MapleSoft customers include automotive OEMs and Tier1, Automation and Machine Design companies, Aerospace companies, and major Consumer Electronics Manufacturers, etc.EnginSoft is excited by the new opportunities brought by the agreement with MapleSoft, and is looking forward fruitful collaboration with current and new Maplesoft product Customers.

For more information: [email protected]

Fig. 3 - MapleNet allows to easily capture and share design knowledge via the web

Fig. 4 - Maplesoft products overview

ANSYS AIM: Integrated Multiphysics Simulation Environment for All Engineers

ANSYS 16.0 delivered the first integrated and comprehensive multiphysics simulation platform designed for all engineers. This platform, called AIM, has derived from the ANSYS Workbench strategy to revolutionize engineering processes through Simulation Driven Product Development.

AIM extends the value of simulation beyond single engineering disciplines by providing a full array of physics in a unified, immersive user platform that is readily deployable across the organization. Whether a multiphysics simulation includes structures, fluids, thermal properties, or electromagnetics, all aspects of simulation workflow are supported in the single-window design, reducing the training and deployment costs typically associated with fragmented tool chains. Moreover, AIM’s all new simulation process paradigm guides engineers through multiphysics workflows that consider the interaction of physical phenomena, while automating tedious tasks to free-up engineering time and resources.

View the EnginSoft webinar and a demo live on AIM, visit the page:http://www.enginsoft.it/form/downloadwebinar.html

Research & Technology Transfer53 - Newsletter EnginSoft Year 12 n°4

all the 13 original uncertainties. For this purpose we have evaluated different series of designs, applying a Polynomial Chaos Expansion of order 2 for the UQ, keeping originally all the polynomial terms (105).Applying the Adaptive Sparse methodology described in section 2, we have found that only 11 terms are needed to give acceptable errors on a database, which can then be reduced to 30 samples only. Table 2 below therefore reports the UQ results for the test case, after the application of Polynomial Chaos expansion (in modeFRONTIER software) of order 2 for the different DOEs analyzed, and in particular: 1) 1000 samples with all the uncertainties, 2) 105 samples with all the uncertainties, 3) 30 samples with 11 most important PCE terms. The results of the test are satisfactory. Applying the Adaptive methodology, the results are improved, since by a lower number of samples, 30, the highest error on standard deviation has been reduced in the worst case to 2%. In general, it emerges clearly that a significant reduction of the number of needed samples, reachable by any of the two methodologies, produces an accurate evaluation of the statistical moments, with a contained maximum estimated error.

RDO: classical vs Min-max/Reliability approachThe first RDO approach introduced as state of the art, is the classical approach based on the definition of a multi-objective optimization problem, consisting generally on the optimization of the mean value of the performances and on the minimization of their standard deviation.This approach guarantees the definition of a complete Pareto frontier as trade-off of the optimal solutions, in terms of mean performance and in terms of their stability or robustness. The designer has then the freedom to select the best solutions accordingly to a large variety of possibilities depending on which criteria should be privileged.The problem of this approach is that a Multi-objective Optimization algorithm is to be chosen, since the definition of a single objective as weighted sum of the different criteria cannot be proposed for the impossibility of knowing a priori the proper weights of the particular optimization problem. Multi-objective Optimization algorithms are in fact generally very robust, but they require a number of simulations generally very much consistent with respect to a single objective optimization case, and for a RDO problem the number of simulations may be not feasible from a practical point of view (this number being multiplied by the sampling size for each design to obtain the overall number of simulations required).In order to reduce the overall number of simulations for a RDO problem, we then propose in this paper another approach. The basic idea is to reduce the number of objectives, so that a single-objective algorithm, which requires much less simulations for the convergence, could be applied.

To achieve this purpose, the so called min-max or max-min approach is followed: The idea is to maximize the minimum or worst performance of a distribution function that is to be maximized (for instance the aerodynamic efficiency of a wing), or to minimize the maximum or worst limitation that is to be minimized (for instance the drag coefficient of a wing).In the case of a Normal distribution of the performance, since it is unlimited, the concept of the extremes may be replaced by a given percentile of the distribution, for instance 95 or 99%. Usually, the reference value is 99.73% because for a Normal distribution it corresponds to the 3Sigma level (Six-Sigma analysis). A limitation of the proposed RDO based on Six-Sigma occurs when the performance does not follow a Normal distribution: in this case, the Six-Sigma formulation may not correspond exactly to the correct percentile of the distributions, and then a Reliability formulation problem is necessary. Many methodologies exist in literature to solve Reliability problems, such as FORM/SORM, which for a RDO optimization could be very expansive from the numerical point of view. For this reason, we propose here a different methodology, based indeed on the Polynomial Chaos polynomial exploitation. In fact, the evaluation of the performance function in industrial cases can be very demanding, since they often involve expensive CFD or structural numerical simulations. In the approach we propose, these expensive evaluations are required only to determine the coefficients of the PCE (Polynomial Chaos expansion). Once found them, it is possible to express the CDF (cumulative distribution function) of any system response using directly the PCE polynomial, which can be considered as a meta-model of the response, practically free in terms of CPU. Once the CDF is accurately obtained, we can easily retrieve the value corresponding to the needed percentage of the distribution, to define objectives or constraints (failure probability).

RDO test case applicationTo validate and compare the efficiency of the methodologies proposed in the previous chapter, we have applied those to test case derived from the one used to describe the State of the Art techniques of UQ in section 3. For simplicity, we consider only 3 uncertainties for the RAE2822 airfoil, with nominal values equal to 0.734 for free stream Mach number, 2.79° for angle of attach and Reynolds number equal to 6.5E6. The uncertainties are given by a Normal distribution for Thickness-to-chord profile (a single uncertainty factor which multiplies the thickness profile),

Fig. 1 - Workflow for process automation in modeFRONTIER

Fig. 2 - UQ results for test case

Research & Technology Transfer Newsletter EnginSoft Year 12 n°4 - 52

One of the most efficient methodologies in managing uncertainties in industrial design is the application of Polynomial Chaos expansion. However, this methodology requires a minimum number of samples which increases heavily with the number of uncertainties, making a typical industrial optimization case (for instance at least 10 simultaneous uncertainties) an unfeasible task. For this reason, this paper proposes some approaches to handle industrial problems of this kind efficiently.From the Uncertainty Quantification (UQ) side, the proposed solution is to use a methodology that identifies the Polynomial Chaos terms that have the highest statistical effects on the system performance, reduce the number of unknown coefficients and therefore the number of samples needed to complete the UQ, without discarding any stochastic variables from the problem. From the Robust Design Optimization (RDO) side, the proposed solution is to use a methodology based on the min-max formulation of objectives, which guarantees the reduction of objectives with respect to a classical RDO approach, and therefore reducing the number of simulations to be performed drastically. In order to guarantee an accurate application of this methodology, an approach was developed based on the exploitation of Polynomial Chaos coefficients to evaluate the percentiles of the quantities to be optimized/constrained. This methodology is also called reliability-based design optimization, and the solution we propose, based on Polynomial Chaos exploitation, is innovative and very promising in terms of efficiency.

UQ of large number of variables: Adaptive Sparse CollocationAn efficient method to apply UQ with a large number of variables, is based on the application of a regression analysis directly on the Polynomial Chaos expansion (PCE) expression. In other words, the PCE will keep only those terms which actually affect the output, discarding the others.First, the methodology ranks the terms using a Least Angle Regression (LAR) technique and then assesses how many PCE terms should be kept.The order of selection of the PCE terms will reflect a ranking based on how much each term affects the output. Once the ranking is done, it is necessary to establish a way of choosing how many PCE terms should be kept. The criterion for this is based on the Mean Leave one Out Error(ErrLOO).

where N is the number of samples and i=output(xi)-M(xi) is the difference between the output corresponding to the i-th sample and the output computed from the PCE obtained excluding from the training samples the i-th. The criterion to select the number of terms consists in monitoring the quantity:

RLOO is a function of the number of PCE terms: it initially increases as the

number of terms increase, but from a certain number of terms on, it starts showing a decreasing trend. This index is in fact sensitive to overfitting problems, so that an optimal number of terms is generally much lower than the maximum possible number equal to the original degree of freedom of the complete polynomial.The described approach gives the important benefit of reducing the global number of unknown coefficients for the PCE expansion, and therefore giving the possibility to reduce the number of sampling points needed for the PCE training.

UQ Test Case applicationTo validate the methodology proposed in previous section, we have applied it to the test case BC-02 of UMRIDA European Project. The test case consists in the UQ quantification of a RAE 2822 airfoil, for a specified conditions, and for a total of 13 uncertainties (operational and geometrical). Nominal parameters are given by Mach number=0.734, Re= 6.5 106 and angle of attack=2.54°, with uncertainties given by a Symmetric Beta Pdf with a range from 4 to 10% of the baseline values. In particular, the geometrical uncertainties refer to the camber line and the thickness-to-chord ratio of the nominal profile, which have been fitted by a Bezier parametric curve, of respectively 7 and 8 control points, and also Mach and angle of attack are defined as uncertain parameters.Figure 1 reports the process workflow created for the set-up of this test case in modeFRONTIER software from ESTECO. Each component of the process is defined by dedicated modules (input variables, CAD/CAE interfaces, output variables) inter-connected between them, in order to allow the automatic execution of the simulations for each design sample which is proposed by the selected algorithm for the UQ. modeFRONTIER software contains as well all the tools needed to complete automatically the UQ of the required parameters.The mesh provided for this test case has been elaborated in FINE/Open software from NUMECA. The mesh is characterized by an overall number of cells equal to about 1/2 million, which require an average time to complete the simulation of one design sample in about 1 hr, using a 2-cpu machine. Around the airfoil the mesh is adaptively refined, and a full turbulent (Spalart-Allmaras) model is used.The first step in the UQ of the test case is the definition of a large DOE (Design of Experiments) using a Latin Hypercube algorithm, considering

Innovative methodologies for Robust Design Optimization with large number of uncertainties using modeFRONTIER

Events55 - Newsletter EnginSoft Year 12 n°4 Research & Technology Transfer Newsletter EnginSoft Year 12 n°4 - 54

Mach number and angle of attack, defined by a standard deviation respectively equal to 0.005, 0.005 and 0.1. The target is to perform a RDO minimizing the drag coefficient, keeping lift and momentum coefficients not worse than the baseline configuration.The RDO problem may then be defined in three different ways following the different approaches described in section 4, and they can be resumed in table 3. The classical optimization strategy (multi-objective approach), considers the minimization of the mean Cd value and the minimization of its standard deviation, adding two constraints on the worst values of Cl and Cm, here approximated considering a Normal distribution following the Six-Sigma rule.The second methodology follows the max-min approach, allowing the definition of a single-objective problem, which is the minimization of a given high percentile (still 99.97) of the Cd distribution, that can be considered as the “maximum” target value, computed again following the Six-Sigma rule. Following the third approach, we can compute accurately the needed percentile distribution (99.97%) directly from the CDF distribution function defined by the Polynomial Chaos expansion. The definition of the constraints is therefore slightly different from the Six-Sigma approach (table 3). A number of 10 sampling points for design was found to be necessary to guarantee an accurate UQ using a Polynomial Chaos expansion of the second order. To solve the multi-objective approach, we applied a classical GA algorithm, but after the evaluation of more than 50 designs (for a total of 500 CFD simulations, that corresponds to about 20 days using a double Cpu machine), it was practically impossible to find feasible solutions that improve the original baseline. The optimization approach has been stopped, because the optimization time was considered already excessive for a problem of industrial relevance.To solve the single-objective optimization efficiently we have applied a Simplex algorithm instead, with a global number of simulations not higher than the one considered for the multi-objective case. Fig.4 below reports the results obtained with both approaches (Six-Sigma and PCE-Reliability), reporting the convergence of the objective function (maximum value of Cd). The yellow points are unfeasible designs, i.e. they don’t respect the constraints, while the blue ones are feasible. The convergence trend is clear, in particular regarding the reliability approach. In order to compare the results, in Figure 5 below we report a comparison of the performances (mean and standard deviation values) of the baseline configuration and of the optimized configuration obtained in each approach (including multi-objective approach, that as explained was not able to find feasible designs by this number of evaluations). It can be noted that, even though both the two single-objective approaches are able to find optimal designs which respect the constraints, the reliability approach based on

the application of PCE is able to find better results, in particular with an improvement of about 20% in the reduction of drag coefficient.This difference may be probably explained by the fact that the first approach (Six-Sigma) follows an hypothesis which in this case is not completely correct (the performance distribution to follow a Normal distribution) and therefore only by the accurate estimation of extreme percentiles of the PCE approach we can evaluate correctly the design objectives/constraints, obtaining a better solution.

ConclusionsWe have illustrated some innovative methodologies for Robust Design Optimizations with a large number of uncertainties, which is a typical requirement from the industry. An efficient UQ methodology has been proposed, based on a regression methodology, which can be used to reduce the number of sampling points for an accurate uncertainty quantification (reducing the number of Polynomial Chaos terms).In addition, a methodology for efficient Robust Design Optimization (based on the application of min-max criteria combined with reliability-based optimization formulation and Polynomial Chaos exploitation for percentiles estimation) has been presented. All the methodologies have been validated by the application of selected test cases in the aeronautical field; in the future steps of UMRIDA Project, the proposed methodologies will be applied to industrial problems of challenging relevance.

Alberto Clarich, Rosario Russo - ESTECO

AcknowledgementsThis UMRIDA has project received funding from the European Union’s Seventh Framework Programme for research, technological development and demonstration under grant agreement no ACP3-GA-2013-605036 (website www.umrida.eu).

For more information:Francesco Franchini, EnginSoft - [email protected]

Fig. 3 - RDO definition of test case accordingly to three different approaches

Fig. 4 - Optimization results using min-max approach (Six-Sigma left – PCE-Reliability right)

Fig. 5 - Results of Six-Sigma based and RBDO method

Mentor Graphics Corporation announced the winners of its first annual Don Miller Award for Excellence in System-Level Thermo-Fluid Design. Mentor Graphics established the award to recognize exceptional use of Flowmaster software in research and real-world applications. The inaugural award was presented to Andrea Tradii, Stefano Rossin and Riccardo De Paolis of GE Oil and Gas on October 19th at the 31st International CAE Conference in Verona, Italy. Their work, “Experimental Validation of Steam Turbine Control Oil Actuation Systems Transient Behavior,” was presented at the same event last year and was nominated for the inaugural award this year.

Tradii, Rossin and De Paolis’ work dramatically demonstrates the value of simulating the multi-physics of the fluid-mechanical interactions of a steam turbine trip valve using the Mentor Graphics Flowmaster tool. With nearly 40 years of combined experience at GE Oil and Gas and the turbomachinery industry, the authors demonstrated creativity and innovative thinking in the application of the Flowmaster software to solve a significant engineering challenge.Two runners-up were also selected by the panel of expert judges, including Don Miller. The team of Morten Kjeldsen from Flow Design Bureau AS, and Christoffer Jarpner from EnginSoft Nordic received a runner-up award for their work at Salt Ship Design SX.

They investigated use of 1D and 3D CFD as complementary methods of simulation and optimized piping systems onboard ships with the aid of Flowmaster software and other CAE tools. The team of Roberto Conti, Emanuele Galardi, Enrico Meli, Daniele Nocciolini, Luca Pugi, Andrea Rindi of Florence University, with Dr. Stefano Rossin and Riccardo De Paolis of General Electric Nuovo Pignone S.p.A., was recognized for their innovative work on understanding the interactions between the fluid, mechanical control and control logic of a steam turbine control valve.

“I am deeply honored to have an award in my name, and to see the tremendous body of work executed by these outstanding engineers,” stated Don Miller, former research director for British Hydromechanics Research (now BHR Group) in the U.K. “I had no idea the capabilities Flowmaster would provide users with when I

sought venture capital funding 35 years ago. The prize-winning entry clearly demonstrates that from a core of validated data, not only can the thermal/fluid flow performance of complete systems be simulated, but also the functioning of complex individual components that carry out vital control and safety functions.” Miller is also the author of Internal Flow Systems, whose book serves as the foundation for Mentor Graphics Flowmaster software technology.

For more information on the Don Miller Award, and to review the winning submissions for the 2015 awards, visit: http://www.mentor.com/products/mechanical/awards-publications/don-miller

Mentor Graphics Announces Winners of Don Miller Award for Thermo-Fluid Design Excellence during the International CAE Conference 2015

Events Newsletter EnginSoft Year 12 n°4 - 56

Innovative perspectives along with avant-garde science and technologies are once again protagonists at the International CAE Conference, the most significant European event dedicated to computer simulation and its countless application possibilities continues to progress the unstoppable steps towards the future.

This year marks the thirty-first of the event which has become an internationally established reference point; held on the 19th – 20th October 2015 in Lazise at the Congress Centre Hotel Parchi del Garda. The event provided the engineering community with the platform to share the latest achievements and experiences amongst their peers and diverse span of industry professionals, researchers and technicians; whom are involved in various aspects of product and process innovation, including all application fields: Manufacturing, Aerospace & Defense, Oil&Gas, Energy, Medical, Automotive, Transportation, Civil Engineering etc..

Participation: the numbers speak for itselfMore than 800 delegates participated at this year’s event, with 150 speakers alternating between the two plenary sessions and 10 industrial focused sessions, where over 35 companies sponsors were also present. There was no lack of attendance from Universities, Faculties and Polytechnic Institutes; represented by professors, researchers and students, which in the development of the numerical simulation epitomising the best collective minds, Europeans and overseas, but also Italians. It can’t be forgotten, indeed, that the Country plays a crucial role of excellence in the industry and can advocate a leading role in the landscape of technological innovations.

The prestigious setting was complete with the high calibre of guests, who shared their progress and future advances from pioneering projects demonstrated in their own fields. As highlighted by the founder and President of the Scientific Committee of the International CAE Conference, Stefano Odorizzi, “The International CAE Conference is a rich environment to discover current developments and new technological breakthroughs, network, explore new business opportunities and share experiences”.

In connection with the Expo, the International CAE Conference has had the pleasure of hosting prestigious guests such as two of the deus ex machina: Alessandro Gasparini, engineer and triumphant in the race against time for the construction of the sculpture – a symbolic installation at the Expo, the “Tree of life”; along with Massimo Maffeis, designer of tensile structures, who created at the international solutions for both the major walkway (cardo and decumano), between the Mexican, Kwait and German pavilions.

INTERNATIONAL CAE CONFERENCEYour opportunity to be part of the future19th – 20th October 2015, Lazise (Congress Centre Hotel Parchi del Garda)

Events57 - Newsletter EnginSoft Year 12 n°4

In particular the presence of Don Miller is of significance, at 86 years old, he is among one of the predecessors for the use of computer simulation through fluid dynamic processes. The English professor who is considered a visionary for thirty years, verifies the excellence his knowledge and he was chosen to award Andrea Tradii and Riccardo De Paolis of General Electric – Nuovo Pignone – Florence, for the innovative approach applied to a turbine project.

Feed the planet: challenges for the future or challenges for the present?The future is imagined today, but today we also need to prepare for the future.As we strive to improve there is a natural theme that links the International CAE Conference to the Expo2015in Milan: “Feed the planet”. This can be done – explained CHRISTOPHIE LASSEUR, coordinator of the ESA(European Space Agency)Life support R&D– which also applies on Earth; the techniques for the reproduction of oxygen, water and raw materials (including edible crops) in an hostile environment, is currently under study with the view of landing on Mars. This is the base concept of the Melissa project, of which Lasser is leading ,it takes inspiration from the anaerobic principles typical of aquatic ecosystems.

Nutrition and food were also at the centre of the event. Professor Davide Cassi, physicist and founder of the Faculty of Physics Gastronomy at the University of Parma, the only Italian institution of its kind. Cassi retraced the steps of the molecular cuisine, explaining some of the newest ways of preparing food through scientific techniques: the use of alcohol or liquid nitrogen, for example, as tools of transforming the gastronomic matter with the capability of maintaining the components and the aromas. Professor Cassi amazed the audience, creating recipes of the tradition cuisine live in a scientific way.

Life quality, the answers from technologyRobots that take care of people, with the ability to assist an elderly: for example if he forgets to take the medicines, the robot will remind him, or if it detects something abnormal, a fall or they collapse, it

can call the ambulance or alert the relatives. In addition, small robots are able to interact with autistic children, to accompany them through their learning. They are no longer just science fiction, but all the models are in the testing phase at the IAS Lab (Intelligent Autonomous Systems), in the Robotic Laboratory at the University of Padua. During the International CAE Conference their designers showed the potential and operation controlled by brain impulses.

Speed is a matter of numbersIn the exhibition area, among the others achievements showcased, the attendees could admire PULSAR, a high speed bicycle (HPV – Human Powered Vehicle) specifically designed by the Turin Polytechnic to compete at the WHPSC (World Human Powered Speed Challenge), an event that takes place every year in Nevada. The prototype was built in 2015 and took part to the WHPSC in September, after about 850 Km of tests. Andrea Gallo, present with Pulsar at International CAE Conference, reached the Italian speed record with 116.19 kilometers per hour, becoming the 24th fastest pilot in the history.

The research Agorà, the science finds spaceIlluminate the present with the ideas of the future. In the Research Agorà, displayed the most innovative projects, developed through key international partnerships. In the dissemination throughout the scientific sessions, the attendees not only had the opportunity to presented their work, but they also to go on the stage to conduct live experiments and demonstrations of their projects.

Poster Awards: recognizing the future of engineering each year, in fact five…The 2015 edition continued to provide the creative foundation to next generation with the Poster Awards, the contest exclusively open to the undergraduates and PhD students, which awards the best 5 Posters selected by the Examining committee. This year there were 44 distinguished finalists from more than 200 Poster entries received.

www.caeconference.com

Fig. 1 - Experiment on molecular couisine during the plenary session by Prof. Davide Cassi from Parma University

CAE Poster Award Newsletter EnginSoft Year 12 n°4 - 58

The fourth edition of the “International Poster Award: a poster for CAE” has concluded with the winners ceremony in Pacengo, Garda Lake, on the 19th October at the end of the first day of the International CAE Conference, the annual event on “Simulation Based Engineering and Sciences”, a perfect marriage with this event. The contest was once again sponsored by EnginSoft who strives to constantly disseminate the scientific culture and to create opportunities for the next generation of talent. This year ESTECO co-sponsored, the contest. Their participation was aligned with the recent launch of the ESTECO Academy project that aims to promote the use of the optimization technologies in Universities and Research Centers. A Scientific Committee was assigned with the difficult task of evaluating the posters. The committee consisted of university professors from various academic research backgrounds: Bruno Atzori (University of Padova - Italy), Roberto Battiti (University of Trento - Italy), Beatrice Belletti (University of Parma - Italy), Marco Evangelos Biancolini (University of Rome “Tor Vergata” - Italy), Gabriele Angelo Dubini (Politecnico di Milano - Italy), Giuseppe Gambolati (University of Padova - Italy), Michael Gasik (University of Aalto - Finland), Gianluca Iaccarino (University of Stanford - USA), Clovis Maliska (University of Santa Catarina - Brazil), Dubrawka Mijuca (University of Belgrade - Serbia), Enrico Nobile (University of Trieste - Italy), Stefano Odorizzi (University of Padova - Italy), Bernardo Schrefler (University of Padova - Italy) and Giorgio Zavarise (University of Salento - Italy). The poster contest was also opened to registered users on the event website, who were able to add their votes to the Scientific Committee in an online pool.

The final 44 Poster were selected among the pool of posters submitted this year, which can still be viewed on the Poster Award section on the CAE Conference website (www.caeconference.com).The 5 winning posters showcasing their projects are presented in the following pages.The award ceremony was presented by the famous Radio host Luca Viscardi and the winners were awarded by special guests: Stefano Odorizzi, president of EnginSoft and professor at the University of Padova and Enrico Nobile, founder of ESTECO and professor at the University of Trieste. Together with the Scientific Committee, they were both very satisfied from the outstanding level and the variation of contest posters.

For more information please refer to the Poster Award section on: www.caeconference.com, or contact: [email protected]

INTERNATIONAL CAE POSTER AWARD 2015: a great success for the fourth yearAn important showcase for the next generation and for their innovative ideas using CAE technologies

CAE Poster Award59 - Newsletter EnginSoft Year 12 n°4

High Performing Free-Form Design and Material Optimization for Additive Layer ManufacturingA.N. Albini, S. Micheletti, S. Perotto, L. Soli, D.A. Tobia MOX,Politecnico di Milano & Thales Alenia Space

FREE FORM Two optimization strategies:

The reference physical-mathematical model is based on the linear elasticity system:

Topological optimization leads to solve the constrained optimization problem for the compliance /:

Topological optimization enhances the performances:

Free-form design for ALM From 3D model to printing

Unexpected forms are automatically designed: Saving mass and cost by ALM design:

Be stiff but save the mass:

Minimizing material waste

FLY TO BUY OPTIMIZATION

ALM makes your dreams come true!!

BENDING STRAIGHT TRUSSES

Advanced mathematical techniques yield structures and forms that classical industrial processes are not able to design

CAE

POST

ER A

WAR

D 20

15: W

INNE

R


Numerical analysis of blasting: Explosion modeling, FEM code validation and application in explosive metal formingBackgroundFrom the theoretical viewpoint, the explosion is an unexpected and aggressive emission of mechanical, chemical or nuclear energy, usually with production of high-temperature and high-pressure gases.Such gaes spread in the surrounding environment as shock wave, which in the absence of obstacles, expands like a spherical surface centered in the explosion point. The blast wave is followed by a blast wind of negative pressure, which sucks items back in towards the center. Considering the variation of pressure in time, fixed a point in the space, it change with an exponential law achieving two load stage: the first one is positive duo to overpressure while the second one is negative duo to the depression caused by the explosion winds (Figure 1).

Aim of the studyRealization of a FEM model that describe the evolution of pressure and the behavior of the structures under the effects of dynamics loads with high intensity and short duration, which are those produced by the explosions.

Numerical Simulation of Blast and Model ValidationLS.Dyna SIMULATION BLAST METHODSA-Load BlastThis method is based on the CONWEP function and allows the user to simulate bursts using an analytical formulation which depends on the distance from the center of the burst and the amount of the explosive used. The algorithm is based on the equivalent TNT method, indeed, several kind of explosive can be simulated by using an equivalent amount of TNT and appropriate conversion factors. This method allows the simulation of different kind of bursts: - Hemispherical surface burst- Spherical air burst - Air burst with ground reflection - Air burst moving non-spherical warhead

The convenience of this method is its low computational cost, but it does not take into account the reflections due to possible presence of objects situated between the explosive charge and the target. The presented numeric method has been utilized for the simulation of a free-air-spherical-burst: the results have been validated by the comparison with experimental data taken from Kingery and Bulmash (Figure 2).

B-Multi material ALE (Arbitrary Lagrangian Eulerian) The name of the method, Multi-Material, derives from the fact that different domains are involved, respectively for the modeling of the explosive and those of the shock wave propagation medium. It is based on the numerical solution ALE in which one mesh is a Lagrange-type-mesh and the other one is an Euler-type-mesh. One domain is created

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Fig. 1 - Typical trend of the overpressure

Fig. 2 - LBE overpressure valitation

Fig. 3 - MM_ALE overpressure valitation

yp

Giacomo Bolla Politecnico di Torino


for representing the shock wave propagation phenomenon and it is modeled according to the material of the burst event environment; the second material, which is the Lagrange-type one, is used for simulating the explosive domain. Both domains do have an assigned material which allows the modeling of the kind of explosive and those of the shock wave propagation space. Furthermore, a pressure propagation equation is used to characterize the shock wave propagation environment. This method is appropriate for simulating enclosed bursts, ground-level-bursts and free-air-spherical-bursts taking into account the interactions (the reflections) with secondary objects placed between the explosive charge and the simulation main object. The same free-air-spherical-burst simulation has been conducted using this technique and also in this case the results successfully validated the Load_Blast data-method (Figure 3).

C-Mixed Method LBE-ALE As a last validation test, a mixed method has been adopted. It has been developed by considering the two previously described simulation techniques: the LBE and MM_ALE methods. The idea is to simulate flat shock waves which derive from being at long distances from the detonation center. In this way the computation gets faster if compared to a pure ALE method simulation and this allows employing a smaller number of finite elements within the numeric method. In the domain which is not modeled by using finite elements, the shock wave is simulated with the LBE method, while for the modeled domain, the ALE_MULTI_MATERIAL method is employed. To simulate the air flow moved by the shock wave within the ALE domain, some receptor solid elements have been employed; they have the task of feeling the shock wave

load which is computed by the LBE nume ric method. At a later stage, they propagate within the rest of the air domain which, in turn, is also modeled with solid elements. The receptor elements and the air domain make up the Multi-Material Domain. The same free-air-spherical-burst simulation has been conducted using this technique and also in this case the results successfully validated the Load_Blast and ALE data-method (Figure 4).

Explosive Metal FormingThe explosive forming technique is a material processing method where a shock wave is generated by an explosion and propagated through a suitable pressure medium, such as water or air, and deforming a metal plate, tube, or other object. It provides very high straining of the material and also a suitable plastic deformation. The method tested is called shock bulge forming. The advantages of shock bulge forming are reduced spring back and a high strain rate material processing. The material processing at high strain rate leads to an extremely high deformability of the aluminium based alloys.

ConclusionThe graphics above show that numerical and experimental results have a good correspondence about an spherical charge explosion in air. With Explosive Metal Forming application, has been also simulated the fluid-structure interaction. Currently industrial safety are imposing stricter laws about the explosions, therefore numerical simulation can be an important support in order to make safe the structures near high-danger detonation areas.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Fig. 4 - LBE-MM_ALE overpressure valitation

Fig. 5 - Exsplosive metal forming resultsFig. 5 - Exsplosive metal forming results


Finite Element model of Human Fingertip

The fingertip deformation represents the basic mechanical action that shapes human haptic perception. Everyday, humans use their fingertips (and hands) to explore, manipulate and grasp the external environment for many tasks, ranging from simple object grasping towards complex palpation procedures used for medical diagnoses. Moreover, the investigation of fingertip sensing and mechanical properties has gained an increasing attention not only for modeling human behavior, but also in humanoid robotics, where the need for compliant robotic fingers endowed with tactile sensors has become crucial. In this work, we present an experimental set up to provide a characterization of human fingertip mechanical properties, in terms of contact area, fingertip deformation and pressure distribution. Such measures, obtained from experimental tests, are then correlated with the output of a 3D Finite Element (FE) Model of fingertip developed in order to validate the proposed study.

Experimental Tests• The displacement is given by a linear

DC actuator with 9.2 N of full scale.• The forces and torques are recorded

using a 6-DoF force/torque sensor (ATI 6-DoF Nano17).

• The contact area is acquired visually using a video camera, with 1500 x 1120.

• Subject placed the finger pad on the finger-holder, and the indenter surface was moved toward the finger pad.

• The finger was fixed to the finger-holder on the top of the nail, and it was oriented at a 15 deg angle.

• Displacement of indentation 3 mm• Velocity of identation 1 mm/s

Numerical Model• Dermis, bone, nail and plate have

a linear elastic behavior that was modelled using the mechanical properties in Tab 2.

• The subcutaneous tissue was present an hyperelastic behavior and it was modelled by using a Mooney-Rivilin model.

• The solid was meshed by using a 3D tetrahedral element that exhibit a quadratic displacement behavior.

• 31672 elements and 77829 DOF.• In order to optimize the simulation time, the finger model was

divided in two parts according to the sagittal symmetrical plane.• All contacts were modeled as bonded contacts, except the

contact between the finger and the plate which was modeled as a frictionless contact.

• A fixed support constraint was applied on the nail and a displacement of 0.1 mm/s over 30 seconds (3mm overall) was imposed to the lower face of the plate.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Fig.1 Virtual test-rig Fig.2 Physical test-rig

Fig.3 Experimental test concept Fig.4 Experimental test

Fig.5 Materials

Tab.1 Geometrical properties

Fig.6 MeshTab 2 Linear-elatic mechanical properties

Maria Laura D’Angelo, Mariapaola D’Imperio, Ferdinando Cannella, Darwin G. Caldwell

Advanced Robotics Department, Istituto Italiano di Tecnologia

Matteo BianchiUniversity of Pisa

Results1. Experimental Tests

2. Numerical SimulationsThe numerical results were mainly two: Contact Area: the one between the finger and the plate (blue one in Fig. 10)Pressure Area: the one corresponding to the maximum value of pressure [Pa] (red on in Fig. 10).

3. Comparison and Validation


CAE

POST

ER A

WAR

D 20

15: W

INNE

RFig.7 Symmetric plane

Fig.8 Constraints and displacement

Fig.9 Experimental test results

Fig 10 Numerical results

Tab 3 Comparison between Experimental and Numerical Values in terms of Contact Area and Force

Tab.3 Hyperelastic mechanical properties


PulsaR design: CFD comparative study of speed-record Human Powered Vehicles

ContextHigh speed Human Powered Vehicles (HPVs) are specifically designed in order to race at the World Human Powered Speed Challenge (WHPSC), an event taking place in Battle Mountain, Nevada, with the purpose of pushing streamlined bike technology to the limits. The speed is measured on a -0.6% average slope road by timing the HPVs through a 200 m trap after a run-up of 5 miles ( ~ 8km). Record-legal wind must be < 6 km/h.

Shape Variation ComparisonThe design of PulsaR was aimed at the minimum frontal area, given the “compact” size of the rider Andrea Gallo. Crank lenght and Q-factor (the pedal lateral distance) were reduced after experimental confirmation that the power output was not affected by a not-standard configuration.

Given a first draft shape accounting for the main volume constrains (head and helmet, shoulders, legs and feed movement), 5 design versions were created by slight variations of length, nose, tail and wheel cover profile.

Shape Ranking and SelectionThe CFD analysis was performed with CD-Adapco StarCCM+. Particular attention was given to the mesh of the boundary layers, where laminar flow is expected. A shape ranking by drag was obtained, allowing to select the most efficient solution.

Comparative Assessment: PulsaR vs. Varna vs. VeloX IThe same CFD analysis was repeated on two historical reference HPVs in order to assess the overall drag. Left

The selected design solution resulted competitive with these reference prototypes in terms of drag and, as a consequence, in terms of power required to travel at a given speed.Also lift forces were compared showing a reduced down-force for PulsaR, lowering the effect of tire load increase with speed and resulting in less overall rolling resistance.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

P. Baldissera, C. Delprete, A. Gallo Politecnico di Torino


ResultsTre prototype was build in the 1st semester of 2015 and raced at the WHPSC in September after about 850 km of testing. Andrea Gallo achieved the Italian Speed Record at 116.19 km/h (with a peak of 118.4 km/h at the end of the 200m trap), becoming the 24th fastest rider in the history of this discipline.

Further developmentsThe Team will continue to improve PulsaR to approach 120-125 km/h in 2016. In parallel, a new prototype design is being developed aiming at the world speed record in 2017-2018 (now 139.45 km/h).Higher aerodynamic efficiency can be obtained by solving some manufacturing critical aspects: a better thermoforming procedure for the front window plexiglass and a smoother window-fairing connection.

Moreover, while the pressure profile was used to choose air inlet and outlet areas, a CFD analysis including internal ventilation is needed as far as a sensitivity analysis to lateral wind. Win tunnel and on-road experimental validation are also needed. For the new design to be manufactured in 2017, the use of CFD analysis combined with shape optimization tools is considered as a promising approach in order to achieve a stable laminar flow and a reduced drag.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Exhibition of PulsaR vehicle during the International CAE Conference 2015


A computer-aided methodologyfor the design of de-manufacturing processes for waste recycling

• Recycling of Waste Electrical and Electronic Equipment (WEEE) is a challenging task due to their complex material structure.

• Currently, thermal and metallurgical recovery processes are used.

• Corona Electrostatic Separation is a promising technology for the mechanical pre-treatment of shredded waste.

• However, the efficiency of separation is highly affected by: the presence of non-liberated particles in the mixture and the influence of particle-particle interactions and impacts.

• Accurate physical modeling and tight process parameters control is needed.

• State-of-the-art models only model single particle trajectories.

• The aim of this study is to develop a multi-body physical model of CES to capture the effect of the particle impacts on the output recovery and grade.

Method• Multibody environment and colliding surfaces are

automatically generated from CAD (SolidWorks).• The particles in the mixture have been modeled

with nominal particle shapes materials and size classes.

• The 3D electrostatic field has been modeled by FEM and by 2D analytical approximation.

• The Electric forces acting on metal and non metal particles are computed on the basis of the electric field.

• The Aerodynamic forces, the Centrifugal force and the Gravity force acting on the particles are modeled.

• The impacts are simulated by the DVI - based (Differential Variational Inequality) solver integrated in the Chrono::Engine environment.

• About 1 million impacts can be simulated in the granular flow.

Results and Conclusions• Experiments were performed at ITIA-CNR pilot plant using a

controlled mixture of material.

• The validation has been done using a DOE, with three levels for the potential (factor 1) and five levels for the splitters’ configuration (factor 2).

• Simulations with the same experimental conditions have been done.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Fig. 1 - The Corona Electrostatic Separation process

M. Colledani, I. Critelli Politecnico di MilanoM. Diani Istituto di Tecnologie Industriali e Automazione (ITIA-CNR)A. Tasora Università di Parma


• The squared error in the prediction of the metallic fraction recovery rate has been calculated.

• Statistical analysis of the experimental results has been performed using ANOVA method.

• Accuracy: it has been shown by real experiments performed at the ITIA-CNR “Demanufacturing Plant”, that the developed simulation model well predicts the distribution of the throw of conductive and non-conductive particles (Figure 3). As a matter of fact, maximum squared error is about 0,013.

• Statistical analysis: ANOVA method demonstrates that the error is not statistically significantly dependent on factors.

CAE

POST

ER A

WAR

D 20

15: W

INNE

R

Fig. 2 - The CAD of the CES machine is imported in the multi-body simulation environment (Chrono::Engine). Then, the granular flow is simulated

Fig. 3 - Output distribution of metal particles

Poster Award 2015 - Special MentionsA special mentions was given by the Scientific Commitee to the following posters during the Poster Award 2015 Award Cerimony:

On the performance of pressure-velocity coupling techniques for solving incompressible flows using unstructured gridsHermínio Tasinafo Honório - Federal University of Santa Catarina - Florianópolis - BrazilWhen solving Navier-Stokes equations for incompressible flows, the main task one has to adress is how to deal with the well known pressure-velocity coupling. Putting in simple words, it means how to find a pressure field that, when introduced into the momentum equations, will result in a velocity field that satisfies not just the momentum equations themselves, but also the continuity equation. There are a number of strategies that can be employed to deal with this problem and the performance of some of them are compared in this work. Two dimensional unstructured grids are employed and the Element based Finite Volume Method is used to obtain the discretized equations.

Validation of the unsteady friction model in Flowmaster 1D thermofluid simulationJames Flood - Brunel University - UKExperimental results from “Developments in unsteady pipe flow friction modelling” by A. Bergant et. Al. are compared with two pipe friction models in Flowaster for a fast transient case. Friction models used in Flowmaster are the quasi-steady friction model, (using the Colebrook-White approximation), and the unsteady friction model, (based on the Vítkovský formulation of the Bronone unsteady friction model). The unsteady friction model used by Flowmater demonstrates very close agreement with the experimental results, while the quasi-steady friction model demonstrates reasonable agreement with descrepancies as the simulation progresses.

Innovative Numerical Methodologies for Structural Optimization of Civil Steel FrameCarmen Bernardini - University of Salento - ItalyTwo multi-objectives optimization procedures for civil steel frame design are presented. This work is to provides two different alternative procedures to optimize not only the elements’ dimension but also their position within a prototype of a steel frame structure whose goals are user-defined. The structure is fully modeled through beams and joint connections. The integration of CivilFEM and ANSYS WB with modeFRONTIER allows the software to work in synergy in a semi-automatic process, to generate the optimized configuration of the civil design and to check it according to the standards.

For more information: www.caeconference.com

www.enginsoft.com

Global Partner for Master's Programs

Would you like to becomea CAE Professional?

Register now!

Your journey begins here:

International on-line Master'sin Theoretical & Practical Application

of Finite Element Method andCAE Simulation

www.enginsoft.com/news/onlinemastercae.html

newsletter - enginsoft · shape with optimization ... design exploration and optimization ... [icem...

Documents