Structural Non-linear Analysis of Steel/ Aluminium Cover for Hydraulic Actuated Gearbox in ANSYS Structural Environment

Magneti Marelli Powertrain S.p.A. progetta, testa e realizza una versione elettro-idraulica di scatole di cambio automatizzate, nota con il marchio “SeleSpeed”®.
Durante le fasi di progettazione, le varie parti dei componenti meccanici vengono simulate numericamente, utilizzando codici FEM, CFD e 1D, in modo da garantirne resistenza, buone prestazioni e sicurezza. Proprio per quest’ultimo motivo, una delle parti principali dello studio è l’analisi della copertura del freno, un piccolo coperchio che internamente può subire una pressione fino a centinaia di bar.
In questo lavoro si riferisce dell’analisi di una flangia di acciaio inossidabile e un coperchio in alluminio, collegati da quattro viti. I risultati dello studio - particolarmente complesso nella simulazione del problema del contatto superficie/superficie, nella descrizione non lineare della flangia per tener conto di deformazioni plastiche, e nella descrizione del pretensionamento delle viti – sono stati confrontati con quelli di prove sperimentali a banco, per assicurarsi che il metodo di calcolo usato fosse attendibile. L’esito del confronto, del tutto positivo, ha consentito a Magneti Marelli di acquisire sensibilità allo strumento di calcolo, utilizzato, quindi, per assumere decisioni importanti in sede di progettazione.
Questo risultato è dovuto largamente all’ottimizzazione dei parametri utilizzati nella simulazione ottenuta con modeFRONTIER.

Picture 1 – Gearbox Actuator and Studied component (on the right)

Magneti Marelli Powertrain S.p.A. designs, tests and realizes an electro-hydraulic version of gearbox, which is mostly known with the brand name “SeleSpeed”®.
During design phases, various parts of the mechanical component are numerically calculated by means of FEM, CFD, and 1D codes to grant their integrity, good performance and safety.
Mainly for safety reasons, one of the most important part to study is the so-called brake cover, which is a small cover internally stressed by a pressure that can raise up to hundred bars.
In this case study, a flange in stainless steel and a cover in aluminium, which are kept together by four screws, are analyzed, including the following key features:

1. Full use of surface to surface contact elements to simulate the real contact between various sub-parts.
   
2. Non linear flange material, to take into account plastic deformations.
   
3. Pre-stress on screw due to the initial phase of their tightening.

The results obtained were checked ed against test bed results, to validate the reliability of the method.

Introduction
The mechanical behaviour of the Brake Cover for SeleSpeed ® series of gearbox is investigated by means of FE models. The structural set-up is represented in figure 1. The structure is made by two parts:

1. A cover in Aluminium (ERGAL)
   
2. A Flange in Steel

Picture 2 – ANSYS Model

These two parts are joined together and to the valve group (in dark purple in picture 1) by means of four tightening screws.
Different design variables have to be considered, including:

1. Steel flange thickness and shape, to be identified as the best compromise with respect to strength, machinability and strain behaviour;
   
2. Cover shape, in terms of wall thickness and radii of the cap.
   
3. Screw tightening torque.
   
4. Clearance between steel flange and Valve Group.

To obtain the best design, various of the above mentioned variables have to be taken into account.

The Model
The FE model of the brake cover is represented in picture 2.

The steel flange is expected to be the most stressed component. Hence a bilinear kinematic hardening material model is chosen to have a good representation of the behaviour.

The contact between the various parts is described by surface to surface contact elements, and more specifically between:

1. steel flange and aluminium cover (flexible target),
   
2. steel flange and screws heads (flexible target)
   
3. steel flange and valve group surface (rigid target)
   
4. aluminium cover and valve group (rigid target)

To simulate the valve group, only its face in contact with brake cover is taken into account - mainly in order to reduce the size of the model – and this is treated as a rigid fixed region.

The results of the analysis
From the variety of analysis outputs, those on which the design decision are taken, include:

• the Von Mises stress distribution on the aluminium cover, steel flange and screws;
  
• the total plastic strain distribution on the steel flange;
  
• the total displacements,
  
• the Von Mises stress path in the section of the screws.

The results were used to choose the best configuration of the components. Some prototypes were then built, to carry out a series of tests, including the standard Magneti Marelli endurance test.

Picture 3 – Comparison between test and simulation (Steel Flange)

Among the laboratory tests, some were carried out with the aim of validating the FE models, particularly as regards the simulation of the contact between the parts, and the correctness of the stiffnesses involved. This work was done by applying, to a SeleSpeed prototype assembly, a growing oil pressure into the brake cover, up to 150 bars, that is still below the ultimate strength of the system. Then the system was disassembled and the steel flange was analysed. Some photos are shown in picture 3. The results of the FE model proved to match well those obtained with the physical prototype.

Conclusion
Having proved the reliability of the FE models to simulate the performances of the system, some key design decisions were taken by Magneti Marelli on the virtual prototypes, with relevant time saving in the design phase. The gear box is now in production with success. At this stage, a further activity is being carried out to investigate the assembly more in depth, so that to obtain the best balance between the flange thickness and the screw torque. This is typically done taking advantage of modeFRONTIER software environment for the multi objective design optimization.

Ing. Alessandro Brazzi
CAE Senior Analyst
Magneti Marelli Powertrain
Analisys and Simulation
Responsible Ing. Nazario Bellato