Milling analysis by 3D FEM and experimental tests.
Constantin, Corina ; Bisu, Claudiu-Florinel ; Croitoru, Sorin Mihai 等
1. INTRODUCTION
Milling is a common form of machining designated for creating a
great variety of surfaces. It becomes very important to have an approach
for predicting cutting forces, chip formation, thermal aspects, etc. The
importance comes from the necessity of using optimum technological
parameters for processing by milling different materials, and also for
determining loads (forces and torque) during operation. They are useful
for tool designing and also in tool functioning (damage and wear rate).
The analysis by FEM modelling and simulation becomes very powerful in
this field. It still needs for the confirmation of the method the
support of experimental tests for validation.
Finite Element Method (FEM) permits prediction of cutting forces,
stresses, tool wear, and temperatures of the cutting process so that the
cutting tool can be designed. FEM has some advantages such as (Kirichek
& Afonin, 2007 ): solves contact problems, uses bodies made from
different materials, a curvilinear region can be approximated by means
of finite elements or described precisely etc. There are two types of
finite element formulations to describe a continuous medium: Lagrangian
and Eulerian (Bareggi & O'Donnell, 2007). Based on the success
of FEM simulations for different processes, many researchers developed
their own FEM codes to analyze metal cutting processes (Cerenitti et
al., 1996).
Applications of FEM models for machining can be divided in six
groups: tool edge design, tool wear, tool coating, chip flow, burr
formation plus residual stress, and surface integrity (H. Yanda et al.,
2009).
The right choice of finite element software is very important in
determining the scope and quality of the analysis that will be
performed. The most important software codes used for simulation of
metal cutting are: Abaqus, Deform 2D and 3D (Uhlmann et al., 2007), and
AdvantEdge.
In this paper the Deform 3D commercial software is used to simulate
the milling process (www.custompartnet.com).
2. FEM ANALYSIS AND EXPERIMENTAL
VALIDATION
The FEM analysis consists of three steps: pre-processor, simulation
and post-processor (Deform 3D-V6.1, User's Manual). In the
pre-processor the initial data for modelling and simulation must be set.
The process parameters and cutting conditions are described in Table 1.
The next steps are tool and workpiece setup, material choice, and mesh
generation. For the studied process, a milling tool Sandvik
R365-080Q27-S15M (www.sandvik.com) of 80 mm diameter with inserts was
designed in Aut [degrees]AD and then imported in the software. $$ The
software generates a workpiece based on the properties presented in
Table 2. The tool is made of WC and the workpiece material is AISI1045
(Steel).
The mesh generation is very important for accuracy of the
simulation. The mesh is reformulated at nearly every time step, in order
to manage the material deformation. Fig. 1 shows an example of deformed
workpiece mesh for the milling process.
The end of the pre-processor step contains the simulation controls
and data base generation (Table 3). After completing these steps, the
database can be generated. At this step, the simulation can be started.
The simulation initiates a series of operations and generates a new mesh
if necessary.
The last step is the post-processor. The user can check and use the
simulation results after the data extraction.
The most important data obtained from the FEM simulation are:
geometry of workpiece and tool after the simulation; tool movements and
deformed mesh at each saved step (Fig. 1); distribution of state
variables: stress, strain, temperature, wear, damage (Fig. 2);
displacement and velocity; chip formation; predicted cutting forces and
torque (Fig. 3).
[FIGURE 1 OMITTED]
For an assessment of the cutting simulations, experimental tests
have been carried out. The experimental setup consisted of a vertical
machining centre FIRST MCV 300, a Kystler dynamometer, an amplifier
connected to the computer acquisition motherboard, a workpiece of
pre-shaped of AISI 1045 steel and a Sandvik R365-080Q27-S15M milling
head.
The cutting conditions were the same as those presented in Table 1.
The difference between the simulation and experiment was the following:
the simulation was conducted with a tool with one tooth and the tool
used in experiments had 6 teeth.
The measured cutting forces are presented in Fig. 4: the feed force
[F.sub.X] = 50 N, cutting force [F.sub.Y] = 110 N and axial force
[F.sub.Z] = 60 N.
To be able to compare the calculated cutting forces with the
measured ones, the user had to determine the simulated specific cutting
forces. For this, the integral average was computed. After that, the
quotient of the integral and the time fragment was calculated. The
simulated and the measured cutting forces show small differences; this
can be ascribed to problems concerning the model.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Besides model improvement, also the model validation vill have in
view the friction parameter establishing for different tool-part
material couples in high speed.
As a future research goal we can mention the use of 3D FEM analysis
for inverse simulation to deduce the Johnson-Cook parameters that
describe the material law (Shrot & Baker, 2010) used in machining
simulation for high speed processes. This will be done on the basis of
the adiabatic stress-strain curves obtained by FEM.
3. CONCLUSION
This paper proposed an overview of the approach of FEM analysis of
a milling process considering 3D modelling and also an experimental
validation. The simulation was conducted with a single tooth tool and
the program generated the workpiece. The tool used in the experiment was
a Sandvik milling tool with 6 inserts. The experimental results validate
in a largely way the measured cutting forces but for a better agreement
the model can be improved. Building improved models and further
experiments on different materials will be among the main tasks for
further work including material law parameter finding.
4. ACKNOWLEDGEMENTS
The work has been funded by the Sectoral Operational Programme
Human Resources Development 2007-2013 of the Romanian Ministry of
Labour, Family and Social Protection through the Financial Agreement
POSDRU/88/1.5/S/61178.
5. REFERENCES
Bareggi A.; O'Donnell G.E. (2007). Modelling Thermal Effects
in Machining by Finite Element Methods, Proceedings of the 24th
International Manufacturing Conference, Vol. 1, 2007, pp. 263-272
Cerenitti E.; Fallbohmer P.; W.T. Wu & Altan T. (1996).
Application of 2D FEM to Chip Formation in Orthogonal Cutting, Journal
of Material Processing Technology, Vol. 59, 1996, pp. 169-180
Hendri Y.; Ghani J. A.; Hassan C. & Haron C. (2009). Effect of
rake and clearance angles on the wear of carbide cutting tool, Eng.
e-Transaction, Vol. 4, No. 1, 2009, pp. 7-13
Kirichek A.V.; Afonin A.N. (2007). Stress-Strain State of the
Thread-Milling Tool and Blank, Russian Engineering Research, Vol. 27,
No. 10, 2007, pp. 715-718
Shrot A.; Baker M. (2010). Is it possible to identify Johnson-Cook
law parameters from machining simulations? Int J Mater Form, Vol. 3,
Suppl 1, 2010, pp. 443-446
Uhlmann E.; Graf von der Schulenburg M.; Zettier R. (2007). Finite
Element Modelling and Cutting Simulation of Inconel 718, Annals of the
CIRP, Vol. 56, No. 1, 2007, pp. 61-64
*** Deform 3D-V6.1 User's Manual
*** www.custompartnet.com, Accessed on: 2010-08-20
*** www.sandvik.com, Accessed on: 2010-07-27
Tab. 1. Process and condition setup
Process and condition parameters Milling
Cutting speed 75.36 m/min
Feed 2.4 mm/sec
Depth of cut 0.5 mm
Shear friction coefficient 0.5
Interface heat transfer coefficient 45[degrees]C
Convection coefficient 0.02
Environment temperature 20[degrees]C
Tab. 2. Workpiece properties
Workpiece parameters
Geometry Modelled as plastic
Length 20mm
Material AISI1045 (Steel)
Tab. 3. Simulation controls
Simulation controls
Nr. of steps 10 000
Steps to save / steps def. 25
Tool wear calculation Usui's Model: