Lopez de Lacalle L.N., Lamikiz A. Machine Tools for High Performance Machining

Подождите немного. Документ загружается.

2 New Concepts for Structural Components 51

• The path length from the tool to the workpiece through the structure should be

minimal with the aim of minimising thermal and elastic structural loops [5].

The less material that separates the parts of the structural machine components

that hold the tool and the part the quicker the machine will reach a stable equi-

librium.

With these basic rules, and taking into account that there is not a unique opti-

mal architecture, the design task will focus on making the components of the se-

lected configurations as stiff and light as possible, as well as other considerations

such as the cost, the ease of assembly, the workzone accessibility, the footprint

and the total space occupied by the machine.

Figure 2.2 shows the different architectures that a machine builder considers

when designing a hybrid heavy-duty milling and friction stir welding (FSW) ma-

chine. The company studied six architectures defining seven indicators to allow an

even weighting of the drawbacks and advantages of each architecture. The seven

indicators considered were the following: i) stiffness at the tool tip, ii) cost, iii)

accessibility, iv) flexibility, v) homogeneous and symmetrical behaviour, vi) oc-

cupied space and vii) safety for the end user. Moreover, due to the extraordinarily

high forces associated with FSW processes, which are more than the double the

forces of heavy-duty milling processes, the machine builder has ranked as the

“most important” the characteristics of stiffness, and after that, the characteristics

of cost and risk, and has ranked as the “least important” the flexibility of the ma-

chine and the space that the machine occupies.

The main characteristics of the six considered architectures were:

1. C Structure, with a fixed column, and crossed slides on which a table moves in

X and Y, and vertical ram in Z

2. Fixed bridge structure, with a movable table in X and crossed slides in Y and Z

axes of the tool

3. Fixed bridge structure, with a movable table in X, a movable slide in Y and

a vertical embedded ram in Z

4. C Structure with a fixed column, a movable table in X, and a movable embed-

ded slide in Y and vertical ram in Z

5. Travelling bridge structure, with fixed table, travelling bridge in X, embedded

movable slide in Y and vertically movable crossbeam in Z

6. Fixed bridge structure, with movable table in X, movable embedded slide in Y

and vertical ram in Z

The main conclusion of this comparison study was that number 5 produced the

stiffest architecture. The most economical solution was produced by number 1,

which also provided the most workspace accessible solution. Furthermore, the

most homogeneous behaviour was seen in solutions 5 and 6, both of which also

provide the safest solution. Finally, there are no notable differences as regards the

room that each solution occupies.

Table 2.1 summarises, using a ranking system 1 to 4 (where 1 means a “poor per-

formance” and 4 means an “excellent performance”) the main results of this study:

52 J. Zulaika and F. J. Campa

Fig. 2.2 Different machine architectures for a hybrid milling and FSW machine

1) Open-loop, C frame

2) Closed-loop, C frame

3) Closed-loop, Bridge frame

5) Closed-loop, Bridge frame

4) Open-loop, C frame

6) Closed-loop, Bridge frame

2 New Concepts for Structural Components 53

Table 2.1 Comparisons among different machine architectures

Machine architecture

1 2 3 4 5 6

Stiffness 1.0 2.0 2.5 2.0 4.0 1.0

Cost 4.0 2.0 1.5 3.0 2.0 2.5

Accessibility 4.0 1.0 1.0 2.0 2.0 2.5

Flexibility 3.0 1.0 1.0 1.0 1.0 3.0

Homogenous behaviour 3.0 1.0 2.0 1.0 4.0 4.0

Occupied room 2.0 2.0 2.0 2.0 2.0 2.0

Risk 3.0 1.0 2.0 1.0 4.0 4.0

In the case of selecting adequate machine architecture for a heavy-duty milling

and FSW process and taking into account the high importance of stiffness for

these processes, the selected architecture was the 5th.

2.2.2 Structural Components in Machine Structures

Figure 2.3 shows two machines built by the machine tool builder Nicolas Correa

The drawing in Fig. 2.3a depicts a C type machine: an asymmetrical, open-loop

machine, with a fixed column, a movable table in X, a movable outer slide in Z

and a movable ram in Y. The drawing on the right depicts a Bridge type, symmet-

rical, a closed-loop machine, with a fixed bridge, a movable table in X, a movable

outer slide in Y and a vertical ram in Z.

Despite their remarkable differences in dimensions and occupied room, their stiff-

ness values at the three axes are somehow similar. Therefore the information of inter-

est consists in knowing the most compliant elements of the machine, because as struc-

tural components are mostly in serial, the total compliance of the machine will be

larger than the most compliant component. Moreover, when reinforcing the machine,

the maximum effect is achieved when the most compliant component is acted upon.

Fig. 2.3 Two machines of Nicolas Correa

: a C-type. b Bridge type

54 J. Zulaika and F. J. Campa

Table 2.2 shows the distribution of compliance among the different structural

components of the machine.

The data in this table confirms that concerning the static compliance of ma-

chines, the ram is the most critical component in almost all directions and archi-

tectures. There is also a notable influence from the column with respect to the

bending directions of C-type machines. Therefore, designing optimised rams and

columns for the case of C-frame machines is an issue of special relevance.

The following section will analyse the main rules for designing robust, stiff and

lightweight rams and columns.

2.2.3 Robust Rams and Columns

Rams and columns have in most cases a square section to achieve a symmetrical

and balanced behaviour concerning bending and torsion resistance. As the length

of these components is defined by the strokes of the machine, there are two im-

portant aspects in designing rams and columns: 1) to define the appropriate thick-

ness for their outer walls, and 2) to appropriately place internal ribs as a means of

reaching an optimised stiffness-to-mass ratio.

Table 2.2 Comparisons among different machine architectures

C-type machine Bridge type machine

Component % Static compliance

Component

% Static compliance

X Y Z X Y Z

Ram 49 27 43 Ram 53 42 26

Slide 16 12 24 Slide 10 30 18

Column 24 31 19 Bridge 35 26 46

Others 11 30 14 Others 02 02 10

K/M

Mass M

Stiffness K Stiffness-to-mass K/M

Fig. 2.4 Influence of wall thickness (and mass) on ram stiffness

2 New Concepts for Structural Components 55

With respect to the thickness of the walls, maintaining as a constant the outer

dimensions of the beam, increasing the thickness of outer walls involves an in-

crease of stiffness that is lower than the increase of mass, so that excessively thick

outer walls are not an optimal solution from the eco-efficiency point of view, as

Fig. 2.4 shows. Thickness and mass are directly related.

With respect to ribs, they can be either longitudinal or transversal. Longitudinal

ribs are not an optimal solution for rams and columns from the mechanical point

of view, because for the same amount of mass, an appropriate increase of the

thickness of the outer walls allows for achieving a higher stiffness both for bend-

ing and torsion loads. Therefore, longitudinal ribs are not an optimised solution

according to the stiffness-to-mass ratio.

Fig. 2.5 a Influence of thickness of transversal ribs on stiffness of a ram. b Influence of number

of transversal ribs on stiffness of a ram

56 J. Zulaika and F. J. Campa

Unlike longitudinal ribs, transversal ribs do improve optimally the mechanical

behaviour of rams and columns concerning their bending and torsion resistance.

Therefore, an important issue when designing columns and rams is to select the

appropriate thickness of these transversal walls as well as the distance between

these walls. As a reference, for an average machine ram with a square section, an

optimal thickness for transversal ribs is 10

mm. There is also an optimal space

between transversal ribs, which is of the order of 230–240

mm.

Similar to the case of longitudinal ribs, either an excess in the amount of trans-

versal ribs or an excessive thickness of theirs will increase the total stiffness of

the ram or column at a lower pace than the involved mass, as shown in the charts

of Fig. 2.5.

2.3 Structural Optimisation in Machines

A machine tool is a spatial manipulator that is going to support high loads due to

either the cutting forces or the inertial loads. The optimisation of the machine

constitutive structures is an important task for designers.

2.3.1 Mechanical Requirements for Eco-efficient Machines

When designing the structural components of the machine, the target is to mini-

mise the amount of mass of moving machine structures while static and dynamic

mechanic properties maintain the desired threshold values.

As explained in the chapter introduction, the reduction of moving mass has an

extraordinary impact on the environmental impact of a machine. Therefore it is

worthwhile passing from a strategy based on “the stiffer the machine, the better”

to a strategy based on “the lighter the machine, the better”. Since lighter machines

mean less stiff machines, the latter strategy requires the definition of minimum

“threshold values” for the static and dynamic stiffness, so that an optimal balance

between eco-efficiency, productivity and accuracy can be achieved.

With the aim of contributing to the definition of those threshold values for stiff-

ness, Table 2.3 shows the average process forces associated with milling opera-

tions when cutting AISI 1045 steel, which have been experimentally measured on

Table 2.3 Average forces and acceptable deformations for different milling operations

Process Average force in feed direction Acceptable deformation

Roughing Conventional tools: 1,500

100–125

mm. diam. tools: 3,000

Average: 100

μm

Semi-finishing Conventional tools: 1,000

N Average: 150

μm

Finishing Conventional tools: ,0200

N Average: 110

μm

2 New Concepts for Structural Components 57

different machines and workshops. Table 2.3 also collects the accuracy require-

ments that machine end-users have provided in the survey reported in [7].

Combining the experimental forces and the average requirements coming from

the end users, it was stated that a static stiffness of 20

N/μm at the tool tip can be con-

sidered as a current typical threshold value for general-purpose milling machines.

In addition to the requirement for static stiffness, dynamic stiffness is also an

issue of special concern. Indeed, it is the primary reason to define the dimensions

and shapes of the machine tool structural components. For an accurate definition

of the threshold values for dynamic stiffness, the best utility available is the stabil-

ity lobes diagram.

The stability lobes diagram is a plot that separates stable and unstable machining

operations for different spindle speeds. Stable cuts occur in the region below the

stability boundary, while unstable cuts (chatter) occur above the stability boundary

[3]. An example of the stable and unstable regions is shown in Fig. 2.6. In Chap. 3

the current methods to obtain these diagrams are explained. Stability lobes are func-

tions of the dynamic stiffness at the tool center point (TCP), the tool geometry, the

radial immersion of tool into material, as well as of the material to be machined.

Machine manufacturers and users distinguish between two types of chatter, the

so-called “structural chatter” or “machine chatter”, which is a low frequency chat-

ter, and the so-called “tool chatter”, which appears at higher frequencies. In the

former case, recognisable for a low-pitched sound, the chatter is associated with the

structural modes of the machine, shown in Fig. 2.7. In the latter case, recognisable

for a high-pitched sound, the chatter is associated with the modes of the tool or

spindle. The type of chatter depends on the cutting frequencies; low cutting fre-

quencies excite structural modes and high cutting frequencies excite spindle-

toolholder-tool modes.

Therefore, the aim of an optimised machine is to ensure an acceptable critical

depth of cut in every cutting direction, using the minimum movable mass. As

a conclusion, the “minimum static stiffness” mentioned at the beginning of this sec-

tion and the “minimum critical depth of cut” mentioned just above are the two thresh-

Fig. 2.6 Example of a stability lobe diagram

200 400 600 800 1000

Spindle Speed (rpm)

Depth of cut (mm)

critical depth of cut

STABLE AREA

UNSTABLE AREA

58 J. Zulaika and F. J. Campa

old values to consider in the machine design. That assures that an eco-efficient and

lightweight machine is, at the same time, productive and accurate enough.

The machine mechanical modelling is a key to estimate the static and dynamic

behaviour of a machine. Finite element method (FEM) based models are widely

used, and will be explained over the course of the next section.

2.3.2 FEM Modelling

FEM aims at an approximation of the actual behaviour of a mechanical structure

by assembling discrete and simple elements through nodes. The mechanical analy-

ses that are conducted on a machine tool by means of FEM analyses commonly

cover three stages:

1. The analysis of the static deformations and strain in a structure that is bearing

static forces.

2. The analysis of natural frequencies and modes of the machine structure.

3. The analysis of the dynamic stability of cutting processes by means of analyti-

cal stability diagram lobes. The FEM can be used here as a calculation tool for

the assessment of the frequency response function (FRF), instead of using the

experimental modal analysis.

With FEM, the most complex aspect is that related to contacts between struc-

tural components along the degrees of freedom, in which stiffness, backlash and

friction have a decisive influence on the damping of the machine. More than 90%

of damping in a machine tool comes from bolted joints and from contacts in

guideways [4]. In this respect, experimental modal analysis allows for the meas-

uring of the natural frequencies and modes of an already-built machine, and

above all allows for the measuring of their associated damping coefficients,

which is the most difficult mechanical parameter to estimate. The experimental

Fig. 2.7 View of main

vibration sources in

a milling machine

0 10 100 1000

Feed drive

Forced vibration

Machine-tool structure

chatter

Spindle-toolholder-tool

chatter

Workpiece-Fixturing

chatter

Due to the motion of the machine: Inertial forces

Duetothemachiningprocess

2 New Concepts for Structural Components 59

dynamic parameters allow the calibrating of the FEM models that were developed

in the design phase.

The updated FEM models can be used to calculate the analytical stability lobes

of the machine. Figure 2.8a shows the actual (left) and FEM modelled (right)

FRFs of a milling machine at the TCP. Figure 2.8b and c show the stability lobes

associated with previous FRFs for the downmilling case 45º and 225º with respect

Fig. 2.8 a Experimental and FEM FRFs. b Corresponding stability lobes in the feed direction 0º,

in up-milling. c Corresponding stability lobes in the feed direction 270º, in up-milling

60 J. Zulaika and F. J. Campa

to the machine axes. The experimental data shown in the left part of the chart have

been achieved for a specific tool and a specific material to be machined. The right

part shows the same stability lobe diagrams, with the only difference being that in

this case data has been achieved from a FEM-modelled machine, in which the

damping associated with each mode has been obtained from an experimental mo-

dal analysis on an already built machine.

As is shown, there is a good approximation in the 0º case and a 30% error in

the 270º case. Otherwise, both the modelled-FRF and stability lobes diagrams

show an acceptable level of coincidence with experimental-FRF and stability

lobes. The advantage associated with modelled diagrams is that the machine mod-

els enables an evaluation of the effect of mechanical and architecture changes on

the stability of cutting processes. Thus, it is possible to test several design ap-

proaches to lighten a specific machine and at the same time to validate that the

aimed values of productivity are also achieved.

2.3.3 Topological Optimisation

As shown in previous sections, static stiffness and the critical depth of cut within

stability lobe diagrams offer quantitative and objective parameters that allow

a comparison of the performances among different machine concepts and above

all, allow defining minimum thresholds to be fulfilled by eco-efficient machine

tools. In fact, a reduction of masses that at the same time allows surpassing the

threshold values defined for those parameters will allow an optimisation for that

machine tool from the point of view of the eco-efficiency.

Focusing on reducing the movable masses, the topological optimisation of

structural components is a critical issue, because this optimised mass-to-stiffness

ratio will be the key to tackle effective strategies to reduce the total masses.

With the aim of supporting this objective, there are commercial programs with

optimisation algorithms that starting with a given structural component remove

material on that component until no further removal is possible without deteriorat-

ing the static and dynamic properties, achieving thus a topologically optimised

component. What is more, an important additional objective of topological optimi-

sation programs is to assure that the topologically and dynamically optimised

structure can be manufactured in an economic way.

As an example, Fig. 2.9 shows the optimisation result for C-frame machine.

This typical structure, with a fixed support on one end and forces on the other

end, has been used to analyse the performance of optimisation tools. This example

shows that the topologically optimised structure (Fig. 2.9c) maintains almost the

same stiffness as the original one, but with much less material [2].

With regard to the manufacturability and the economical feasibility of these op-

timised structures, the truss-like structures, which are the most frequent results of

these automated topological optimisations, present some difficulties. As an exam-

ple, cavities inside structures are difficult or impossible to manufacture. To allevi-