Klocke F. Manufacturing Processes 1: Cutting

Подождите немного. Документ загружается.

8.1 Finding Economical Cutting Parameters 351

Amortization period

Costs

· i · t

0.5 · (K

–

) · i · t

Fig. 8.6 Imputed interest

calculation

Since the costs of Eq. (8.39) relate to 1 year, the imputed interest for 1 year K

must be inserted in this equation, resulting in:

+ K

· i (8.43)

Amortization, interest, room costs and ﬁxed maintenance costs are all ﬁxed

quantities. On the other hand, energy costs, tool consumption and the variable main-

tenance costs are variable quantities. If the manufacturing labour is dependant on the

machine run-time, it can be included in the rate of payment. In this case, we refer to

it as a cost per man-hour.

8.1.4 Planning Methods and Tools

According to the type of data determination, we distinguish between manual

(external) and computer-aided cutting data determination. The boundaries here,

depending on the computer use, are ﬂuid (Fig. 8.7).

In the case of manual cutting parameter determination and optimization, docu-

ments from various sources are utilized:

• guidelines of different standardization committees

• catalogues and cutting parameter recommendations of the tool manufacturers

• trial results from research papers

• in-house data banks for cutting parameters

Machining standards, such as are published by the Verein Deutscher Ingenieure

(VDI) (English: Association of German Engineers) for various machining processes

as VDI guidelines, give the process planner only very rough reference values for the

target cutting parameters [VDI3206, VDI3208, VDI3209]. Usually, cutting speeds

are recommended for broadly associated material groups independently of chip

cross-section, tool life behaviour, machine type and machine-hour rate that often

352 8 Process Design and Process Monitoring

Methods for

process optimization

Manual fixed

value optimization

Manual

off-line optimization

Computer aided off-

line optimization

Computer aided

on-line optimization

Indicators

- Standard value

tables

- Nomograms

- Formulas

- Process opti-

mization with

intermittent

sensors

- Process opti-

mization by end

of tool life

- Process optimization

with continuous

sensors and adaptive

control adaption

Fig. 8.7 Methods of process optimization, acc. to GEBAUER [Geba80]

lead to an uneconomically high tool life. Many tool manufacturers today offer s oft-

ware systems (e.g. CIMSOURCE, CoroGuide

, WinTool, etc.), with the help of

which the determination of the cutting parameters can be combined directly with

the selection of a suitable tool. More exact results can be obtained from test reports,

which however are valid in most cases for special machining conditions and can

only be applied to other machining cases with reservations.

Based on these sources, internal cutting data collections are often made as part

of a company’s process preparation. These collections make reference to the accu-

mulated practical machining cases in manufacture and also take into consideration

the long-term experiences of specialists and skilled workers. Information centres f or

cutting data, which collect cutting data from industry and research and make these

available to companies in a form speciﬁcally tailored to them, are another modern

phenomenon. Figure 8.8 shows the input mask of a data bank as an example of such

a type of cutting parameter selection. These data bank systems contain cutting data

for set tool life parameters for a comparable workpiece material/cutting tool mate-

rial combination and under consideration of geometrical inﬂuencing variables. Such

a data bank makes it possible to process large collections of previously determined

machining data under consideration of many inﬂuencing parameters and to provide

improved t arget values.

8.2 Process Monitoring

Secure process monitoring has a crucial role in manufacturing technology.

Uninterrupted, fully or partially automated manufacturing processes guarantee high

productivity and optimal capacity utilization. Process disturbances are, however,

unavoidable. Stochastically occurring tool failure, deviations in the composition of

8.2 Process Monitoring 353

Fig. 8.8 Example for technology data organization (Source: CIM GmbH)

the machined material or faulty planning/programming prior to machining are all

potential causes for such interruptions.

The goal of process monitoring in manufacturing technology is to recognize these

process disturbances in a timely fashion and to introduce proper countermeasures.

For example, extensive tool, workpiece or machine tool damage can result from tool

fractures if the proper countermeasures are not promptly applied, be it manually

or automatically (i.e. with the help of monitoring systems). This means, among

other things, that the machining process must be monitored in order to minimize

consequential costs of process disturbances.

Process monitoring systems make it possible to monitor cutting tools. The tools

are thereby monitored for sudden tool fractures or overloads and for increasing

tool wear. Tool monitoring has been developed to varying extents depending on

the various machining processes with deﬁned cutting edge geometries. Monitoring

of the drilling process has been widely researched and ﬁnds extensive use in indus-

try. There are also different approaches from research and practice for realizing

a reliable tool fracture and wear-recognition monitoring system for turning pro-

cesses. In the case of milling, fracture and wear monitoring is much more costly

than in the abovementioned processes due to the more complicated process kine-

matics. Accordingly, there are only a few monitoring solutions that exist for milling

processes.

Common to all monitoring systems is that process-relevant information is mea-

sured with the help of specialized sensors. The most common sensors and sensor

principles today will be introduced in the next chapter.

354 8 Process Design and Process Monitoring

8.2.1 Sensors for Process Monitoring

A number of sensors are available on the market for monitoring manufacturing

processes, of which most can be classiﬁed in six physical functional principles.

Table 8.2 shows an allocation of typical measurement parameters and the physical

functional principles that primarily underlie them.

In order to standardize usage and to lend semantic clarity to the terms used, the

essential fundamental concepts of metrology will ﬁrst be deﬁned [Pfei01]:

• The measurement parameter is the physical parameter that is to be

measured.

• Measurement is the execution of activities planned for the quantitative compari-

son of measurement parameters with the unit.

• The measurement result is the value taken from the measurements estimated as

the true value of a measurement parameter.

• The measurement principle is the physical basis of the measurement.

• The measurement method is deﬁned as a special kind of procedure used in the

measurement independent of the measurement principle.

• A measurement process is the practical application of a measurement principle

and a measurement method.

A further distinction is drawn between direct and indirect measurement meth-

ods. Direct measurement methods record the measurement parameter of interest

by means of a comparison with a normal of the same physical parameter. Indirect

measurement methods, in order to determine the measurement parameter, capture

an auxiliary parameter which stands in a known and describable relation to the

measurement parameter [Pfei01].

While for direct methods, it is often difﬁcult to reach the measurement loca-

tion with suitable sensors, the problem with indirect measurement methods is in the

determination of an assignment rule, the validity of which should not be affected

Table 8.2 Relation of measurement parameters to physical functional principles

Physical functional principles

for process description Associated measurement parameters

Mechanical (motion, displacement, stiffness) Position, acceleration, velocity, force, torque,

longitudinal/torsional stress,

longitudinal/torsional strain, pressure etc.

Thermal (kinetic energy of atoms and

molecules)

Temperature, heat ﬂow, speciﬁc heat, thermal

conductivity etc.

Electric (electric ﬁeld) Voltage, current, charge, conductivity etc.

Magnetic (magnetic ﬁeld) Permeability, magnetic current etc.

Radiation (electromagnetic radiation) Energy, intensity, emission, reﬂection,

permeability etc.

Chemical (forces between atomic nuclei and

electrons, bond energies of molecules)

Chemical components, concentrations etc.

8.2 Process Monitoring 355

Acceleration

Acoustic emission

Heat radiation

Resultant

force

Fig. 8.9 Process emissions

for monitoring of cutting

process

by changes to the process parameters. Yet the advantages and disadvantages of

the direct and indirect measurement methods cannot be generalized and must be

weighed for every particular case.

The following is concentrated on the monitoring of cutting processes. Figure 8.9

shows the process emissions relevant for this area.

Numerous sensors use the strain of an object as an auxiliary function for mea-

suring force, torque and acceleration. For this reason, the following sections will

examine the workings of strain gauge strips and piezoelectric sensors.

8.2.1.1 Strain Gauges

A strain gauge converts a mechanical strain into an electrical resistance change. The

HM resistance R of a metallic conductor is a function of the material (ρ

), length

(L) and cross-section (A) according to the f ollowing relation:

R = ρ

(8.44)

In the simplest case, a strain gauge consists of a resistance wire, which is fre-

quently incorporated into a supporting matrix. The strain gauge is fastened to

the shifting position on the measuring body. The adhesive is of great importance,

because the strain or compression of the measuring body must be transferred unal-

tered to the strain gauge via the adhesive connection. Assuming that the material

properties of the strain gauge are not changed by the strain (ρ

= constant), the

resistance change is dependent only on the length and cross-section change. In gen-

eral, the relation between strain and relative resistance change can be described by

Eq. (8.45). Strain sensitivity k, for example for a constantan strain gauge, amounts

356 8 Process Design and Process Monitoring

to k = 2.044. The normal operating range of strain gauges extends up to a relative

strain/compression of about 1%.

R/R = k · ε (8.45)

In order to reduce the dimensions in comparison to the single-wire form, usually

meander-formed designs are utilized (Fig. 8.10).

However, the deﬂection areas cause a sensitivity of up to 3% transverse to the

measurement direction. Strain gauges are not limited to the use of resistance wires.

Semiconductor, foil and thin-ﬁlm materials are also employed.

The resistance changes arising during the use of strain gauges are very small

and are only a fraction of its total resistance. With the help of a bridge circuit, it is

possible to measure the arising resistance changes in the form of a voltage variation

. For a balanced bridge circuit with R

= R

= R and a feed voltage

, U

= 0.TheWHEATSTONE bridge circuit is s hown in Fig. 8.11.

Also valid is the relation given in Eq. (8.46)

= U

(

R + R

)(

R + R

)

−

(

R + R

)(

R + R

)

(

2R + R

+ R

)(

2R + R

+ R

)

(8.46)

Depending on whether two or four active strain gauges are used, this relation can

be simpliﬁed in the form of a quarter, half or f ull bridge. For a quarter bridge with

Carrier foil

Connectors

Measurement meander

Fig. 8.10 Sketch of a strain

gauge strip with measurement

meander

Fig. 8.11 Wiring diagram of a WHEATSTONE bridge

8.2 Process Monitoring 357

R

= R and R

= R

= 0 and R << R the Eq. (8.47) is valid. It

is used for simple measurements with a single strain gauge.

= U

R

(8.47)

A half bridge can be used to increase sensitivity by applying two identical strain

gauges so that the amount of strain of one is equal to the amount of compression of

the other, i.e. R

= R, R

=−R and R

= R

= 0. Eq. (8.48)istruein

this case.

= U

R

(8.48)

In the case of the full bridge, all four arms of the bridge circuit are ﬁt with strain

gauges. This makes it possible to raise sensitivity ever further and, in case all the

strain gauges are exposed to the same temperature, a temperature compensation is

also possible. Equation (8.49) is valid here with R

= R

=−R and R

R

=−R.

= U

R

(8.49)

8.2.1.2 Piezoelectric Sensors

Piezoelectricity is caused by the electromagnetic interaction between the mechan-

ical and electrical state of crystals that have no centre of symmetry, e.g. Quartz

(SiO

), tourmaline and ferroelectrical ceramics. By deforming the crystal lattice

along the polar axes, the positive and negative lattice elements are displaced relative

to one another, generating an electrical dipole moment. This charge transfer can be

converted into a voltage signal with the help of a charge ampliﬁer. In the case of

piezo-actuators, the inverse piezoelectric effect is utilized, in which a mechanical

change in length is produced by applying voltage.

Piezoelectric materials have one or more polar axes. Their piezoelectric coef-

ﬁcient is direction-dependent and indicates sensitivity in the direction of the

corresponding axis. Piezoelectric coefﬁcients range from 2 to a present maxi-

mum of 1500 pC/N. This should be taken into consideration when cutting crystals

(Fig. 8.12).

Fig. 8.12 Possible cuts

through a synthetic crystal

(Source: Kistler)

358 8 Process Design and Process Monitoring

Stressed crystal

++++++++++

–

Unstressed crystal

–

Fig. 8.13 Longitudinal effect

in piezoelectric materials

In the case of the longitudinal effect, a charge transfer occurs on the force contact

surface as a result of a mechanical deformation along a polar axis of the crystal

perpendicular to the pickup of the charge (Fig. 8.13).

The charge difference should not be increased with the geometrical dimensions

of the crystal plates, but only by mechanical series connection and electrical par-

allel connection of several plates. The amount of charge can be calculated for the

longitudinal effect with Eq. (8.50), with d

as the piezoelectric coefﬁcient in the

x-direction, F

as force in the x-direction and n as the number of crystal plates.

= d

· F

· n (8.50)

In the case of the shear effect, stress occurs as in the longitudinal effect along the

polar axis of the crystal. However, in this case a shear force arises in the direction

of the charge pickup (Fig. 8.14).

Unstressed crystal

–

Stressed crystal

–

+++++++++

–

Fig. 8.14 Shear effect in

piezoelectric materials

8.2 Process Monitoring 359

Unstressed crystal

–

Stressed crystal

–

Fig. 8.15 Transversal effect in piezoelectric materials

The charge thereby released can be computed using Eq. (8.51), with d

as the

piezoelectric coefﬁcient in the x-direction, F

as force in the x-direction and n as the

number of crystal plates.

= 2 · d

· F

· n (8.51)

In the case of the transversal effect, stress is applied in the direction of a neu-

tral axis of the crystal, and the charge occurs orthogonally to it on a polar axis

(Fig. 8.15).

In this case, the amount of charge should be inﬂuenced by the dimensions of the

piezo-element and can be calculated with Eq. (8.52), with d

as the piezoelectric

coefﬁcient in the x-direction, F

asforceinthey-direction and b/a as the geometric

ratio.

=−d

· F

(8.52)

Another essential component of piezoelectric sensors are the isolators. Isolators

are subject to extremely high requirements in order to prevent small, piezoelectri-

cally generated charge transfers from ﬂowing off rapidly and distorting the reading,

especially in the case of static and quasistatic measurements. The isolation effect

must also be guaranteed during use under high temperatures and under the inﬂu-

ence of moisture. Furthermore, the isolation materials must have a high level of

stiffness for transferring mechanical stresses in order to support the piezo-element

against the housing.

8.2.1.3 Force Sensors

For measuring forces and torques in manufacturing technology, piezoelectric sen-

sors and, occasionally, inductive and capacitive position transducers are utilized. In

general, all systems convert a stress into an electrically evaluable magnitude.

Fundamentally, one can measure either in the main force connection or in the sec-

ondary force connection. Especially when the sensor is arranged in the main force

360 8 Process Design and Process Monitoring

connection, one must be sure that, on the one hand, the stiffness properties of the

entire system are not negatively affected by the integration of the sensor and, on the

other hand, that the sensor itself is not damaged by an overload. The high stiffness

and high overload stability are advantageous in this regard. However, the sensors

can sustain only extremely low tensile stresses, so that they must be sufﬁciently

prestressed with a compressive stress in the case of tensile or tumescent stress.

Figure 8.16 shows the structure of a piezoelectric sensor used for measuring force

components. Two longitudinally sensitive quartz plates serve as the sensor elements.

In order to measure torque, two basic principles can be used. When longitudi-

nally sensitive force sensors are used, it is possible to make inferences about the

torque from non-parallel sensors under consideration of their reference coordinates.

If shear-sensitive force sensors are used, the torque applied can be directly mea-

sured (Fig. 8.17). The plates are arranged mechanically in a series and connected

electrically parallel so that the charge is proportional to the torque applied. Since

the shear stresses are non-positively transferred, the sensors require a high level of

pre-stressing.

In the case of sensors designed to capture several components, a corresponding

number of additional quartz plate pairs of varying orientation are used. Figure 8.18

shows a four-component dynamometer used for measuring three force components

and torque.

Circular covering plate

Connector

Housing

Base plate

Electrode

Quartz plates

Fig. 8.16 Schematic view of a piezoelectric dynamometer

Shear sensitive

quartz plate

Fig. 8.17 Schematic view of

assembly of shear sensitive

quartz plates in a torque

sensor