THEORETICAL AND EXPERIMENTAL STUDY OF TOOLING SYSTEMS
‐ PASSIVE CONTROL OF MACHINING VIBRATION ‐
Lorenzo Daghini Licentiate Thesis
School of Industrial Engineering and Management Department of Production Engineering The Royal Institute of Technology, Stockholm
JUNE 2008
TRITA‐IIP‐08‐04 ISSN 1650‐1888
Copyright © Lorenzo Daghini
Department of Production Engineering
The Royal Institute of Technology
S‐100 44 Stockholm
“Nissuna umana investigazione si pò dimandare vera scienzia s'essa non passa per le matematiche dimostrazioni, e se tu dirai che le scienzie, che principiano e finiscono nella mente, abbiano verità, questo non si concede, ma si niega, per molte ragioni, e prima, che in tali discorsi mentali non accade esperienzia, sanza la quale nulla dà di sé certezza”.
No human investigation can be called true science without going through the mathematical tests… the sciences which begin and end in the mind cannot be considered to contain truth either, because such discourses lack experience, without which nothing reveals itself with certainty.
Leonardo da Vinci
I
Abstract
Vibration control has been and still remains a subject of primary importance in modern manufacturing industry. To be able to remove high volumes of material in shorter time as well as to be able to get the right quality of the parts at the first time are goals that many shops would like to achieve. Tooling systems, and especially cantilever tools, and cantilever structural units of machine tools are the least rigid components of machining systems and therefore the most prone to vibration. Boring tools are often encountered as rotating tools in machining centres or as stationary tools in internal turning. In this thesis the focus is on internal turning. Internal turning is widely known as a very delicate operation and it is often carried out with cutting parameters far from optimal, from a productivity point of view, due to limitations imposed by vibration. Another type of tooling system whose functionality is impaired by vibration is the parting‐off tool. The design of damped parting‐off tool is one of the focus of this thesis as well.
Vibration control has the purpose to achieve an efficient energy dissipation of a vibrational system. Basically this is achieved by controlling the damping of the system. Since damping involves the conversion of energy associated with a vibration to other forms, there are several mechanisms to remove energy from a vibrating system. Typically these mechanisms are divided in two classes:
1. Mechanisms that convert mechanical energy to heat, i.e. passive damping.
2. Mechanisms that transport energy away from vibrating systems, i.e.
active damping.
Both these techniques have been used during the years and both have been giving excellent results. The active vibration control mechanisms are more expensive and not suitable for machining due to the cables they necessitate that could interfere with the machining operation.
This work proposes an original approach to vibration damping in machining systems, the objects of vibration dissipation being the structural components on the link between turret and cutting insert. The idea is to use composite materials to create damping interfaces between and within the different structural components. Different clamping system designs are being compared in order to see how these influence the performance of the machining system and different cutting inserts have been compared for machining hardened steel.
The newly designed components have been going through both extensive off‐
line (modal analysis) and on‐line dynamic testing (machining test) and the results show that the new tool holders used in combination with hydrostatic clamping system are the most optimal solution among the tested ones.
The new design for the turret has been giving promising results and more can be achieved by bringing minor changes to it, these changes are being implemented at the time of writing this thesis.
III
Acknowledgments
This work would have not been possible without the precious participation and help of Mircona, who provided the specially designed tools, Spirex‐tools with their hydraulic clamping devices and the work on the turret, and SSAB who supplied the necessary material for running the machining tests.
A special thank goes to my supervisor, prof. Cornel‐Mihai Nicolescu, for the invaluable support and help in all the different phases of this work, as well as to my colleagues Anders Berglund, Andreas Archenti and Mathias Werner for the continuous and intensive exchanges of opinions. To be not forgotten are the technicians at the production engineering department, Jan Stamer and Leif Nilsson, for their patient and precious help in handling the machines in the laboratory.
This work was financed by EUREKA program within the Dampcomat project (project nr. 25953‐1), through VINNOVA (Swedish Governmental Agency for Innovation Systems).
V
Table of contents
1 INTRODUCTION __________________________________________________ 1
1.1 B ACKGROUND _________________________________________________ 1
1.2 T HESIS STRUCTURE ______________________________________________ 2
2 DESIGN OF DAMPED COMPONENTS _________________________________ 3
2.1 S TATE OF THE ART ______________________________________________ 3
2.1.1 Active damping ___________________________________________ 3
2.1.2 Passive damping ___________________________________________ 4
2.2 C ONCEPT ____________________________________________________ 5
2.3 D ESIGN DETAILS _______________________________________________ 9
2.3.1 Boring bar ________________________________________________ 9
2.3.2 Parting‐off and grooving tool _______________________________ 12
2.3.3 Turret __________________________________________________ 12
3 PERFORMANCE EVALUATION _____________________________________ 15
3.1 B ORING BAR _________________________________________________ 15
3.1.1 Modal Analysis ___________________________________________ 15
3.1.2 Machining tests __________________________________________ 19
3.2 P ARTING ‐ OFF TOOL ____________________________________________ 35
3.2.1 Modal Analysis ___________________________________________ 35
3.2.2 Machining tests __________________________________________ 39
3.3 T URRET ____________________________________________________ 47
3.3.1 Modal analysis ___________________________________________ 47
4 CONCLUSIONS AND FUTURE WORK ________________________________ 49
4.1 B ORING BAR _________________________________________________ 49
4.2 P ARTING ‐ OFF TOOL ____________________________________________ 49
4.3 T URRET ____________________________________________________ 49
4.4 F UTURE WORK _______________________________________________ 50
REFERENCES ________________________________________________________ 53
PAPERS ____________________________________________________________ 55
P APER I ____________________________________________________________ A
P APER II ___________________________________________________________ O
P APER III __________________________________________________________ EE
VII
Table of figures
Figure 1‐1 Machine tool components taken in consideration in the Dampcomat project. __ 2 Figure 2‐1 Solid model view of the active tool adaptor (4). __________________________ 3 Figure 2‐2 Sandvik Coromant Silent tool boring bar. The pre‐tuning system of the tuned bar
consists mainly of a heavy tuning body (A) with a certain inertia mass, suspended in two rubber bushes (B), one at each end of the tuning body. The tuning body is surrounded by a special oily liquid (C). If vibration tendencies should arise during the machining process using a tuned bar, the damping system will immediately come into force, and the movement‐energy of the bar will be absorbed into the tuning system. 5 Figure 2‐3 Hydrostatic clamp(left) and screw clamp (right). __________________________ 6 Figure 2‐4 Comparison between screw (above) and hydrostatic clamp (below). The real
overhang for the tool clamped with screws (above) is longer than the measured one, usually 12 mm longer, that is the distance between the outer face of the holder and the first screw. _________________________________________________________ 7 Figure 2‐5 Parting‐off tool. ____________________________________________________ 8 Figure 2‐6 Deflection of the workpiece due to the radial character of the cutting force and
the distance of the cutting area from the chuck. ______________________________ 8 Figure 2‐7 Energy propagation paths. If there is a metal‐to‐metal contact in the structure the energy will prefer to follow that path instead of going through the damping material. 9 Figure 2‐8 Damping ring, front view (right) and axonometric view (left). ______________ 10 Figure 2‐9 Channels for the glue flow. __________________________________________ 10 Figure 2‐10 Specially shaped ring for allowing the coolant flow through the tool. _______ 10 Figure 2‐11 Section of the damping structure. In yellow the tool body, in green the damping
material and in blue the protective discs. In the detail bubble is shown how the protective disc does not touch the collet. ___________________________________ 11 Figure 2‐12 CAD view of the boring bar. _________________________________________ 11 Figure 2‐13 CAD model of the parting‐off tool. (a) Exploded view. (b) view of the assembled
tool. (c) View of the back of the tool. ______________________________________ 12 Figure 2‐14 On the left a standard turret with VDI system. On the right the new turret, front
view. The available tool‐holder sizes are: two ∅42, one ∅12, one ∅16, one ∅25 and one ∅32 mm. _________________________________________________________ 13 Figure 3‐1 Tightening order. __________________________________________________ 15 Figure 3‐2 Measuring configuration. The figure shows the accelerometers position and the
excitation point. _______________________________________________________ 16 Figure 3‐3 Modal analysis result. The tools were clamped on the conventional turret with a
VDI adapter with screws. ________________________________________________ 17 Figure 3‐4 Result of the modal analysis: compliance. In dark green the damped tool in VDI
adapter for Hydrostatic clamp. In black the damped tool in the Hydrostatic turret. In pink the damped tool in VDI with screws. In red the conventional tool in VDI adapter for hydrostatic clamp. In light green the conventional tool in hydrostatic turret. ___ 18 Figure 3‐5 How to read the waterfall diagram. On the vertical axis the signal amplitude, on
the horizontal the frequency and on the third axis is the time. Since cutting speed has
been increased in every recording, the time indicates the increase of the speed as well. _________________________________________________________________ 21 Figure 3‐6 How to read the time record diagram. Here only time and signal amplitude are
represented. The cutting speed increases from left (40 m/min) to right (80 m/min). 21 Figure 3‐7 Time record of noise generated during machining at v
c= 40 to 80 m/min f = 0.15
mm/rev and a
p= 1 mm. In red the conventional tool and in green the damped one. 22 Figure 3‐8 Frequency content of the signal produced when machining at v
c= 40 to 80 m/min
f=0.15 mm/rev and a
p= 1 mm. On the left the conventional tool, on the right the damped one. __________________________________________________________ 22 Figure 3‐9 Time record of noise generated during machining at v
c= 40 to 80 m/min f = 0.15
mm/rev and a
p= 2 mm. In red the conventional tool and in green the damped one. 23 Figure 3‐10 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f = 0.15 mm/rev and a
p= 2 mm. On the left the conventional tool, on the right the damped one. _______________________________________________________ 23 Figure 3‐11 Time record of noise generated during machining at v
c= 40 to 80 m/min f = 0.15
mm/rev and a
p= 0.5 mm. In red the conventional tool and in green the damped one.
_____________________________________________________________________ 24 Figure 3‐12 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f=0.15 mm/rev and a
p=0.5 mm. On the left the conventional tool, on the right the damped one. _______________________________________________________ 24 Figure 3‐13 Time record of noise generated during machining at v
c= 40 to 80 m/min f=0.15
mm/rev and a
p= 1 mm. In red the conventional tool and in green the damped one. 25 Figure 3‐14 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f=0.15 mm/rev and a
p= 1 mm. On the left the conventional tool, on the right the damped one. _______________________________________________________ 25 Figure 3‐15 Time record of noise generated during machining at v
c= 40 to 80 m/min f = 0.15
mm/rev and a
p= 1.5 mm. In red the conventional tool and in green the damped one.
_____________________________________________________________________ 26 Figure 3‐16 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f=0.15 mm/rev and a
p= 1.5 mm. On the left the conventional tool, on the right the damped one. _______________________________________________________ 26 Figure 3‐17 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.3mm/rev and a
p= 1 mm. In red the conventional tool and in green the damped one. _________________________________________________________________ 27 Figure 3‐18 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f= 0.3 mm/rev and a
p= 1 mm. On the left the conventional tool, on the right the damped one. _______________________________________________________ 27 Figure 3‐19 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.15mm/rev and a
p= 2 mm. In red the conventional tool and in green the damped one. _________________________________________________________________ 28 Figure 3‐20 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f= 0.15 mm/rev and a
p= 2 mm. On the left the conventional tool, on the right
the damped one. _______________________________________________________ 28
IX
Figure 3‐21 Time record of noise generated during machining at v
c= 40 to 80 m/min f=0.15mm/rev and a
p= 0.5 mm. In red the conventional tool and in green the damped one. _________________________________________________________________ 29 Figure 3‐22 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f=0.15 mm/rev and a
p= 0.5 mm. On the left the conventional tool, on the right the damped one. ______________________________________________________ 29 Figure 3‐23 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.3mm/rev and a
p= 0.5 mm. In red the conventional tool and in green the damped one. _________________________________________________________________ 30 Figure 3‐24 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f=0.3 mm/rev and a
p= 0.5 mm. On the left the conventional tool, on the right the damped one. ______________________________________________________ 30 Figure 3‐25 Time record of noise generated during machining at v
c= 40 to 80 m/min and
a
p=0.5 mm. In red the conventional tool at f=0.15 mm/rev and in green the damped one at f=0.3mm/rev. ___________________________________________________ 31 Figure 3‐26 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.3mm/rev and a
p= 1 mm. In red the conventional tool and in green the damped one. _________________________________________________________________ 31 Figure 3‐27 Frequency content of the signal produced when machining at v
c= 40 to 80
m/min f= 0.3 mm/rev and a
p= 1 mm. On the left the conventional tool, on the right the damped one. ______________________________________________________ 32 Figure 3‐28 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.15mm/rev and a
p= 1 mm. In green the conventional tool clamped in Hydrostatic turret, in red in VDI with screws and in blue in VDI with hydrostatic clamp. _______ 32 Figure 3‐29 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.15mm/rev and a
p= 1 mm on hydrostatic turret (green) and a
p= 2 mm on
hydrostatic VE turret (red). ______________________________________________ 33 Figure 3‐30 Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.15mm/rev and a
p= 1 mm. In green the damped tool clamped in Hydrostatic turret, in red in VDI with screws in blue in hydrostatic VE turret and in purple in VDI with hydrostatic clamp. _________________________________________________ 34 Figure 3‐31 Left: Time record of noise generated during machining at v
c= 40 to 80 m/min
f=0.15 mm/rev and a
p= 1.5 mm. In green the damped tool clamped in Hydrostatic turret, in red in VDI with screws in blue in hydrostatic VE turret and in purple in VDI adapter for hydrostatic clamp. Right: enlarged view of the initial part of the diagram.
____________________________________________________________________ 35 Figure 3‐32 Modal analysis measuring configuration. In orange the accelerometers and in
green the excitation point. ______________________________________________ 36 Figure 3‐33 Modal analysis result: compliance. In red the parting‐off v1 clamped on the
hanging hydrostatic VE turret, in green the parting‐off v2 clamped on the hanging hydrostatic VEv2 turret. _________________________________________________ 37 Figure 3‐34 Comparison between the shaft sections of the two parting‐off tools (v1 on the
left and v2 on the right). ________________________________________________ 37
Figure 3‐35 Modal analysis result: compliance. The parting‐off v2 tool is clamped in the turret mounted in the machine tool (same scale as figure 3‐34). ________________ 38 Figure 3‐36 Simulation of the first mode shape. Point 1,2 and 3 represents the
accelerometers on the tool and 4,5 and 6 on the turret (see also figure 3‐32). _____ 38 Figure 3‐37 Simulation of the second mode shape. Point 1,2 and 3 represents the
accelerometers on the tool and 4,5 and 6 on the turret (see also figure 3‐32). _____ 38 Figure 3‐38 Simulation of the third mode shape. Point 1,2 and 3 represents the
accelerometers on the tool and 4,5 and 6 on the turret (see also figure 3‐32). _____ 39 Figure 3‐39 Machining configuration. After every groove was machined, the workpiece was
moved 14mm out of the chuck. ___________________________________________ 40 Figure 3‐40 Time record of noise generated during machining at v
c=105 m/min and feed
f=0.1mm/rev (red), f = 0.15mm/rev (green) and f=0.2 (blue). ___________________ 41 Figure 3‐41 Frequency content of the signal produced when machining at v
c= 105 m/min
f=0.1mm/rev (left), f=0.15 mm/rev (center) and f=0.2mm/rev (right). ____________ 41 Figure 3‐42 Surface produced when machining at v
c=105m/min and f=0.1mm/rev and
f=0.15mm/rev. ________________________________________________________ 42 Figure 3‐43 Time record of noise generated during machining at v
c=150 m/min and feed
f=0.1mm/rev (red), and f=0.2 (green). ______________________________________ 42 Figure 3‐44 Frequency content of the signal produced when machining at v
c=150 m/min
f=0.1mm/rev (left) and f=0.2mm/rev (right). ________________________________ 43 Figure 3‐45 Surface produced when machining at v
c=150 m/min and f=0.1mm/rev and
f=0.2mm/rev. _________________________________________________________ 43 Figure 3‐46 Time record of noise generated during machining at v
c= 200 m/min and feed
f=0.1mm/rev (red), f = 0.15mm/rev (green) and f=0.2 (blue). ___________________ 44 Figure 3‐47 Frequency content of the signal produced when machining at v
c= 200 m/min
f=0.1mm/rev (left), f=0.15mm/rev (center) and f=0.2mm/rev (right). ____________ 44 Figure 3‐48 Surface produced when machining at v
c=200 m/min and f=0.15mm/rev and
f=0.2mm/rev. _________________________________________________________ 45 Figure 3‐49 Surface produced when machining at v
c=200 m/min and f=0.1mm/rev. ______ 45 Figure 3‐50 Effect of excessive vibration. The tool broke after barely two seconds of
machining. ____________________________________________________________ 45 Figure 3‐51 Even when using the tailstock the conventional parting‐off tool did not manage
to produce a smooth surface. _____________________________________________ 46 Figure 3‐52 Time record of noise generated during machining at v
c= 105 m/min and feed
f=0.1mm/rev. In red the conventional tool and in green the parting‐off v2 tool. ____ 46 Figure 3‐53 Modal analysis measuring configuration. In orange the accelerometers and in
green the excitation point. _______________________________________________ 47 Figure 3‐54 Modal analysis result: compliance. In green the standard turret, in blue the
hydrostatic VE and in purple the hydrostatic VEv2 turret. ______________________ 48 Figure 4‐1 Simple representation of the VDI adapter for hydrostatic clamp (left) and the
hydrostatic turret (right). The VDI adapter is composed of joints, one that couples the
adapter with the tool and one with the turret. The clamping system on the
XI
conventional turret for fastening the VDI adapter is as well composed of further joints, while the hydrostatic turret is simply one joint. ________________________ 50 Figure 4‐2 Influence of joints on damping. The damping ratio increases with the number of
joints (10). ____________________________________________________________ 50
1
1 Introduction
1.1 Background
In later years environmental and productivity issues have been demanding more and more from manufacturing industries. Stricter rules on emissions from combustion engines for example have caused the leading vehicle manufacturing industries to invest in research for new materials for their engines. These materials are usually harder to machine. In order to save energy by skipping the heat treatment, a workshop could decide to work on already hardened steel instead. These materials are already on the market and they are machinable, according to their manufacturers. To machine a hardened tooling steel such as Toolox® 44 (~45HRC) is possible but it requires some precautions. This kind of steel does not require heat treatment and this ensures energy and investment savings for the workshop, and, of course, it minimizes the risks for deformation and subsequent correction of the workpiece. This does not come without a price to pay though. The harder the material the more difficult problem arise in machining with respect to dynamic instability, this means that cutting parameters must be reduced. Due to the high grade of hardness it would be almost impossible to raise the cutting speed enough to get rid of the instability, the insert would wear very fast with an increase of cutting speed. Therefore a different approach has to be taken in consideration. The machine tool elastic structure becomes important, in all its parts: insert, tool holder, turret, bed, workpiece fixture and workpiece. If all the parts are optimized, a higher level of productivity can be achieved. Therefore nearly every tool supplier has in its product range a “vibration damping” or “silent” tool holders. Usually these tools are a lot more expensive than the conventional ones, but the cost is easily compensated (and justified) by being able to machine at higher removal rates in stable conditions, the work presented here takes in consideration not only the tool but other components within the machine tool elastic structure, such as the turret and the cutting insert.
Basically, there are two main categories of vibration damping systems: active and passive. Active damping requires sensors, processors and actuators in order to sample the vibrations frequency and amplitude and compensate for those.
Such tools need cables for data and energy transfer that can interfere with the machining process (and therefore not suitable), and usually are considerably expensive. Such an approach has been giving good results but its industrial application has not been welcomed by the end‐users due to the complexity of the hardware. The passive damping technique does not need any complicated hardware; the end‐user does not need to change her/his routines too much.
Implementations of passive damping in tooling equipment are already available
2
on the market and have been welcomed by the user community although the considerably higher price compared to conventional tools.
In this thesis are proposed several designs of machine tool components based on passive damping technique.
1.2 Thesis structure
This thesis is based on the Dampcomat
1project results. The scope of the project was to develop machine tool components for vibration control in machining operations (see figure 1‐1).
Figure 1‐1 Machine tool components taken in consideration in the Dampcomat project.
This work presents the thought solutions for boring bar, grooving and parting‐
off tool, turret, tool adapter (tool clamping device) and cutting tool (insert).
Three papers have been produced based on this research, they are presented at the end of this thesis. The first two papers are essentially excerpts from this thesis, where the design and validation process for the boring bar and the turret are described. The third paper describes how the right insert substrate and coating for machining hardened steel was chosen among a selection of specially designed inserts.
After this introductory chapter this thesis is divided in three more chapters.
Chapter 2 presents the state of the art in vibration control as well as detailed description of the tooling system design proposed and tested in this work.
Chapter 3 describes the different tests and their results and chapter 4 summarizes this research discussing the developments achieved and the future steps of this research.
1
Dampcomat project (project nr. 25953-1) within the EUREKA program.
3
2 Design of damped components
2.1 State of the art
As stated above, two are the main categories of vibration damping systems:
active and passive. The following chapters present how these techniques have been applied in the tooling field.
2.1.1 Active damping
The principle of active damping is to analyze in real time the signal emitted during machining, recognize instability (chatter) and compensate for it. For this purpose different techniques can be used. One way is to change the cutting speed (spindle rotation) when chatter arises. A first strategy for automatic chatter recognition and online modification of the cutting speed to a stable area was proposed by Weck et al (1). This idea has been further developed by Tlusty et al. (2), (3), where the signal emitted by the cutting process is sensed and used to recognize chatter, the cutting speed is modified thereafter. Another approach for active damping is to compensate in real time for the dynamic forces that arise during the cutting process. Harms et al. (4) suggest a tool design with piezoelectric actuators and force sensors with interchangeable tool head, see figure 2‐1.
Figure 2‐1 Solid model view of the active tool adaptor (4).
4
Browning et al (5) propose an active control system for boring bars using accelerometers on the tool for providing the controller with both reference and error signal. The signals are processed and sent eventually to the actuators located in the tool clamp, which compensates by providing dynamic forces to the boring bar. The clear advantage of the active damping approach is the perfect adaptability; all the above mentioned techniques are based on online adaptation to the ongoing process to ensure stability. The downside of this approach is the required calculation power and hardware: the system has to process the acquired signal for chatter recognition in real time, and the amount of data can be massive.
2.1.2 Passive damping
The principle of passive damping is to enhance the damping ability of the tool without actively compensating for the upcoming vibrations. A common way to achieve passive damping is by using viscoelastic materials to dissipate the energy that causes vibration. The use of viscoelastic materials for damping purposes is quite common, this technique has been used in other fields of application, such as automotive and aeronautics (6). In fact it is possible to acquire already prepared damping foils of different shapes and thicknesses on the market, such is the popularity of these materials. Viscoelastic materials are used for damping enhancement generically in three different ways: as free‐layer dampers (FLD), as constrained‐layer dampers (CLD) and in tuned viscoelastic dampers (TVD) (6). An example of TVD are the dynamic vibration absorbers (DVA) with inertia mass. The basic principle of this technique is to add a mass residing on a spring and a viscous damper at the point of maximum displacement. The viscous damper is usually implemented using viscoelastic materials. This additional single degree of freedom (SDOF) system must have approximately the same natural frequency of the boring bar in order to obtain large relative displacements, and if the viscous damper is properly designed it will dissipate the mechanical energy (7). A solution in this direction is proposed by Rivin et al. (8) where the inertia weight is integrated in the tool, hanging on rubber rings. The absorber is tuned by changing the stiffness of the additional system. Another example of application of DVA principle is proposed by Lee et al. (9), the DVA is, in this work, tuned by changing the inertia mass. This technique has been presented even for milling operations by Rashid (10). DVA technique is already successfully used in several successful commercial products such as Sandvik Silent tool product range (see figure 2‐2).
Design of damped components
5
Figure 2‐2 Sandvik Coromant Silent tool boring bar. The pre‐tuning system of the tuned bar consists mainly of a heavy tuning body (A) with a certain inertia mass, suspended in two rubber bushes (B), one at each end of the tuning body. The tuning body is surrounded by a special oily liquid (C). If vibration tendencies should arise during the machining process using a tuned bar, the damping system will immediately come into force, and the movement‐energy of the bar will be absorbed into the tuning system
2.
Rashid (10) presents as well a solution with integrated damping interface applied to workholding systems for milling operations. When implementing such a solution it is of vital importance for the success of the design to properly locate the viscoelastic layer in the structure since viscoelastic materials give the best results through shear deformation (11).
2.2 Concept
Vibration in machining is a problem for all sorts of processes but the most exposed are certainly internal turning and grooving operations. Internal turning is very sensitive to vibration due to the geometry of the boring bar, which behaves just like a cantilever beam. The deflection of a cantilever beam is strongly dependent on the overhang and the cross‐section as shown in equation 1:
I E
L F
⋅
⋅
= ⋅ 3
δ 3 eq.1
Where F is the load, L the overhang, E the Young module and I the moment of inertia. When expressing the moment of inertia for a round section beam (as the common boring bars have) the equation becomes:
3 4
3
3 64 3
64 ⎟
⎠
⎜ ⎞
⎝
⋅ ⎛
⋅
⋅
⋅
= ⋅
⋅
⋅
⋅
⋅
= ⋅
d L d E
F d
E L F
π π
δ eq.2
2
http://www2.coromant.sandvik.com/coromant/pdf/Silent_tools/Silent_tools.pdf.
6
Where d is the beam section diameter. Equation 2 puts in evidence the importance of the ratio between the overhang and the section diameter. If, for example, a boring bar with section diameter of 25 mm is clamped with an overhang of 120 mm and another one with 132 mm overhang, the ratio will be 4.80 for the first one and 5.28 for the second. If the deflection is calculated for both cases according to equation 2, considering same load and Young module, the result is:
111 80
4 3 ≈ ⋅
⋅
= K . K
δ
147 28
5 3 ≈ ⋅
⋅
= K . K
δ
Where
d E K F
⋅
⋅
⋅
= ⋅ π 3
64
The deflection calculated for an overhang of 120 mm is nearly the 75% of the one calculated for 132 mm overhang. This means the vibration amplitude can be drastically decreased by minimizing the overhang. Practically this can be done by using hydrostatic clamp instead of the conventional screw clamp (see figure 2‐3).
Figure 2‐3 Hydrostatic clamp(left) and screw clamp (right).
It is in fact common knowledge that the overhang has to be measured from the outmost fixed point, that is the first screw on the conventional VDI adapter with screws, and the outmost face on the hydrostatic clamp (see figure 2‐4).
Design of damped components
7
This solution helps indeed to minimize vibration amplitude but it does not create any vibration dissipation; what is achieved with such clamping technique is an enhancement of the static stiffness. This means that it is more difficult to excite the system tool‐clamp at its natural frequency but when that is done the system will not oppose resistance. Therefore dissipation of the mechanical energy deriving from vibration is necessary as well. To achieve this, there are several techniques, as mentioned in the previous chapters, and the choice fell on implementing viscoelastic material layers in the tool. The reason for this choice is due to two main aspects:
1. Tool production cost.
2. End‐users reluctance in changing their routines
The effectiveness of viscoelastic material for vibration damping purposes has been proved already in several fields of application (see section 2.1.2) and there is a wide range of products already available on the market at reasonable prices.
The integration of the material in the tool, as it will be explained in the next section, does not necessarily imply high production cost such as active damping or DVA solutions. For what it concerns the end‐user attitude to new routines, the use of passive damping technique insures that, first of all, the new tool does not differs too much from the traditional one in terms of look, and secondly that there will not be any additional hardware (cables and computers) and software that an actively damped tool would need.
For what it concerns parting‐off operation the problem lays both in the geometry of the tool and the peculiar disposition of the cutting force. The parting‐off tool has a slender protrusion, like a blade, where the cutting edge resides, see figure 2‐5. The longer and thinner this protrusion the easier is to get unstable cutting conditions. The dimension of the tool is though constrained by the feature it has to create on the workpiece, this means that most of the time the tool geometry is not negotiable.
Figure 2‐4 Comparison between screw (above) and hydrostatic clamp (below). The real overhang for the tool clamped with screws (above) is longer than the measured one, usually 12 mm longer, that is the distance between the outer face of the holder and the first screw.
Real overhang
Real overhang Measured overhang
Measured overhang
Screws
8
Figure 2‐5 Parting‐off tool.
The other major issue in parting‐off or grooving operations is the disposition of the cutting forces. The feed direction is radial, therefore the longitudinal component of the force is null and the radial is very high, while in longitudinal turning the radial component is almost negligible if compared to the longitudinal one (feed force). This means that vibration can be due to the deflection workpiece as well if the cutting operation is carried out far away from the spindle, see figure 2‐6.
Figure 2‐6 Deflection of the workpiece due to the radial character of the cutting force and the distance of the cutting area from the chuck.
Even in the case of parting‐off operations it is possible to minimize vibration by
using proper clamping systems and by implementing passive damping
enhancement as for internal turning. The key is to dissipate the mechanical
energy as near as possible the source, therefore the hydrostatic clamping system
is most suitable since it allows to shorten the overhang of the tool and place the
damper as near as possible the tool blade (see section 2.3.2).
Design of damped components
9
A general issue to think about when designing dampers with viscoelastic material is how the mechanical energy flows through the structure. It is important that most of the energy (if not all) flows through the damper. If the energy has an alternative way of propagation than the damper, the energy will by‐pass the damper for the easier way (12). For this reason the damping interface should not be by‐passed, see figure 2‐7, .
Tool
Clamp
Damping material
Metal supports
Tool
Clamp
Damping material
Metal supports
Energy propagation
path
Energy propagation
path Insert
Insert
Figure 2‐7 Energy propagation paths. If there is a metal‐to‐metal contact in the structure the energy will prefer to follow that path instead of going through the damping material.
The use of hydrostatic clamps helps in this matter as well, since the contact surface is much larger than the one obtained with the conventional screw clamps, and the clamping force is evenly distributed over the whole surface. This makes the energy spread through the whole damping material homogeneously, while when using screw clamps the energy would concentrate in a small contact area.
2.3 Design details
2.3.1 Boring bar
The boring bar used in this work has a diameter 25 mm with milled faces for
clamping on common VDI adapter with screws. The damped tool has been
produced by adding damping rings especially formed to avoid any possible
rotation around the tool (see figure 2‐8). These rings are made of composite
material composed of an aluminum foil 0.26 mm thick and viscoelastic material
0.12 mm thick. Their external diameter is 40 mm.
10
Figure 2‐8 Damping ring, front view (right) and axonometric view (left).
Three protective steel rings have been shaped likewise. Their thickness is of 5 mm and the external diameter is 38 mm. The function of these rings is to avoid possible bending of the damping rings. Two of the protective rings are positioned at the extremities of the damping structure while the third is in the middle (see figure 2‐11). The damping structure has been glued on the tool, where small channels have been produced in order to let glue cover the area homogeneously (see figure 2‐9).
Figure 2‐9 Channels for the glue flow.
To protect the damping structure from the coolant, a specially shaped ring has been screwed at the bottom where channels have been milled to let the coolant flow through the tool and not around it (see figure 2‐10). The screw is provided with a coolant channel as well.
Figure 2‐10 Specially shaped ring for allowing the coolant flow through the tool.
Design of damped components
11 A collet with an external diameter of 42 mm is glued on top of the whole structure. This design allows the clamping to be completely isolated from the tool, i.e. all the energy produced during the cutting process propagates through the damping material (12) (see figure 2‐11). The complete tool is shown in figure 2‐12
Figure 2‐11 Section of the damping structure. In yellow the tool body, in green the damping material and in blue the protective discs. In the detail bubble is shown how the protective disc does not touch the collet.
Figure 2‐12 CAD view of the boring bar.
12
2.3.2 Partingoff and grooving tool
To design the parting‐off tool the same technique used for the design of the boring bar has been used. The tool has been produced with a shaft shaped like the boring bar (see previous chapter for more details) instead of the traditional square one, in order to be able to glue the damping structure on it, see figure 2‐
13.
Figure 2‐13 CAD model of the parting‐off tool. (a) Exploded view. (b) view of the assembled tool. (c) View of the back of the tool.
This solution is designed for being clamped in a diameter 42 mm hydrostatic clamp. A version for diameter 32 mm clamp has been produced previously, the tool shaft was reduced from 25 mm to 16 mm diameter, and the damping rings were shaped likewise but with reduced dimensions. In this work the tool will be addressed as it follows:
• Parting‐off v1, the one for 32 mm diameter clamp.
• Parting‐off v2, the one for 42 mm diameter clamp.
2.3.3 Turret
An important, but often forgotten, part of the machine tool system is the turret, where the different tools are clamped. The turret has to be able to rotate around its center for tool change operation without interfering with the machining operations, and therefore special consideration must be regarded for the design.
Design of damped components
13 A special turret has been designed and manufactured using composite material for vibration damping and cast‐iron for the body. The choice of the material for the turret is for exploring the possibility of using the slightly higher loss factor of cast‐iron. For what it concerns the tool‐holding system, the turret has been designed for use with Hydrofix® holders (hydrostatic holders) in different sizes (see figure 2‐14). Hydrostatic clamping system ensures a uniformly distributed force over the contact surface, as it is suggested by tooling companies as first measure of vibration control.
Figure 2‐14 On the left a standard turret with VDI system. On the right the new turret, front view. The available tool‐holder sizes are: two ∅ 42, one ∅ 12, one ∅ 16, one ∅ 25 and one ∅ 32 mm.
The composite material layer is located on the interface between the turret and the machine‐tool. This design guarantees the absence of by‐passes between the turret and the machine‐tool, in order to maximize the effect of the composite material layers.
The most evident advantage of using the new turret is the reduced number of components, minimizing the possibility positioning error. This gives as well an increased working space in the machine, since no adapters are required. Two versions of this turret have been tested in machining, one with integrated composite layers and one without. To avoid misunderstanding, the turrets will be called as follows:
• Hydrostatic VE turret, the one with viscoelastic material layers
• Hydrostatic turret, the one without viscoelastic material layers.
The Hydrostatic VE turret has been further improved by manufacturing the
turret’s core in aluminum and adding more composite foil on the interface
between the core and the turret. This turret will be addressed as Hydrostatic
VEv2.
15
3 Performance evaluation
3.1 Boring bar
3.1.1 Modal Analysis Experimental Setup
The modal analysis has been carried out in three different setups: the tools clamped in the standard turret through a conventional VDI adapter (with screws), through a Hydrostatic clamp VDI adapter and in the hydrostatic turret.
Clamped boring bar in conventional VDIadapter with screws
The tools have been clamped in the turret using a VDI adapter with screws. The specification for the holders are the following:
• VDI
o ∅25 mm for conventional o ∅42 mm for the damped tool
The screws where tightened with 20 Nm torque each, in the order shown in figure 3‐1. In both cases the measuring configuration, that is the hitting point and the accelerometer position, was the one showed in figure 3‐2.
T U R R E Tool
2 1 3
T U R R E T
Figure 3‐1 Tightening order.
16
Figure 3‐2 Measuring configuration. The figure shows the accelerometers position and the excitation point.
In both cases the overhang was 123 mm.
Clamped tools in hydrostatic holder (on standard turret)
The tools have been clamped in VDI adapter for hydrostatic holder type Hydrofix NBE 42/50‐62. This holder is designed for ∅42 mm bars, therefore the conventional tool has been clamped with a collet (∅25 to ∅42 mm). The measuring configuration and the overhang was the same as previously.
Clamped tools in Hydrostatic and Hydrostatic VE turret
The tools have been clamped in hydrostatic holder type Hydrofix NBC 42/50‐62.
This holder is designed for ∅42 mm bars, therefore the conventional tool has
been clamped with a collet (∅25 to ∅42 mm). The measuring configuration and
the overhang was the same as previously.
Performance evaluation
17 Results
The conventional and the damped tool show almost the same natural frequency when clamped on conventional VDI adapter with screws. The conventional tool has a higher compliance magnitude (lower dynamic stiffness). The result is though not satisfactory enough. In fact, the light green curve (damped tool in figure 3‐3) has a sharp peak as the dark green one (conventional tool), this means that there is not much difference in damping ratio.
Figure 3‐3 Modal analysis result. The tools were clamped on the conventional turret with a VDI adapter with screws.
The most interesting result comes from the comparison of the different clamping
systems. Figure 3‐4 shows a comparison of the two tools in the different holders
and turrets.
18
Figure 3‐4 Result of the modal analysis: compliance. In dark green the damped tool in VDI adapter for Hydrostatic clamp. In black the damped tool in the Hydrostatic turret. In pink the damped tool in VDI with screws. In red the conventional tool in VDI adapter for hydrostatic clamp. In light green the conventional tool in hydrostatic turret.
The damped tool has the highest dynamic stiffness and damping ratio (the curve is round at the top) when clamped in a hydrostatic clamp, it does not make much difference if it is on the conventional or the hydrostatic turret. The conventional tool improves its dynamic stiffness (but not the damping ratio) when clamped on VDI adapter for hydrostatic clamp, but when mounted on the Hydrostatic turret the situation changes for the worse.
Conventional tool in hydrostatic turet
Conventional tool in VDI adpter for hydrostatic clamp
Conventional tool in VDI
Damped tool in VDI
Damped tool in VDI adapter for hydrostatic clamp
Damped tool in hydrostatic
turret
Performance evaluation
19
3.1.2 Machining tests Experimental Setup
The objective of these tests is to compare not only the two tools but how the tools’ behavior changes when the different clamping systems are used and if the results reflect the ones obtained in modal analysis. The tools have been tested clamped in the same configurations as they underwent the modal analysis. The workpiece was a cylinder of diameter 150 mm and height 170 mm. The material was TOOLOX® 44 (see following section). The operation was internal turning and the starting inner diameter was 48 mm. The tests were carried out at different cutting parameters, as shown in table 1.
Table 1 Cutting parameters.
Cutting speed (v
c)
[m/min] Feed (f) [mm/rev] Depth of cut (a
p) [mm]
40 – 60 – 80 0.15 0.5
40 – 60 – 80 0.3 0.5
40 – 60 – 80 0.15 1
40 – 60 – 80 0.3 1
40 – 60 – 80 0.15 1.5
40 – 60 – 80 0.15 2
3In the machining tests the sound emitted during the process was recorded and processed using LMS Test.Lab
®. Sound is easy to work with, since the microphone has to be positioned away from the sound source (the cutting process area) and cables are not needed in delicate areas. The cutting speed was increased while recording from 40 to 80 m/min, in order to fully use the available test material.
Material – Toolox ®44
Toolox is a pre‐hardened tool steel with a hardness between 410 and 475 HBW (corresponds approximately to 41 – 47 HRC)
4.The chemical composition is shown in table 2 and table 3 shows the mechanical properties.
3
Not always possible
4 Data from http://www.toolox.com/Pages/English/Main%20Frame_eng.htm
20
Table 2 Chemical composition of TOOLOX 44 (typical values)
5.
Element Quantity
C 0.31%
Si 0.60%
Mn 0.90%
P, max 100 ppm
S, max 40 ppm
Cr 1.35%
Ni 0.70%
Mo 0.80%
V 0.145%
CEV (IIW) 0.96
CET 0.57
Table 3 Mechanical properties of TOOLOX 44 (typical values).
5Temp
°C Rm
[MPa] Rp0,2
[MPa] Elongation A5
[%] Impact
toughness [J]
20 1450 1300 13 30
200 1380 1200 10
TOOLOX 44 has a compressive strength (R
c0,2) of about 1040 MPa after the material has been kept at 400°C for 175 minutes and impact toughness of 85 J when the has been heated up at 400° C
6.
Results
Two kinds of diagram are used to show the results, a time record diagram and a waterfall diagram. The time record shows the signal amplitude in time and the waterfall shows the frequency content of the signal and how does it changes (in terms of amplitude and frequency) with time. The recording was made at three different cutting speed settings; therefore the diagrams have to be interpreted as follows (see figures 3‐5 and 3‐6).
5
Data from http://www.toolox.com/PDF/English/ENG_Anvdndning_T44.pdf
6