- Email: [email protected]
In situ investigation of production processes in a large chamber scanning electron microscope
Accepted Manuscript
In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope ¨ , S. Wiesner , A. Aretz , L. Ehle , A. Haeusler , K. Bobzin , M. Ote A. Schmidt , A. Gillner , R. Poprawe , J. Mayer PII: DOI: Reference:
S0304-3991(17)30420-5 10.1016/j.ultramic.2018.07.002 ULTRAM 12608
To appear in:
Ultramicroscopy
Received date: Revised date: Accepted date:
9 October 2017 27 June 2018 3 July 2018
¨ , S. Wiesner , Please cite this article as: A. Aretz , L. Ehle , A. Haeusler , K. Bobzin , M. Ote A. Schmidt , A. Gillner , R. Poprawe , J. Mayer , In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope, Ultramicroscopy (2018), doi: 10.1016/j.ultramic.2018.07.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Highlights:
This paper describes the development of in situ test equipment for the study of fundamental processes in production engineering.
An in situ turning device was developed, tested and used to observe the chip formation on the microstructure scale of a steel sample. Laser beam micro welding was integrated into the LC-SEM to achieve in situ analysis of the welding process of stainless steel.
CR IP T
A heating module was employed for in situ wetting experiments to observe the formation and solidification of the melt of a tin-copper brazing filler on an aluminium
AC
CE
PT
ED
M
AN US
cast alloy.
ACCEPTED MANUSCRIPT
In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope A. Aretz1, L. Ehle1, A. Haeusler2, K.Bobzin3, M. Öte3, S. Wiesner3, A. Schmidt3, A. Gillner2, R. Poprawe2, J. Mayer1 1
CR IP T
Central Facility for Electron Microscopy, RWTH Aachen University, Ahornstr. 55, 52074 Aachen, Germany 2 Chair for Laser Technology LLT, RWTH Aachen University, Steinbachstr. 15, 52074 Aachen, Germany 3 Surface Engineering Institute, RWTH Aachen University, Kackertstr. 15, 52072 Aachen, Germany
Abstract
AN US
A large-chamber scanning electron microscope (LC-SEM) provides an ideal platform for the installation of large-scale in situ experiments. Our LC-SEM has internal chamber dimensions of 1,2 x 1,3 x 1,4 m3 ( W x H x D) (Fig.1) and makes it possible to incorporate novel in situ experimental devices, which are reported on here. The present manuscript describes in detail
M
the development of in situ test equipment for the study of a broad range of processes in
ED
production engineering. Direct observation of the materials modification mechanisms provides fundamental insight into the underlying process characteristics. An in situ turning
PT
device was developed, tested and used to observe the chip formation on the microstructure scale of a 43CrMo4-sample. Laser beam micro welding was integrated into the LC-SEM to
CE
achieve in situ analysis of the welding process on stainless steel 1.4310. A heating module
AC
was employed for in situ wetting experiments to observe the formation and solidification of the melt of a tin-copper brazing filler on an aluminium cast alloy.
Keywords:
Large Chamber-SEM (LC-SEM), in situ devices, turning device, laser beam micro welding, heating module, wetting experiments
1. Introduction
ACCEPTED MANUSCRIPT In many technological areas, scanning electron microscopes (SEMs) have become indispensable tools. Due to the limited size of conventional units, however, only small parts of the actual sample may be inspected. In October 2005, a large-chamber scanning electron microscope (LC-SEM) was installed at the Central Facility of Electron Microscopy (GFE) at the RWTH Aachen University. With the large chamber scanning electron microscope (LC-
CR IP T
SEM) this size restriction no longer applies. A completely new approach has been adopted, in which the SEM column itself is suspended and can move around the object [Kle97; Kle08; Kle10]. The object itself can thus have dimensions of up to 70 cm in diameter and 300 kg in weight. As an example, we have investigated a diverter element, which is in construction for
AN US
the ITER fusion reactor project, in our large chamber SEM without having to section it [May07]. Heidenblut et al. described in their work the wear characterization of forging dies using the LC-SEM, as well as investigations regarding the size influences during soldering
M
with an additional electron beam integrated in the LC-SEM [Hei07]. In situ fatigue tests in an LC-SEM, which was equipped with an integrated servo-hydraulic testing device, were
ED
reported by [Nol04] and [Kre13]. Luo et al reported on the setup of a pin-on-disc tribometer, which was used to perform in situ wear tests in the LC-SEM [Luo14]. Tillman et al used their
CE
900°C [Til15].
PT
LC-SEM for analyzing the spreading kinetics of AgCuTi melts on silicon carbide below
AC
2. Large Chamber-SEM
AN US
CR IP T
ACCEPTED MANUSCRIPT
Fig.1: LC- SEM at Central Facility for Electron Microscopy (GFE); inner dimensions of the vacuum chamber: 1,2m (w) x 1,3m (h) x 1,4m (d)
The Large Chamber SEM was designed and developed as a commercial instrument by the company Visitec (Fig.1). The instrument has been developed as a suitable tool for the study of
M
large technological parts without having to section them, but also allows a visual control of
ED
any micromechanical manufacturing, assembling, or testing process by implementing them in the chamber. The electron optics is based on an electron optical column for a conventional
PT
SEM equipped with a thermial emitter and an image resolution of 10 nm. As conventional SEM columns are not built to be operated within the vacuum chamber, modifications had to
CE
be developed for the adaption of the electron optics in the vacuum chamber. The SEM column
AC
itself has been integrated into a positioning system, [Kle97] that allows to “walk” around the sample and to image it from freely selectable viewing directions (Fig.2) [Kle08; Kle10].
Fig.2: Schematic drawing of the positioning system.
CR IP T
ACCEPTED MANUSCRIPT
AN US
The positioning system makes it possible to freely position the electron optical columns within the chamber. Two translatory axes and two rotatory axes are moving the electron optics. In addition to the mechanical positioning system, an electronic beam shift allows a
M
movement of the SEM probe of about ± 2 mm. Furthermore, a rotating device for the channeltron secondary electron detector at the objective lens of the electron optics is installed
ED
for contrast tuning in the SE mode. The positioning system enables the user to investigate objects up to a size of 70 cm in diameter and 300 kg in weight. The rectangular vacuum
PT
chamber is made of aluminium with an overall wall-thickness of 40 mm, due to the
CE
magnetically shielding properties of aluminium. In addition, the inner wall of the chamber is plated with a special nickel alloy as a further magnetic screen. The chamber is equipped with
AC
a large front door, opening completely. This allows a comfortable access to the working space and an easy installation of objects to be investigated. As we will show in the present manuscript, integration of various testing and analyzing devices is possible because of the large vacuum chamber. The LC-SEM at the GFE has a combination of analytical capabilities and accessories, which consist of the standard secondary electrons (SE) and backscattered electrons (BSE) detectors, and a liquid nitrogen free energy dispersive X-ray (EDX) detector. In addition, the chamber
ACCEPTED MANUSCRIPT can be operated in variable pressure mode in low vacuum conditions of up to p = 30 mbar. Due to the large chamber and the flexible positioning of the electron-optical column, the LCSEM offers one-of-a-kind conditions for constructing devices for in situ investigations in the chamber. Figure 1 shows a view into the chamber and illustrates the available space for experimental installations. This space gives the LC-SEM unique capabilities in various
CR IP T
research projects [Ram14]. The projects we will report on include the development and installation of in situ test equipment for turning, laser beam micro welding and the integration of a heating module for carrying out brazing experiments. The in situ devices will be
AN US
described in detail and preliminary experimental results will be presented.
3. In situ Turning Device
Functionally relevant geometric component properties, i.e. dimension, shape and position
M
tolerances, can now be produced reliably within predefined production tolerances. State variables, e.g. the inherent stress state of the component edge zone, however, cannot be set in
ED
a reproducible manner during the production process [Jaw11]. Material properties of great relevance for the later functionality of components are, among other things, the hardness, the
PT
inherent stress, the plastic deformation and the surface roughness.
CE
In the case of components produced by machining technologies (turning, milling, etc.), various effects occur in the contact zone [Gha02], [Hos12] [Liu17]. These in turn influence
AC
the properties of the produced component. Such effects and surface modifications are investigated in the field of our research [Ehl16, Hei16]. The term "process signatures" describes the relationships between material stress and the resulting modification of the material near the contact zone. The aim of this research area is the evaluation of these process signatures, so that specific modifications and state variables can be generated reproducibly in the contact zone. Consequently these potentials can be utilized for tailoring materials properties in a surface-near zone.
ACCEPTED MANUSCRIPT Within the scope of this field of research are in situ investigations in the LC-SEM during simulated material processing processes, e.g. material removal with a defined cutting edge. For these studies, a special, high vacuum-capable turning device was designed and produced. Using this turning device, a metallic sample can be turned on its surface with a precision in the range of a few µm and velocities between 10µm/s and 100µm/s. The resulting turn track,
CR IP T
the deformation on the surface, and the resulting chip are examined in situ with the LC-SEM. Deformations with a chip width of approximately 20 μm and a chip thickness of less than 3 μm can be analyzed. This allows, on the one hand, the in situ investigation of the deformation
AN US
on the sample surface plane; on the other hand, it is possible to consider the deformation of individual grains. Furthermore the surfaces produced can be subsequently characterized by conventional microscopic methods such as Atomic Force Microscopy (AFM), Transmission
3.1 Turning Device Setup
M
Electron Microscopy (TEM) and Scanning Electron Microscopy (SEM).
ED
The concept of the turning test relies on two linear axes which are arranged orthogonally to
AC
CE
PT
one another and a workpiece driven in rotation (Fig.3).
Fig.3: Concept of the turning device
The turning device consists of the following components: linear axes, electric motor, gear unit, turning tool, and the turned sample. The linear axis in the Y direction moves the turning
ACCEPTED MANUSCRIPT tool and the linear axis in the X direction moves the turned sample. Here, two axes are required to produce a plurality of turn tracks next to each other on a single turn sample. This eliminates the need to produce a new sample for each turning track. The electric motor displaces the turning sample into a rotary movement via the corresponding gear, with the turning tool guided into the outer surface of the turn sample by means of the linear axes.
CR IP T
Furthermore, during the assembly of the turning device, the circular running precision of the turned sample was achieved to be less than 10µm. This precision is a prerequisite for turning
PT
ED
M
AN US
in the μm range.
CE
Fig.4: Detailed image of the tool and the sample
By rotating the turning tool holder by 90 °, the planar end face of the sample can also be
AC
turned (Fig.4). Compared to the surface of the cylindrical shell, this has the advantage of facilitated preparation on a microstructure scale. A preparation on the cylindrical surface must ensure a high dimensional accuracy of the outer diameter. If the preparation was not performed in an absolutely uniform manner, this surface would begin to wobble. As a result, the diamond tip of the turning tool would be sometimes more and sometimes less engaged during the turning operation. This would result in turn tracks of different depth and width, up to the point where the track is interrupted.
ACCEPTED MANUSCRIPT
3.2 Experiments Figure 5 shows the process of turning during an in situ experiment in LC-SEM. A sample of 42CrMo4 is turned on the planar end face of the sample. The polished surface was etched
M
AN US
CR IP T
with Nital [see also Video 1].
ED
Fig.5: In situ experiment with 42CrMo4; Detector: SE, WD: 38mm, HT: 15kV
Chip formation mechanisms in machining processes have not yet been fully understood. Until
PT
now, this process could only be observed with optical methods, placing a strong limitation on the magnification. Our in situ LC-SEM investigations allow for the first time the observation
CE
of chip formation on the microstructure scale. Thus, conclusions on the deformation behavior
AC
of individual grains are made possible for the first time.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.6: Magnification of a chip
Figure 6 shows a section of the formed chip. Remains of the former microstructure are recognizable despite high deformation. The chip appears to be composed of individually strongly squeezed grains, which are stacked together under high deformation. Figure 7 shows
M
a FIB lamellae cut out from the turned track. The upper layers of the microstructure show a
ED
high deformation in the direction of the track. The deeper one looks into the material, the more uninfluenced the microstructure appears. By means of these investigations, the
PT
fundamentals for the understanding of the mechanisms which cause modifications of the surface near zone during straining of the material are created. Structural modifications such as
AC
CE
the grain size distribution and the grain deformation can be determined in situ.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.7: TEM bright field image of a FIB Lamella cut as a section parallel to the turned track
4. Laser Welding and Brazing in the LC-SEM
If a material passes into a molten phase during the process chain, such as in the case of metal
M
casting, plastic injection molding, all fusion welding processes, and thermal cutting processes, but also in thermal coating processes, the desired requirements for component precision might
ED
not be maintained. Volumetric contraction during solidification, uneven cooling due to restricted energy transport, and uncontrolled microstructure formation result in a large number
PT
of structural defects which significantly influence the precision of the component. In
CE
particular processes, in which the component geometry is produced by a molding tool and rapid casting and melting processes, cost-effective mass production is possible. Examples of
AC
detailed investigations of process-induced microstructural and property changes are given in the work of [Eng06; Oes08] in the field of laser processing. The central research questions in this research project can be formulated as follows: • Which physical effects and processes influence melting and solidification processes, and how is their interaction related to the achievable precision?
ACCEPTED MANUSCRIPT • How can these processes be selectively and predictively influenced to increase component precision, and how is component precision affected by thermal separation processes that produce the materials removal by local heating of the workpiece? The effect of removal is produced by heat conduction and phase transformation of the material - in the case of melt cutting processes, by the conversion of the volume to be
CR IP T
removed into the molten phase - and the removal at least of parts of this melt from the forming cutting gap. The removal geometry can thus not be derived directly from a welldefined cutting edge. Rather, the distribution of locally injected energy, the heat conduction,
AN US
and the melting process, as well as the processes of the partial material discharge on the one hand and the residual melt adherence and stiffening on the other, determine in particular the precision (reproducibility and regularity) and shape retention of the resulting cutting flanks. During laser beam cutting, instabilities of the laser cutting edge cause undesirable quality
M
losses in the form of removal and solidification ruptures. They can generate burrs. The
ED
analysis of the causative phenomena is hampered mainly by restricted diagnostic methods due to time and location of domains that are difficult to access. The physical mechanisms of the
PT
formation of the shape of the incision (cut) and the generation of the burr at the cutting edge,
CE
the two most important precision features of a laser cutting edge, are therefore still unclear. The formation of a molten phase during manufacturing processes activates complex
AC
thermomechanical mechanisms that have a lasting effect on the microstructure, the resulting properties, and the component geometry. This includes both the coarsening of existing phases and the conversion into new phases, as well as the formation of diffusion zones and residual stress zones. With high-resolution electron microscopy methods, the microstructures can be characterized both in the area of the solidified molten liquid phase and in the surrounding heat influence zones. A deeper understanding of the mechanisms involved during the formation and solidification of the melt is obtained by electron microscopy in situ experiments.
ACCEPTED MANUSCRIPT In the large-chamber scanning electron microscope installed at the GFE, a practical simulation of the processes during laser welding and brazing is carried out with the aid of corresponding devices, and the processes taking place during this activity are directly observed with the electron beam. The aim of these investigations is the basic understanding and a further quantitative description of the functional relationship between the determined microstructural
material combinations.
4.1 Laser Welding Experimental Setup
CR IP T
parameters and the macroscopic geometrical and mechanical properties of the investigated
AN US
By means of an electron microscopic investigation of the laser beam micro-welding process, a better understanding of the solidification processes and the associated capillary dynamics can be achieved. With the knowledge gained in this way, laser beam welding processes in
M
microtechnology can be used in a manner that is more effective and more processor-oriented, thereby increasing productivity [Sch12].
ED
In order to achieve a high resolution in situ analysis of the welding process, the laser system
PT
technology for laser beam micro welding was integrated into the LC-SEM. By means of this integration, the microstructures and the affected material areas of the dynamic and complex
CE
melting process can be visualized.
AC
The experiments were carried out using a SPI200C fibre laser (SPI Lasers UK Ltd, United Kingdom) with a central emission wavelength of 1070 nm and a maximum output power of 200 W. At the end of the fibre, a permanent collimator was mounted, which provided a parallel beam with a diameter of 5 mm. The resulting focused laser beam at the end of the optical lens system had a diameter of 10 µm to 50 µm on the sample surface.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.8 : LC-SEM with experimental setup with the green coloured pilot laser beam
To create a more flexible experimental setup, the laser beam was integrated in the large vacuum chamber by using an anti-reflective vacuum window (Fig.8). This setup allows observation of different laser processes in the vacuum chamber with a simple change of the
M
laser beam source. Inside the chamber, the laser beam is reflected by different mirrors, which
ED
compensate for the linear offset between the inlet of the laser beam and the actual processing area.
PT
The laser beam is focused by using a lens with a focal length of f = 100 mm and a further
CE
angled mirror, which reflects the converging laser beam under an angle of α = 30° to the normal of the surface of the sample. To realize a vertical welding process, the sample material
AC
is inclined at the angle α. The observing electron beam is directed in a vertical direction towards the tilted sample (Fig.9).
ACCEPTED MANUSCRIPT
elect ron gun
f ocal lens
BSE det ect or
CR IP T
a sample
AN US
Fig.9: Optical setup with a tilted sample
To realize a feed motion during the laser welding process, the sample material is mounted on a linear translation table (Steinmeyer Mechatronik GmbH, Germany), which can generate a slow movement vf << 1 mm/s. This practical implementation leads on the one hand to a fixed
M
observation point of the process on the sample; on the other hand, the slow movement is
frame rates up to 2 f/s.
PT
4.2 Experiments
ED
necessary to record a video by using the scanning electron microscope, which can realize
CE
The stainless steel 1.4310 (X10CrNi18-8) was used as sample material during the laser
AC
welding tests. First, the power parameters of this experimental setup for laser beam micro welding were determined. For this purpose the power of the laser was gradually increased from 20 W to 60 W (Fig.10).
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.10: Comparison of laser power from 20 watts to 60 watts post mortem; Detector: SE, WD: 44mm, HT: 15kV
The tests were started with a laser power of 20 watts and produced the right track in Figure 10. In 10 watt increments, the power was gradually increased from 20 watts to 60 watts. At the lower power of 20 watts and 30 watts, no cutting process has yet taken place. Only the
M
heat-affected zone, which is slightly wider for the 30 Watt track than for the 20 Watt track
ED
(5.33 μm versus 4.01 μm), is visible. The first material removal appears at a laser power of 40 watts. For this and higher laser powers the trenches generated by the laser exhibit a width
PT
ranging from 13.20 μm to 20.00 μm (Fig.10). As a result of laser reflections at the maximum power of 60 watts, disturbances occur during the image recording due to beam reflections on
CE
the surface. Figure 11 shows a post-mortem recorded laser track with a laser power of 50
AC
watts [see also Video 2].
Fig.11: Laser welding lane in situ monitored; Laser Power: 50 W
5. Brazing Investigations in the LC-SEM
CR IP T
ACCEPTED MANUSCRIPT
AN US
In joining techniques such as, for example, during brazing, molten phases are locally produced in the workpiece, forming a metallurgical bond between filler metal and base material. In the heat-affected zone, melting and solidification processes can thereby lead to
M
shrinkage, mismatch and warpage. Furthermore, various process parameters, such as heating rate and holding time, influence the formation of the joining zone and in particular the
ED
characteristic of the microstructure [Sch16]. The partly restricted possibilities to directly influence these factors and the mutual interactions of these factors reduce the precision and
PT
reproducibility of the melt-joined composites and thus the mechanical properties of the joint.
CE
As a consequence, rejects may arise or additional post-processing steps become necessary to remove defects, such as shrinkage, mismatch and warpage. Therefore, the goal of an
AC
interdisciplinary research group at the RWTH Aachen University is to achieve high level of precision in the case of joined components by controlling and influencing the precisiondetermining factors during the joining process. In a first step toward this goal, the factors that influence precision are analyzed. In addition, we aim to gain a deeper understanding of the mechanisms involved such as wetting of the base material during the melting and interactions of the filler metals and base material during solidification of the filler metals and therefore formation of a reaction zone
ACCEPTED MANUSCRIPT and of the joint by electron microscopy in situ experiments. In the LC-SEM installed at the GFE, a mechanism-oriented observation of the processes during brazing is attempted and the processes taking place during this experiment are directly mapped with the electron beam. The aim of this study is to provide a basic understanding and a further quantitative description of the thermodynamic and thermomechanical properties as well as the relationship between
CR IP T
the structures formed during the solidification.
5.1 Heating Module
A heating module from the company Kammrath & Weiss was employed for in situ wetting
PT
ED
M
AN US
experiments. Figure 12 shows the installation set-up of the heating module in the LC-SEM.
CE
Fig.12: Heating module in the LC-SEM (bottom). The electron optical column with an SE and BSE detector can be seen at the top.
AC
In this heating module, samples up to a size of 10 × 10 mm can be heated at temperatures ranging from room temperature to 1500 °-C. Samples can then be examined in situ in the scanning electron microscope. The heat transfer to the LC-SEM table is kept as low as possible. The small chamber of the heating module is equipped with a motor-driven cover, which can be closed to protect the pole piece of the column and detectors from excessive heat loads and from meld deposition from the vapour. The planar samples can be fixed with small tungsten springs.
ACCEPTED MANUSCRIPT
5.2 Experiments Within the framework of the first experiments, in situ wetting experiments of 78Sn22Cu (wt%) melt spinning braze ribbons on an aluminium cast alloy Al7Si0.3Mg (wt%) are conducted at thixotropic temperatures. Advanced information about the manufacturing
CR IP T
process and an analytical characterization of the utilized braze ribbons are published in [Bob16]. The braze ribbon is clamped on the base material with a tungsten spring of the
PT
ED
M
AN US
heating module, as shown in Figure 13, Picture 1, at the bottom left.
CE
Fig.13: Heating experiment up to 390°C for a period of 40 minutes
AC
At temperatures above 250°C, the shape of the soldering tape begins to shrink due to the high surface tension (Pictures 2, 3 in Fig.13). Therefore the melting process starts at temperatures in this range. This observation correlates with Sn-Cu thermodynamic data, Figure 14. After the shrinking process, the braze ribbon deforms more and more into a partially liquified sphere (Pictures 4, 5 in Fig.13) without wetting the base material. A probable reason for the liquid filler metal not to wet the aluminium base material is an alumina layer on top of the base material. At a temperature of 380°C, the braze ribbon suddenly begins to flow on the
ACCEPTED MANUSCRIPT substrate surface starting from the contact point of the tungsten needle (Picture 6 in Fig.13). The braze material is almost completely spread out at 390°C (Picture 7 in Fig.13) and remains virtually unchanged during the holding time of 30 minutes (Picture 8 in Fig.13). We assume that this rapid process started when the alumina layer was locally removed from the top of the
M
AN US
CR IP T
base material [see also Video 3].
Fig. 14: Computed equilibrium phase fractions for the Sn22weight-%Cu-brazing filler (database TCAl4) [Bob17]
ED
In order to get insight in the underlying mechanisms, we performed subsequent
PT
characterization of the microstructure of the braze area and the surrounding base material in
AC
CE
our LC-SEM by energy-dispersive-X-ray-spectroscopy (EDXS, Bruker GmbH).
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
PT
Fig.15: Element distribution maps for AlSi7Mg0.3 (wt%) base material and Sn78Cu22 (wt%) brazing ribbon from SEM-EDXS: (a) SE-Image, (b) aluminium, (c) silicon, (d) magnesium, (e) tin and (f) copper.
CE
Corresponding SEM-EDXS element maps from this sample (Fig.15) show a non-homogenous distribution of the elements. The base material shows aluminium as the main material, on the
AC
grain boundaries of which silicon is enriched and to a lesser extent magnesium. In the area of the braze point, as expected, no base material is to be found, only copper and tin. In addition, it can be seen clearly that tin spreads beyond the braze point along the silicon-enriched grain boundaries of the base material.
ACCEPTED MANUSCRIPT
6. Conclusions We report on the development and initial experiments with three different in situ units in our LC-SEM. A turning device setup for investigating details of the turning process of metallic samples was developed and first results were presented. The cylindrical sample can be
CR IP T
machined both on its cylindrical surface and on the planar end surface. With this equipment the deformation of the sample surface can be documented during the turning process. Furthermore post mortem examination methods using conventional SEM and TEM provide a deeper insight into the depth to which the deformation extends into the sample and how the
AN US
resulting chip is constructed. We also report on the incorporation of a laser beam into the LCSEM to investigate laser welding processes in situ. We have succeeded in producing laser traces in situ and to analyze the results of a variation of the power of the laser. Finally, we report on in situ wetting experiments in our LC-SEM. In this simulation of a brazing process,
M
it was possible for the first time to observe all stages of the melting and the wetting process of
ED
aluminium base material including the removal of alumina layer directly in the electron microscope. Subsequent analytical studies will provide an accurate insight into the
PT
composition of the individual phases as well as their local distribution.
CE
Acknowledgment:
Our large-chamber scanning electron microscope was acquired within the framework of a
AC
microscope infrastructure initiative of the German Research Foundation (DFG. We gratefully acknowledge support in the context of the Collaborative Research Centre SFB 1120 "Precision Melt Engineering” and the Collaborative Research Centre SFB TRR136 „Process Signatures“ funded by the German Research Foundation (DFG).
ACCEPTED MANUSCRIPT References: [Bob16] K. Bobzin, M. Öte, S. Wiesner, L. Pongratz, J. Mayer, A. Aretz, R. Iskandar, A. Schwedt, presented at WTK 2016, Chemnitz, Germany, March 10 – March 11, 2016, 76-87. [Bob17]
K. Bobzin, M. Öte, S. Wiesner, A. Schmidt, M. Apel, R. Berger, A. Aretz, J. Mayer, Formation of the reaction zone between Sn-Cu brazing fillers and Al-Si-Mg alloys: experiments and thermodynamic analysis, Material Science and Engineering Technology, submitted.
[Ehl16]
L. Ehle, J. Kämmler, D. Meyer, A. Schwedt, J. Mayer, 3rd CIRP Conference on Surface Integrity, Procedia CIRP 45C, 2016, pp. 367-370.
CR IP T
[Eng06] L. Engelbrecht, O. Meier, A. Ostendorf, H. Haferkamp, Materialwissenschaft und Werkstofftechnik 37 (2006) 272–27. [Gha02] F. Ghanem, C. Braham, M.E. Fitzpatrick, and H. Sidhom: Effect of Near-Surface Residual Stress and Microstructure Modification From Machining on the Fatigue Endurance of a Tool Steel. Journal of Materials Engineering and Performance 11 (2002) 631-639. Heidenblut, T.; Möhwald, K.; Deißer, T. A.; Bistron, M.; Behrens, B.-A.; Bach, Fr.-W. (2007): Wear Characterization of Forging Dies Using a Large Chamber Scanning Electron Microscope (Microscopy and Microanalysis, 13), S. 104-105.
[Hei16]
C. Heinzel, F. Borchers, D. Berger, L. Ehle, 3 rd CIRP Conference on Surface Integrity, Procedia CIRP 45C, 2016, pp. 191-194.
[Hos12]
S. B. Hosseini, U. Klement, J. Kaminski: Microstructure Characterization of White Layer formed by Hard Turning and Wire Electric Discharge Machining in High Carbon Steel (AISI 52100). Advanced Materials Research Vol. 409 (2012) 684-689.
[Jaw11]
Jawahir, I.S.; Brinksmeier, E.; M’Saoubi, R.; Aspinwall, D.K.; Outeiro, J.C.; Meyer, D.; Umbrello, D.; Jayal, A.D.: Surface integrity in material removal processes: Recent advances. Annals of the CIRP,60/2, 2011, Seite 603-626.
[Kle97]
M. Klein and S. Klein, Adapting Human Behaviour for the Development of a New Scanning Electron Microscope. Proc. Int. Conf. Micromechatronics for Information and Precision Equipment MIPE'97, Tokyo, Japan, 1997, pp. 324 - 329.
ED
M
AN US
[Hei07]
[Kle08] M. Klein, Oberflächenbearbeitungsvorrichtung, Patentschrift DE10 2006 032 303 A1 2008.01.17
M. Fried, C. Krechel, E.E. Affeldt, B. Eckert, S. Kimmig, U. Retze, H.W. Höppel and M. Göken : Fatigue crack initiation in nickel-based superalloys studied by microstructure-based FE modeling and scanning electron microscopy, MATEC Web of Conferences 14, 16001 (2014), DOI: 10.1051/matecconf/2014416001
CE
[Kre13]
PT
[Kle10] M. Klein, Oberflächenbearbeitungsvorrichtung, Patentschrift DE10 2006 032 303 B4 2010.08.19
AC
[Liu17] B. Liu, X. Zhou, X. Zhang, Orthogonal machining introduced microstructure modification in AA7150-T651 aluminium alloy, Materials Characterization 123 (2017) 91 – 98.
[Luo14] W. Luo, W. Tillmann and U. Selvadurai; In Situ Wear Test on Thermal Spray Coatings in a Large Chamber Scanning Electron Microscope, Journal of Thermal Spray Technology, Vol. 24 (1-2) January 2015, pp. 263-270 [May07] J. Mayer, J. Kallinna, P. Watermeyer, A. Aretz, W. Rehbach, J. Linke*, and A. Schmidt*, Nondestructive Analysis of Engineering Components in the Large-Chamber Scanning Electron Microscope, Proceedings Microscopy and Microanalysis 2007. [Nol04]
R. Nolte, PhD thesis, Friedrich Alexander University Erlangen-Nürnberg (2004)
[Oes08]
Österle, W. ; Krause, S. ; Moelders, T. ; Neidel, A. ; Oder, G. ; Völker, J.; Materials Characterization 59 (2008) 1564-1571.
[Ram14] A. Ramazani, A. Schwedt, A. Aretz, U. Prahl, Key Eng.Mater. 2014, 586, pp. 67-71.
ACCEPTED MANUSCRIPT Schmitt, F., “Laserstrahl-Mikroschweißen mit Strahlquellen hoher Brillanz und örtlicher Leistungsmodulation,” Ph.D. thesis, Shaker, Aachen, 2012.
[Sch16]
G.J. Schmitz, B. Böttger, M. Apel, IOP Conf Ser-Mat Sci, 2016, 117, 012041.
[Til15]
W. Tillmann, J.Pfeiffer, N.Sievers and K.Böttcher, Analyses of the spreading kinetics of AgCuTi melts on silicon carbide below 900 °C, using a large-chamber SEM, Colloids and Surfaces A: Physicochemical and Engineering Aspects 468; pp. 167--173, 2015
AC
CE
PT
ED
M
AN US
CR IP T
[Sch12]
In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope ¨ , S. Wiesner , A. Aretz , L. Ehle , A. Haeusler , K. Bobzin , M. Ote A. Schmidt , A. Gillner , R. Poprawe , J. Mayer PII: DOI: Reference:
S0304-3991(17)30420-5 10.1016/j.ultramic.2018.07.002 ULTRAM 12608
To appear in:
Ultramicroscopy
Received date: Revised date: Accepted date:
9 October 2017 27 June 2018 3 July 2018
¨ , S. Wiesner , Please cite this article as: A. Aretz , L. Ehle , A. Haeusler , K. Bobzin , M. Ote A. Schmidt , A. Gillner , R. Poprawe , J. Mayer , In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope, Ultramicroscopy (2018), doi: 10.1016/j.ultramic.2018.07.002
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT Highlights:
This paper describes the development of in situ test equipment for the study of fundamental processes in production engineering.
An in situ turning device was developed, tested and used to observe the chip formation on the microstructure scale of a steel sample. Laser beam micro welding was integrated into the LC-SEM to achieve in situ analysis of the welding process of stainless steel.
CR IP T
A heating module was employed for in situ wetting experiments to observe the formation and solidification of the melt of a tin-copper brazing filler on an aluminium
AC
CE
PT
ED
M
AN US
cast alloy.
ACCEPTED MANUSCRIPT
In Situ investigation of production processes in a Large Chamber Scanning Electron Microscope A. Aretz1, L. Ehle1, A. Haeusler2, K.Bobzin3, M. Öte3, S. Wiesner3, A. Schmidt3, A. Gillner2, R. Poprawe2, J. Mayer1 1
CR IP T
Central Facility for Electron Microscopy, RWTH Aachen University, Ahornstr. 55, 52074 Aachen, Germany 2 Chair for Laser Technology LLT, RWTH Aachen University, Steinbachstr. 15, 52074 Aachen, Germany 3 Surface Engineering Institute, RWTH Aachen University, Kackertstr. 15, 52072 Aachen, Germany
Abstract
AN US
A large-chamber scanning electron microscope (LC-SEM) provides an ideal platform for the installation of large-scale in situ experiments. Our LC-SEM has internal chamber dimensions of 1,2 x 1,3 x 1,4 m3 ( W x H x D) (Fig.1) and makes it possible to incorporate novel in situ experimental devices, which are reported on here. The present manuscript describes in detail
M
the development of in situ test equipment for the study of a broad range of processes in
ED
production engineering. Direct observation of the materials modification mechanisms provides fundamental insight into the underlying process characteristics. An in situ turning
PT
device was developed, tested and used to observe the chip formation on the microstructure scale of a 43CrMo4-sample. Laser beam micro welding was integrated into the LC-SEM to
CE
achieve in situ analysis of the welding process on stainless steel 1.4310. A heating module
AC
was employed for in situ wetting experiments to observe the formation and solidification of the melt of a tin-copper brazing filler on an aluminium cast alloy.
Keywords:
Large Chamber-SEM (LC-SEM), in situ devices, turning device, laser beam micro welding, heating module, wetting experiments
1. Introduction
ACCEPTED MANUSCRIPT In many technological areas, scanning electron microscopes (SEMs) have become indispensable tools. Due to the limited size of conventional units, however, only small parts of the actual sample may be inspected. In October 2005, a large-chamber scanning electron microscope (LC-SEM) was installed at the Central Facility of Electron Microscopy (GFE) at the RWTH Aachen University. With the large chamber scanning electron microscope (LC-
CR IP T
SEM) this size restriction no longer applies. A completely new approach has been adopted, in which the SEM column itself is suspended and can move around the object [Kle97; Kle08; Kle10]. The object itself can thus have dimensions of up to 70 cm in diameter and 300 kg in weight. As an example, we have investigated a diverter element, which is in construction for
AN US
the ITER fusion reactor project, in our large chamber SEM without having to section it [May07]. Heidenblut et al. described in their work the wear characterization of forging dies using the LC-SEM, as well as investigations regarding the size influences during soldering
M
with an additional electron beam integrated in the LC-SEM [Hei07]. In situ fatigue tests in an LC-SEM, which was equipped with an integrated servo-hydraulic testing device, were
ED
reported by [Nol04] and [Kre13]. Luo et al reported on the setup of a pin-on-disc tribometer, which was used to perform in situ wear tests in the LC-SEM [Luo14]. Tillman et al used their
CE
900°C [Til15].
PT
LC-SEM for analyzing the spreading kinetics of AgCuTi melts on silicon carbide below
AC
2. Large Chamber-SEM
AN US
CR IP T
ACCEPTED MANUSCRIPT
Fig.1: LC- SEM at Central Facility for Electron Microscopy (GFE); inner dimensions of the vacuum chamber: 1,2m (w) x 1,3m (h) x 1,4m (d)
The Large Chamber SEM was designed and developed as a commercial instrument by the company Visitec (Fig.1). The instrument has been developed as a suitable tool for the study of
M
large technological parts without having to section them, but also allows a visual control of
ED
any micromechanical manufacturing, assembling, or testing process by implementing them in the chamber. The electron optics is based on an electron optical column for a conventional
PT
SEM equipped with a thermial emitter and an image resolution of 10 nm. As conventional SEM columns are not built to be operated within the vacuum chamber, modifications had to
CE
be developed for the adaption of the electron optics in the vacuum chamber. The SEM column
AC
itself has been integrated into a positioning system, [Kle97] that allows to “walk” around the sample and to image it from freely selectable viewing directions (Fig.2) [Kle08; Kle10].
Fig.2: Schematic drawing of the positioning system.
CR IP T
ACCEPTED MANUSCRIPT
AN US
The positioning system makes it possible to freely position the electron optical columns within the chamber. Two translatory axes and two rotatory axes are moving the electron optics. In addition to the mechanical positioning system, an electronic beam shift allows a
M
movement of the SEM probe of about ± 2 mm. Furthermore, a rotating device for the channeltron secondary electron detector at the objective lens of the electron optics is installed
ED
for contrast tuning in the SE mode. The positioning system enables the user to investigate objects up to a size of 70 cm in diameter and 300 kg in weight. The rectangular vacuum
PT
chamber is made of aluminium with an overall wall-thickness of 40 mm, due to the
CE
magnetically shielding properties of aluminium. In addition, the inner wall of the chamber is plated with a special nickel alloy as a further magnetic screen. The chamber is equipped with
AC
a large front door, opening completely. This allows a comfortable access to the working space and an easy installation of objects to be investigated. As we will show in the present manuscript, integration of various testing and analyzing devices is possible because of the large vacuum chamber. The LC-SEM at the GFE has a combination of analytical capabilities and accessories, which consist of the standard secondary electrons (SE) and backscattered electrons (BSE) detectors, and a liquid nitrogen free energy dispersive X-ray (EDX) detector. In addition, the chamber
ACCEPTED MANUSCRIPT can be operated in variable pressure mode in low vacuum conditions of up to p = 30 mbar. Due to the large chamber and the flexible positioning of the electron-optical column, the LCSEM offers one-of-a-kind conditions for constructing devices for in situ investigations in the chamber. Figure 1 shows a view into the chamber and illustrates the available space for experimental installations. This space gives the LC-SEM unique capabilities in various
CR IP T
research projects [Ram14]. The projects we will report on include the development and installation of in situ test equipment for turning, laser beam micro welding and the integration of a heating module for carrying out brazing experiments. The in situ devices will be
AN US
described in detail and preliminary experimental results will be presented.
3. In situ Turning Device
Functionally relevant geometric component properties, i.e. dimension, shape and position
M
tolerances, can now be produced reliably within predefined production tolerances. State variables, e.g. the inherent stress state of the component edge zone, however, cannot be set in
ED
a reproducible manner during the production process [Jaw11]. Material properties of great relevance for the later functionality of components are, among other things, the hardness, the
PT
inherent stress, the plastic deformation and the surface roughness.
CE
In the case of components produced by machining technologies (turning, milling, etc.), various effects occur in the contact zone [Gha02], [Hos12] [Liu17]. These in turn influence
AC
the properties of the produced component. Such effects and surface modifications are investigated in the field of our research [Ehl16, Hei16]. The term "process signatures" describes the relationships between material stress and the resulting modification of the material near the contact zone. The aim of this research area is the evaluation of these process signatures, so that specific modifications and state variables can be generated reproducibly in the contact zone. Consequently these potentials can be utilized for tailoring materials properties in a surface-near zone.
ACCEPTED MANUSCRIPT Within the scope of this field of research are in situ investigations in the LC-SEM during simulated material processing processes, e.g. material removal with a defined cutting edge. For these studies, a special, high vacuum-capable turning device was designed and produced. Using this turning device, a metallic sample can be turned on its surface with a precision in the range of a few µm and velocities between 10µm/s and 100µm/s. The resulting turn track,
CR IP T
the deformation on the surface, and the resulting chip are examined in situ with the LC-SEM. Deformations with a chip width of approximately 20 μm and a chip thickness of less than 3 μm can be analyzed. This allows, on the one hand, the in situ investigation of the deformation
AN US
on the sample surface plane; on the other hand, it is possible to consider the deformation of individual grains. Furthermore the surfaces produced can be subsequently characterized by conventional microscopic methods such as Atomic Force Microscopy (AFM), Transmission
3.1 Turning Device Setup
M
Electron Microscopy (TEM) and Scanning Electron Microscopy (SEM).
ED
The concept of the turning test relies on two linear axes which are arranged orthogonally to
AC
CE
PT
one another and a workpiece driven in rotation (Fig.3).
Fig.3: Concept of the turning device
The turning device consists of the following components: linear axes, electric motor, gear unit, turning tool, and the turned sample. The linear axis in the Y direction moves the turning
ACCEPTED MANUSCRIPT tool and the linear axis in the X direction moves the turned sample. Here, two axes are required to produce a plurality of turn tracks next to each other on a single turn sample. This eliminates the need to produce a new sample for each turning track. The electric motor displaces the turning sample into a rotary movement via the corresponding gear, with the turning tool guided into the outer surface of the turn sample by means of the linear axes.
CR IP T
Furthermore, during the assembly of the turning device, the circular running precision of the turned sample was achieved to be less than 10µm. This precision is a prerequisite for turning
PT
ED
M
AN US
in the μm range.
CE
Fig.4: Detailed image of the tool and the sample
By rotating the turning tool holder by 90 °, the planar end face of the sample can also be
AC
turned (Fig.4). Compared to the surface of the cylindrical shell, this has the advantage of facilitated preparation on a microstructure scale. A preparation on the cylindrical surface must ensure a high dimensional accuracy of the outer diameter. If the preparation was not performed in an absolutely uniform manner, this surface would begin to wobble. As a result, the diamond tip of the turning tool would be sometimes more and sometimes less engaged during the turning operation. This would result in turn tracks of different depth and width, up to the point where the track is interrupted.
ACCEPTED MANUSCRIPT
3.2 Experiments Figure 5 shows the process of turning during an in situ experiment in LC-SEM. A sample of 42CrMo4 is turned on the planar end face of the sample. The polished surface was etched
M
AN US
CR IP T
with Nital [see also Video 1].
ED
Fig.5: In situ experiment with 42CrMo4; Detector: SE, WD: 38mm, HT: 15kV
Chip formation mechanisms in machining processes have not yet been fully understood. Until
PT
now, this process could only be observed with optical methods, placing a strong limitation on the magnification. Our in situ LC-SEM investigations allow for the first time the observation
CE
of chip formation on the microstructure scale. Thus, conclusions on the deformation behavior
AC
of individual grains are made possible for the first time.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.6: Magnification of a chip
Figure 6 shows a section of the formed chip. Remains of the former microstructure are recognizable despite high deformation. The chip appears to be composed of individually strongly squeezed grains, which are stacked together under high deformation. Figure 7 shows
M
a FIB lamellae cut out from the turned track. The upper layers of the microstructure show a
ED
high deformation in the direction of the track. The deeper one looks into the material, the more uninfluenced the microstructure appears. By means of these investigations, the
PT
fundamentals for the understanding of the mechanisms which cause modifications of the surface near zone during straining of the material are created. Structural modifications such as
AC
CE
the grain size distribution and the grain deformation can be determined in situ.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.7: TEM bright field image of a FIB Lamella cut as a section parallel to the turned track
4. Laser Welding and Brazing in the LC-SEM
If a material passes into a molten phase during the process chain, such as in the case of metal
M
casting, plastic injection molding, all fusion welding processes, and thermal cutting processes, but also in thermal coating processes, the desired requirements for component precision might
ED
not be maintained. Volumetric contraction during solidification, uneven cooling due to restricted energy transport, and uncontrolled microstructure formation result in a large number
PT
of structural defects which significantly influence the precision of the component. In
CE
particular processes, in which the component geometry is produced by a molding tool and rapid casting and melting processes, cost-effective mass production is possible. Examples of
AC
detailed investigations of process-induced microstructural and property changes are given in the work of [Eng06; Oes08] in the field of laser processing. The central research questions in this research project can be formulated as follows: • Which physical effects and processes influence melting and solidification processes, and how is their interaction related to the achievable precision?
ACCEPTED MANUSCRIPT • How can these processes be selectively and predictively influenced to increase component precision, and how is component precision affected by thermal separation processes that produce the materials removal by local heating of the workpiece? The effect of removal is produced by heat conduction and phase transformation of the material - in the case of melt cutting processes, by the conversion of the volume to be
CR IP T
removed into the molten phase - and the removal at least of parts of this melt from the forming cutting gap. The removal geometry can thus not be derived directly from a welldefined cutting edge. Rather, the distribution of locally injected energy, the heat conduction,
AN US
and the melting process, as well as the processes of the partial material discharge on the one hand and the residual melt adherence and stiffening on the other, determine in particular the precision (reproducibility and regularity) and shape retention of the resulting cutting flanks. During laser beam cutting, instabilities of the laser cutting edge cause undesirable quality
M
losses in the form of removal and solidification ruptures. They can generate burrs. The
ED
analysis of the causative phenomena is hampered mainly by restricted diagnostic methods due to time and location of domains that are difficult to access. The physical mechanisms of the
PT
formation of the shape of the incision (cut) and the generation of the burr at the cutting edge,
CE
the two most important precision features of a laser cutting edge, are therefore still unclear. The formation of a molten phase during manufacturing processes activates complex
AC
thermomechanical mechanisms that have a lasting effect on the microstructure, the resulting properties, and the component geometry. This includes both the coarsening of existing phases and the conversion into new phases, as well as the formation of diffusion zones and residual stress zones. With high-resolution electron microscopy methods, the microstructures can be characterized both in the area of the solidified molten liquid phase and in the surrounding heat influence zones. A deeper understanding of the mechanisms involved during the formation and solidification of the melt is obtained by electron microscopy in situ experiments.
ACCEPTED MANUSCRIPT In the large-chamber scanning electron microscope installed at the GFE, a practical simulation of the processes during laser welding and brazing is carried out with the aid of corresponding devices, and the processes taking place during this activity are directly observed with the electron beam. The aim of these investigations is the basic understanding and a further quantitative description of the functional relationship between the determined microstructural
material combinations.
4.1 Laser Welding Experimental Setup
CR IP T
parameters and the macroscopic geometrical and mechanical properties of the investigated
AN US
By means of an electron microscopic investigation of the laser beam micro-welding process, a better understanding of the solidification processes and the associated capillary dynamics can be achieved. With the knowledge gained in this way, laser beam welding processes in
M
microtechnology can be used in a manner that is more effective and more processor-oriented, thereby increasing productivity [Sch12].
ED
In order to achieve a high resolution in situ analysis of the welding process, the laser system
PT
technology for laser beam micro welding was integrated into the LC-SEM. By means of this integration, the microstructures and the affected material areas of the dynamic and complex
CE
melting process can be visualized.
AC
The experiments were carried out using a SPI200C fibre laser (SPI Lasers UK Ltd, United Kingdom) with a central emission wavelength of 1070 nm and a maximum output power of 200 W. At the end of the fibre, a permanent collimator was mounted, which provided a parallel beam with a diameter of 5 mm. The resulting focused laser beam at the end of the optical lens system had a diameter of 10 µm to 50 µm on the sample surface.
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.8 : LC-SEM with experimental setup with the green coloured pilot laser beam
To create a more flexible experimental setup, the laser beam was integrated in the large vacuum chamber by using an anti-reflective vacuum window (Fig.8). This setup allows observation of different laser processes in the vacuum chamber with a simple change of the
M
laser beam source. Inside the chamber, the laser beam is reflected by different mirrors, which
ED
compensate for the linear offset between the inlet of the laser beam and the actual processing area.
PT
The laser beam is focused by using a lens with a focal length of f = 100 mm and a further
CE
angled mirror, which reflects the converging laser beam under an angle of α = 30° to the normal of the surface of the sample. To realize a vertical welding process, the sample material
AC
is inclined at the angle α. The observing electron beam is directed in a vertical direction towards the tilted sample (Fig.9).
ACCEPTED MANUSCRIPT
elect ron gun
f ocal lens
BSE det ect or
CR IP T
a sample
AN US
Fig.9: Optical setup with a tilted sample
To realize a feed motion during the laser welding process, the sample material is mounted on a linear translation table (Steinmeyer Mechatronik GmbH, Germany), which can generate a slow movement vf << 1 mm/s. This practical implementation leads on the one hand to a fixed
M
observation point of the process on the sample; on the other hand, the slow movement is
frame rates up to 2 f/s.
PT
4.2 Experiments
ED
necessary to record a video by using the scanning electron microscope, which can realize
CE
The stainless steel 1.4310 (X10CrNi18-8) was used as sample material during the laser
AC
welding tests. First, the power parameters of this experimental setup for laser beam micro welding were determined. For this purpose the power of the laser was gradually increased from 20 W to 60 W (Fig.10).
CR IP T
ACCEPTED MANUSCRIPT
AN US
Fig.10: Comparison of laser power from 20 watts to 60 watts post mortem; Detector: SE, WD: 44mm, HT: 15kV
The tests were started with a laser power of 20 watts and produced the right track in Figure 10. In 10 watt increments, the power was gradually increased from 20 watts to 60 watts. At the lower power of 20 watts and 30 watts, no cutting process has yet taken place. Only the
M
heat-affected zone, which is slightly wider for the 30 Watt track than for the 20 Watt track
ED
(5.33 μm versus 4.01 μm), is visible. The first material removal appears at a laser power of 40 watts. For this and higher laser powers the trenches generated by the laser exhibit a width
PT
ranging from 13.20 μm to 20.00 μm (Fig.10). As a result of laser reflections at the maximum power of 60 watts, disturbances occur during the image recording due to beam reflections on
CE
the surface. Figure 11 shows a post-mortem recorded laser track with a laser power of 50
AC
watts [see also Video 2].
Fig.11: Laser welding lane in situ monitored; Laser Power: 50 W
5. Brazing Investigations in the LC-SEM
CR IP T
ACCEPTED MANUSCRIPT
AN US
In joining techniques such as, for example, during brazing, molten phases are locally produced in the workpiece, forming a metallurgical bond between filler metal and base material. In the heat-affected zone, melting and solidification processes can thereby lead to
M
shrinkage, mismatch and warpage. Furthermore, various process parameters, such as heating rate and holding time, influence the formation of the joining zone and in particular the
ED
characteristic of the microstructure [Sch16]. The partly restricted possibilities to directly influence these factors and the mutual interactions of these factors reduce the precision and
PT
reproducibility of the melt-joined composites and thus the mechanical properties of the joint.
CE
As a consequence, rejects may arise or additional post-processing steps become necessary to remove defects, such as shrinkage, mismatch and warpage. Therefore, the goal of an
AC
interdisciplinary research group at the RWTH Aachen University is to achieve high level of precision in the case of joined components by controlling and influencing the precisiondetermining factors during the joining process. In a first step toward this goal, the factors that influence precision are analyzed. In addition, we aim to gain a deeper understanding of the mechanisms involved such as wetting of the base material during the melting and interactions of the filler metals and base material during solidification of the filler metals and therefore formation of a reaction zone
ACCEPTED MANUSCRIPT and of the joint by electron microscopy in situ experiments. In the LC-SEM installed at the GFE, a mechanism-oriented observation of the processes during brazing is attempted and the processes taking place during this experiment are directly mapped with the electron beam. The aim of this study is to provide a basic understanding and a further quantitative description of the thermodynamic and thermomechanical properties as well as the relationship between
CR IP T
the structures formed during the solidification.
5.1 Heating Module
A heating module from the company Kammrath & Weiss was employed for in situ wetting
PT
ED
M
AN US
experiments. Figure 12 shows the installation set-up of the heating module in the LC-SEM.
CE
Fig.12: Heating module in the LC-SEM (bottom). The electron optical column with an SE and BSE detector can be seen at the top.
AC
In this heating module, samples up to a size of 10 × 10 mm can be heated at temperatures ranging from room temperature to 1500 °-C. Samples can then be examined in situ in the scanning electron microscope. The heat transfer to the LC-SEM table is kept as low as possible. The small chamber of the heating module is equipped with a motor-driven cover, which can be closed to protect the pole piece of the column and detectors from excessive heat loads and from meld deposition from the vapour. The planar samples can be fixed with small tungsten springs.
ACCEPTED MANUSCRIPT
5.2 Experiments Within the framework of the first experiments, in situ wetting experiments of 78Sn22Cu (wt%) melt spinning braze ribbons on an aluminium cast alloy Al7Si0.3Mg (wt%) are conducted at thixotropic temperatures. Advanced information about the manufacturing
CR IP T
process and an analytical characterization of the utilized braze ribbons are published in [Bob16]. The braze ribbon is clamped on the base material with a tungsten spring of the
PT
ED
M
AN US
heating module, as shown in Figure 13, Picture 1, at the bottom left.
CE
Fig.13: Heating experiment up to 390°C for a period of 40 minutes
AC
At temperatures above 250°C, the shape of the soldering tape begins to shrink due to the high surface tension (Pictures 2, 3 in Fig.13). Therefore the melting process starts at temperatures in this range. This observation correlates with Sn-Cu thermodynamic data, Figure 14. After the shrinking process, the braze ribbon deforms more and more into a partially liquified sphere (Pictures 4, 5 in Fig.13) without wetting the base material. A probable reason for the liquid filler metal not to wet the aluminium base material is an alumina layer on top of the base material. At a temperature of 380°C, the braze ribbon suddenly begins to flow on the
ACCEPTED MANUSCRIPT substrate surface starting from the contact point of the tungsten needle (Picture 6 in Fig.13). The braze material is almost completely spread out at 390°C (Picture 7 in Fig.13) and remains virtually unchanged during the holding time of 30 minutes (Picture 8 in Fig.13). We assume that this rapid process started when the alumina layer was locally removed from the top of the
M
AN US
CR IP T
base material [see also Video 3].
Fig. 14: Computed equilibrium phase fractions for the Sn22weight-%Cu-brazing filler (database TCAl4) [Bob17]
ED
In order to get insight in the underlying mechanisms, we performed subsequent
PT
characterization of the microstructure of the braze area and the surrounding base material in
AC
CE
our LC-SEM by energy-dispersive-X-ray-spectroscopy (EDXS, Bruker GmbH).
ED
M
AN US
CR IP T
ACCEPTED MANUSCRIPT
PT
Fig.15: Element distribution maps for AlSi7Mg0.3 (wt%) base material and Sn78Cu22 (wt%) brazing ribbon from SEM-EDXS: (a) SE-Image, (b) aluminium, (c) silicon, (d) magnesium, (e) tin and (f) copper.
CE
Corresponding SEM-EDXS element maps from this sample (Fig.15) show a non-homogenous distribution of the elements. The base material shows aluminium as the main material, on the
AC
grain boundaries of which silicon is enriched and to a lesser extent magnesium. In the area of the braze point, as expected, no base material is to be found, only copper and tin. In addition, it can be seen clearly that tin spreads beyond the braze point along the silicon-enriched grain boundaries of the base material.
ACCEPTED MANUSCRIPT
6. Conclusions We report on the development and initial experiments with three different in situ units in our LC-SEM. A turning device setup for investigating details of the turning process of metallic samples was developed and first results were presented. The cylindrical sample can be
CR IP T
machined both on its cylindrical surface and on the planar end surface. With this equipment the deformation of the sample surface can be documented during the turning process. Furthermore post mortem examination methods using conventional SEM and TEM provide a deeper insight into the depth to which the deformation extends into the sample and how the
AN US
resulting chip is constructed. We also report on the incorporation of a laser beam into the LCSEM to investigate laser welding processes in situ. We have succeeded in producing laser traces in situ and to analyze the results of a variation of the power of the laser. Finally, we report on in situ wetting experiments in our LC-SEM. In this simulation of a brazing process,
M
it was possible for the first time to observe all stages of the melting and the wetting process of
ED
aluminium base material including the removal of alumina layer directly in the electron microscope. Subsequent analytical studies will provide an accurate insight into the
PT
composition of the individual phases as well as their local distribution.
CE
Acknowledgment:
Our large-chamber scanning electron microscope was acquired within the framework of a
AC
microscope infrastructure initiative of the German Research Foundation (DFG. We gratefully acknowledge support in the context of the Collaborative Research Centre SFB 1120 "Precision Melt Engineering” and the Collaborative Research Centre SFB TRR136 „Process Signatures“ funded by the German Research Foundation (DFG).
ACCEPTED MANUSCRIPT References: [Bob16] K. Bobzin, M. Öte, S. Wiesner, L. Pongratz, J. Mayer, A. Aretz, R. Iskandar, A. Schwedt, presented at WTK 2016, Chemnitz, Germany, March 10 – March 11, 2016, 76-87. [Bob17]
K. Bobzin, M. Öte, S. Wiesner, A. Schmidt, M. Apel, R. Berger, A. Aretz, J. Mayer, Formation of the reaction zone between Sn-Cu brazing fillers and Al-Si-Mg alloys: experiments and thermodynamic analysis, Material Science and Engineering Technology, submitted.
[Ehl16]
L. Ehle, J. Kämmler, D. Meyer, A. Schwedt, J. Mayer, 3rd CIRP Conference on Surface Integrity, Procedia CIRP 45C, 2016, pp. 367-370.
CR IP T
[Eng06] L. Engelbrecht, O. Meier, A. Ostendorf, H. Haferkamp, Materialwissenschaft und Werkstofftechnik 37 (2006) 272–27. [Gha02] F. Ghanem, C. Braham, M.E. Fitzpatrick, and H. Sidhom: Effect of Near-Surface Residual Stress and Microstructure Modification From Machining on the Fatigue Endurance of a Tool Steel. Journal of Materials Engineering and Performance 11 (2002) 631-639. Heidenblut, T.; Möhwald, K.; Deißer, T. A.; Bistron, M.; Behrens, B.-A.; Bach, Fr.-W. (2007): Wear Characterization of Forging Dies Using a Large Chamber Scanning Electron Microscope (Microscopy and Microanalysis, 13), S. 104-105.
[Hei16]
C. Heinzel, F. Borchers, D. Berger, L. Ehle, 3 rd CIRP Conference on Surface Integrity, Procedia CIRP 45C, 2016, pp. 191-194.
[Hos12]
S. B. Hosseini, U. Klement, J. Kaminski: Microstructure Characterization of White Layer formed by Hard Turning and Wire Electric Discharge Machining in High Carbon Steel (AISI 52100). Advanced Materials Research Vol. 409 (2012) 684-689.
[Jaw11]
Jawahir, I.S.; Brinksmeier, E.; M’Saoubi, R.; Aspinwall, D.K.; Outeiro, J.C.; Meyer, D.; Umbrello, D.; Jayal, A.D.: Surface integrity in material removal processes: Recent advances. Annals of the CIRP,60/2, 2011, Seite 603-626.
[Kle97]
M. Klein and S. Klein, Adapting Human Behaviour for the Development of a New Scanning Electron Microscope. Proc. Int. Conf. Micromechatronics for Information and Precision Equipment MIPE'97, Tokyo, Japan, 1997, pp. 324 - 329.
ED
M
AN US
[Hei07]
[Kle08] M. Klein, Oberflächenbearbeitungsvorrichtung, Patentschrift DE10 2006 032 303 A1 2008.01.17
M. Fried, C. Krechel, E.E. Affeldt, B. Eckert, S. Kimmig, U. Retze, H.W. Höppel and M. Göken : Fatigue crack initiation in nickel-based superalloys studied by microstructure-based FE modeling and scanning electron microscopy, MATEC Web of Conferences 14, 16001 (2014), DOI: 10.1051/matecconf/2014416001
CE
[Kre13]
PT
[Kle10] M. Klein, Oberflächenbearbeitungsvorrichtung, Patentschrift DE10 2006 032 303 B4 2010.08.19
AC
[Liu17] B. Liu, X. Zhou, X. Zhang, Orthogonal machining introduced microstructure modification in AA7150-T651 aluminium alloy, Materials Characterization 123 (2017) 91 – 98.
[Luo14] W. Luo, W. Tillmann and U. Selvadurai; In Situ Wear Test on Thermal Spray Coatings in a Large Chamber Scanning Electron Microscope, Journal of Thermal Spray Technology, Vol. 24 (1-2) January 2015, pp. 263-270 [May07] J. Mayer, J. Kallinna, P. Watermeyer, A. Aretz, W. Rehbach, J. Linke*, and A. Schmidt*, Nondestructive Analysis of Engineering Components in the Large-Chamber Scanning Electron Microscope, Proceedings Microscopy and Microanalysis 2007. [Nol04]
R. Nolte, PhD thesis, Friedrich Alexander University Erlangen-Nürnberg (2004)
[Oes08]
Österle, W. ; Krause, S. ; Moelders, T. ; Neidel, A. ; Oder, G. ; Völker, J.; Materials Characterization 59 (2008) 1564-1571.
[Ram14] A. Ramazani, A. Schwedt, A. Aretz, U. Prahl, Key Eng.Mater. 2014, 586, pp. 67-71.
ACCEPTED MANUSCRIPT Schmitt, F., “Laserstrahl-Mikroschweißen mit Strahlquellen hoher Brillanz und örtlicher Leistungsmodulation,” Ph.D. thesis, Shaker, Aachen, 2012.
[Sch16]
G.J. Schmitz, B. Böttger, M. Apel, IOP Conf Ser-Mat Sci, 2016, 117, 012041.
[Til15]
W. Tillmann, J.Pfeiffer, N.Sievers and K.Böttcher, Analyses of the spreading kinetics of AgCuTi melts on silicon carbide below 900 °C, using a large-chamber SEM, Colloids and Surfaces A: Physicochemical and Engineering Aspects 468; pp. 167--173, 2015
AC
CE
PT
ED
M
AN US
CR IP T
[Sch12]