Manufacturing of Welded Rings
Evaluation of Post-Weld Operations
Tillverkning av svetsade ringämnen Utvärdering av svetsfogsoperationer
Jim Andersson
Faculty of Health, Science and Technology
Degree Project for Master of science in Engineering, Mechanical Engineering 30 HP
Supervisor: Christer Burman Examiner: Jens Bergström
Abstract
Pipe and ring blanks can be produced in several different ways. Today’s focus on environ- mental effects motivates companies to develop processes that are as efficient as possible in their production. Ringsvets AB is a company that produces pipe and ring blanks from a flat stock by rolling and welding the piece to make it stay in its desired form. The direct benefit of the method is the minimizing of material loss, and it has thereby both environmental and economical advantages. The downside of the method is that the processes involved changes the mechanical behavior of the ring, locally around the weld zone. The focus of this master thesis is the processes and how they affect the material, both microstructure and behavior.
The processes involved are; rolling, welding, shaping, brushing, forging, heat treatment, and calibration.
The purpose of this work was to elevate the knowledge and understanding of the processes at Ringsvets. The goals were to give a theoretical description of them along with practical test results and explanations of how and why they function in reality.
A literature study has been conducted which provided a theoretical basis on how the material reacts on certain processes. Practical examination of samples from current production has been done to get evidence of how well the processes are used, and how well they function, in today’s production. Lastly, the main focus of the thesis, an evaluation has been made; do theory and practice correlate, and should anything be changed to correlate better?
The results showed that the first operations do not alter the material behavior to an unac- ceptable extent. Forging, on the other hand, gives the material a very high hardness in the weld zone, and that needs to be corrected. The following heat treatment should compensate for that in a perfect world, but does not in reality. The finished ring shows good properties in general but with places where the heat treatment has failed to correct the uneven behavior induced by earlier operations.
The heat treatment requires some adjustments before it functions as intended. Some grains has not been recrystallized which makes them very hard and non-ductile. Future tests using a higher temperature or a longer heat treatment time would reveal the best way to adjust the heat treatment to obtain the desired properties. Other changes in the processes could also be beneficial. Interesting things to try and change would, for example, be the degree of deformation in the forging, which affects the recrystallization temperature.
Notes should be taken that this examination is done on just one sample of just one size.
Analyses of different samples of different sizes should be done to ensure of the accuracy of the examination.
Sammanfattning
R¨or- och ring¨amnen ¨ar en produkt som kan produceras p˚a flera olika s¨att. Dagens fokus p˚a att g¨ora ett s˚a litet avtryck p˚a naturen som m¨ojligt motiverar f¨oretag att utveckla processer som ¨ar s˚a milj¨oeffektiva som m¨ojligt i sin produktion. Ringsvets AB ¨ar ett f¨oretag som pro- ducerar r¨or och ring¨amnen fr˚an ett platt grund¨amne genom att rulla och sedan svetsa ihop ringen s˚a att den beh˚aller den ¨onskade formen. Den tydligaste f¨ordelen med metoden ¨ar att materialspillet minimeras, och det finns d¨arf¨or b˚ade milj¨om¨assiga och ekonomiska f¨ordelar med den. Nackdelen med metoden ¨ar att processerna som anv¨ands f¨or¨andrar ringens mekaniska egenskaper i svetsomr˚adet. Detta examensarbete ¨ar fokuserat kring processerna och hur de p˚averkar materialet i ringarna, b˚ade i mikrostruktur och beteende. Processerna som anv¨ands
¨
ar; valsning, svetsning, hyvling, borstning, pr¨agling, v¨armebehandling och kalibrering.
Syftet med projektet var att ¨oka kunskapen och f¨orst˚aelsen hos Ringsvets f¨or processerna de anv¨ander. M˚alet var att f¨orse Ringsvets med teoretiska beskrivningar tillsammans med resultat fr˚an praktiska unders¨okningar samt f¨orklaringar till varf¨or processerna fungerar som de g¨or.
Genom att g¨ora en litteraturstudie har en teoretisk grund erh˚allits f¨or hur materialet reagerar p˚a vissa processer. Praktisk unders¨okning av prover fr˚an nuvarande produktion har gjorts f¨or att f˚a bevis p˚a hur processerna anv¨ands och hur bra de fungerar i dagens produktion.
Slutligen huvuddelen av projektet d¨ar teori och praktik j¨amf¨ors; st¨ammer teori ¨overens med hur det fungerar i praktiken, och b¨or n˚agon eller n˚agra processer ¨andras?
Resultaten visade att de f¨orsta operationerna inte p˚averkar materialbeteendet speciellt mycket.
Pr¨aglingen, ˚a andra sidan, ger materialet en ¨okar h˚ardheten v¨aldigt mycket i svetszonen, vilket beh¨over korrigeras. I en perfekt v¨arld b¨or efterf¨oljande v¨armebehandling kompensera f¨or det, men s˚a fungerar det inte idag. Den f¨ardiga ringen visar goda egenskaper i allm¨anhet men med st¨allen d¨ar v¨armebehandlingen inte lyckats korrigera de oj¨amna egenskaperna som tidigare operationer introducerar.
F¨or¨andringar i v¨armebehandlingen beh¨over g¨oras f¨or att f˚a ett perfekt resultat. Vissa korn har inte omkristalliserats, vilket g¨or dem v¨aldigt h˚arda och ger dem en l˚ag duktilitet. Framtida tester med h¨ogre temperatur och/eller l¨angre tid i h¨og temperatur skulle visa vilken f¨or¨andring som ger ¨onskat resultat, om n˚agon av de tv˚a f¨oreslagna. Andra f¨or¨andringar i processerna kan ocks˚a vara f¨ordelaktiga. Intressanta saker att f¨or¨andra skulle t.ex. kunna vara graden av deformation i pr¨aglingen, vilket exempelvis p˚averkar rekristallisationstemperaturen.
Det b¨or noteras att denna unders¨okning bara g¨ors p˚a ett prov med endast en storlek. Analyser av olika prov eller olika storlekar b¨or g¨oras f¨or att vara s¨aker p˚a att resultatet av unders¨okningen
¨
ar korrekt.
Contents
1 Introduction 1
1.1 Background . . . 1
1.2 Problem statements . . . 2
1.3 Purpose and goal . . . 2
1.4 Delimitations . . . 3
2 Theoretical overview 4 2.1 Current processes . . . 4
2.1.1 Material . . . 4
2.1.2 Welding . . . 4
2.1.3 Shaping . . . 5
2.1.4 Brushing . . . 6
2.1.5 Forging . . . 6
2.1.6 Heat treatment . . . 7
2.1.7 Calibration . . . 7
2.2 Literature study . . . 7
2.2.1 Crystal structure . . . 7
2.2.2 Welding . . . 15
2.2.3 Induction heating . . . 18
2.2.4 Shaping . . . 18
2.2.5 Friction and wear . . . 20
3 Method 21 3.1 Influence of processes in theory . . . 21
3.1.1 Reference . . . 21
3.1.2 Rolled . . . 21
3.1.3 Welding . . . 21
3.1.4 Brushing . . . 21
3.1.5 Forging . . . 22
3.1.6 Heat treatment . . . 22
3.1.7 Calibration . . . 22
3.2 Experimentals . . . 23
3.2.1 Introduction . . . 23
3.2.2 Purpose and goal . . . 23
3.2.3 Material and Measurements . . . 24
3.2.4 Method . . . 24
4.1 Reference . . . 30
4.2 Rolling . . . 31
4.3 Welding . . . 32
4.4 Shaping and brushing . . . 34
4.5 Forging . . . 36
4.6 Heat treatment . . . 37
4.7 Calibration . . . 46
5 Discussion 52 5.1 Microstructure and preparation . . . 52
5.2 Hardness tests . . . 53
5.3 Evaluation of processes . . . 54
5.3.1 Reference and rolling . . . 54
5.3.2 Welding . . . 54
5.3.3 Shaping and brushing . . . 55
5.3.4 Forging . . . 55
5.3.5 Heat treatment . . . 55
5.3.6 Calibration . . . 55
5.4 Summary . . . 56
6 Future work 58 6.1 Complete analysis . . . 58
6.2 Change of parameters . . . 58
6.3 Change of geometry . . . 59
6.4 Change of material . . . 59
6.5 Change of operations . . . 59
7 Conclusion 60
References 62
A Microscope Images A-1
1 Introduction
This master thesis was performed in the facilities of Camatec Industriteknik in Karlstad with additional supervising from Ringsvets and Karlstad University. The following chapter will introduce the problem, the manufacturing company, and also state the questions, purpose, and aim included in this report.
1.1 Background
Components such as flanges for various applications, and for example a specific case; roller re- tainers in roll bearings, are made from circular stocks. The circular stock can be manufactured in a couple of different ways. In the case of flanges, a common way of manufacturing the stock is laser cutting from a metal sheet. In the retainer case one way to go is by deep drawing, and then remove the top and bottom to get a pipe. Both methods give a lot of material loss as there are a lot of cut off material that remains unused [1].
Ringsvets AB is a company specialized in producing welded rings with minimal material loss with metal sheet or rods as the originating shape. All the stock material is used to make the finished ring without any waste of material. The blanks produced by Ringsvets are in the near net shape of the flange or roller retainer that is the final form. The obvious benefits of this are both economical but also environmental. Their lines producing welded rings are located in Kopparberg, Sweden. Ringsvets has its roots in the 70’s as a part of SKF but has since 1987, when SKF decided to move the production, been operating on its own. Ringsvets is now the leading provider of welded rings in Sweden with SKF and Scania as customers [2].
In the making of a welded pipe or ring blank there are a couple of different processes involved.
The flat stock is cut in suitable length, rolled into a circular shape, and then welded to get a complete circle. The resulting joint differs in desired form, but also in microstructure and mechanical properties compared to the rest of the ring. The procedures following welding are therefore done to neutralize the effect of the welding where the goal is to make the joint as equal as possible to the rest of the ring material, geometry and structure wise. The weld joint is therefore shaped, brushed, forged, and heat treated before the ring gets its final shape by calibration, see Figure 1.
Figure 1: The operations done to make the flat stock a finished ring
Ringsvets wants to expand their knowledge about how and why the processes changes the properties of the rings, and with the use of that information expand their production of welded rings and pipes.
The task has been set up as a thesis at Camatec Industriteknik AB. Camatec is an engineering agency based in Karlstad, Sweden, that provides technical solutions in engineering and process industries [3].
1.2 Problem statements
• How is the material affected by the welding process?
• What can be done to minimize the effect of welding on mechanical and geometrical properties?
– How is the material affected by current post-welding operations?
– How should post-welding operations be performed to get as uniform properties as possible?
1.3 Purpose and goal
The purpose of this work is to get a better understanding of the theory behind the operations done post welding, and what the optimal way of getting the desired result would be. The goal is to compare currently used processes with the theoretical result anticipated. The result of the comparison will be used as a base in the construction of new equipment improving the production lines in the future.
microscope and with hardness tests. Examination of the rings in some steps of the production has been done before [4]. The information in that test is however concluded to be incomplete, making a new test a requirement to make a good conclusion. A new and more thorough examination of the rings in each process will therefore be done.
1.4 Delimitations
The scope of the thesis is 20 weeks at 40h/week. The material compositions used at Ringsvets are mostly low alloy steels, and the main objective of the analysis is therefore such steels. The exact composition will not be revealed as it is classified on behalf of the ordering company.
2 Theoretical overview
This project is based on both the processes used in the production line today and also the theory of the behavior of a steel. This chapter will describe current processes together with a declaration of how a steel is structured and its theoretical reactions in the material by material processing.
2.1 Current processes
The understanding and development of post-welding processes has its base in knowing the cur- rently used operations and their purpose in the production line. This chapter will describe the idea of the processes and how they are performed at Ringsvets in present time. This informa- tion was acquired by visiting Ringsvets, including a guided tour by the supervisor at Ringsvets ( ¨Oholm J, oral communication, december 14th, 2018 and januari 25th, 2019), and completing information retrieved from literature.
2.1.1 Material
The material that has the highest volume of production today is a low alloy steel with a composition resembling AISI 1005, which means that the maximum amount of carbon allowed is < 0.06 wt% [5].
2.1.2 Welding
The welding method used to connect the two ends of the sheet, and produce a ring, is called flash butt welding (FBW), or just flash welding. Flash welding is one of
Figure 2: Two separated parts (a) is joined with flash butt welding resulting in one solid piece (b).
several types of electric resistance weld- ing processes (described in Section 2.2.2.2).
Flash welding is conducted in two stages, the first where the material is cleaned and heated, and the second stage where the two ends are displaced towards each other producing a joint [6]. The displacement of the two ends is what produces the upset material (Figure 2), which is the geometrical error that needs to be corrected. The upset consist of previ- ously molten material together with slag and oxides.
2.1.3 Shaping
Most of the upset material produced by the welding is removed in the shaping operation. The oxide and slag in the upset are something that needs to be kept out of the bulk of the ring.
The shaping process is therefore a well-fitting operation that eliminates the chance of getting the bulk contaminated. Another positive thing with the shaping is that the geometry error is also reduced in the process.
The procedure is very similar to regular machining, in particular the single-point method called metal shaping. The difference between the two is that the work piece is still hot from the welding in the operation done at Ringsvets, while the only heat above room temperature in regular machining is induced by the shaping itself, i.e. friction and wear. With the welded ring clamped, an arm with a shaping tool mounted on the tip goes over the surface where the joint is, in a similar manner to Figure 3.
The shaping is done in one stroke and reduces the height of the upset to just a fraction which make the cutting depth rather big. (a) and (b) in Figure 4 shows the upset before and after the shaping.
Figure 3: The shaping procedure with the ring and a part of the ma- chine.
Figure 4: The upset material in unshaped (a) con- dition, and the upset after shaping (b).
2.1.4 Brushing
Figure 5: (a) shows the joint in its shaped con- dition and (b) shows the smoother upset after brushing.
The shaping results in a thinner joint thick- ness, but the cutting also leaves sharp cor- ners of the upset, Figure 5(a). If the joint is struck in that condition the tool would be damaged, and the result of the forging would also be a less uniform geometry. To get a better result, and also spare the forging tool, the brushing is done with steel wire brushes, giving the remaining upset a smoother tran- sition to the bulk material as seen in Figure 5(b). Two brushes in the same axis rotate in opposite directions to make both sides of the upset smooth.
2.1.5 Forging
With a prepared upset the following step is forging to give the ring its uniform thickness. This is acquired by a hydraulic press with a seat on the inner radius and a corresponding seat on the cylinder side pushing on the outer radius, as seen in Figure 6. The ring is pressed between the two seats which makes the upset go down to the same thickness as the rest of the ring.
Figure 6: The ring with its upset and the forg- ing tooling shows in (a), and (b) illustrates the ring in its homogeneous geometry.
2.1.6 Heat treatment
Previous operations gives the ring non-uniform mechanical properties. In most applications uniform properties are crucial, which is the motivation of the heat treatment. The heat treat- ment is done by induction heating of the joint area. Subsequent cooling with the ring travelling through a set of fans gives the ring its desired properties.
2.1.7 Calibration
Figure 7: An underdimensioned ring (a) is cal- ibrated to its correct dimension (b).
The term calibration is an appellation at Ringsvets meaning deformation to the cor- rect dimension and form. The rings are made too small and are deformed to the correct di- mension by putting them on an expanding jig. The jig consists of pie wedges of metal that can be displaced in the radial direction.
Figure 7 illustrates the calibration, where the pie wedges have a conical hole in the center and the expansion is done by driving a corre- sponding conical axle into the hole by a hy- draulic cylinder.
2.2 Literature study
With a clear view of how the operations are performed, one can start investigate what happens theoretically in the material during each operation. This chapter will describe some fundamen- tal concepts of material processing that can be of use when one wants to find an explanation to the results of the operations.
2.2.1 Crystal structure
The fundamental building blocks in a material are the atoms and their way to arrange in the lattice. The lattice is an array of points where identical unit cells are repeated in each point, and the way the atoms are placed in the unit cells in three dimensions produces different types of crystal structures. There are fourteen different ways to distribute the atoms in three dimensions, commonly named Bravais lattices [7].
Figure 8: Illustration of three Bravais lattices;
body-centered cubic, face-centered cubic, and body-centered tetragonal [7].
Three of the Bravais lattices are body- centered cubic (BCC), face-centered cubic (FCC), and body-centered tetragonal (BCT).
The starting point of all three lattices are rectangular unit cells with one atom in each corner. Body-centered cells have an addi- tional atom in the center of the body while face-centered cells have additional atoms in the middle of each face of the unit cell, Fig- ure 8. All three sides of the cubic structure are of equal length (a = b = c), apart from the tetragonal unit cell in which two sides are equal but not the third (a = b 6= c).
2.2.1.1 Solidification
All molten materials freeze when their temperature goes below the freezing point of the ma- terial, known as solidification. The solidification process of crystalline materials, e.g. metals, involves two steps, nucleation and growth [7]. In the nucleation stage, the liquid form ultra- fine crystallites at random points. With continued cooling the thermal energy reaches a point where the ultra-fine crystallites start to grow across the liquid, until all the liquid has solidified.
Figure 9: The principle of polygonized and columnar grain structure
As the crystals nucleate and grow, the tem- perature gradient of the solidified liquid keeps getting lower and will eventually come to a point where grain growth is predominating rather than nucleation [8]. The initially poly- crystalline structure changes as certain direc- tions have a higher growth rate, forming a columnar crystal structure as visualized in Figure 9.
2.2.1.2 Microstructure
The phase of the grains depends on several different conditions, such as composition and cooling rate. The Fe-C equilibrium diagram in Figure 10 shows the stable phase at different temperatures depending on carbon content. Various metastable phases can be obtained by controlling the cooling rate.
Figure 10: The Iron-Carbon equilibrium diagram [9].
CCT-diagram
Figure 11: Constant cooling diagram for a steel with 0.2% Carbon [10].
A frequently used way to perform a heat treatment is with continuous cooling. Con- tinuous cooling can be achieved in different ways with different cooling rates. Cooling in ambient air is one type that is common in both welding and subsequent heat treat- ment. The microstructure obtained in such conditions can be deduced from the contin- uous cooling transformation (CCT) diagram, see Figure 11.
F - Ferrite P - Pearlite B - Bainite M - Martensite
Denotions subscripted with s marks start temperature, and f marks finish temperature.
Austenite
Austenite, the gamma (γ) phase of iron, is the parent phase to most phases, that almost all steels goes through in solidification and further cooling [11].
Ferrite
In iron solidification there are three steps. The first occurs immediately after solidification where the iron forms δ-ferrite, a BCC-structure [7]. At a cooler temperature the iron transforms to austenite (FCC), and with even further cooling it transforms back to the BCC ferrite structure, this time called α-ferrite. α-ferrite is the structure of pure iron in lower temperatures and is often the structure discussed in heat treatment, as just ferrite.
Cementite
Cementite is a compound of iron and carbon, Fe3C, that is formed where the carbon content is higher than the maximum soluble proportion. Cementite can be compared to ceramics as it is very hard and brittle [7]. The carbon content in cementite is 6.67%.
Pearlite
The structure formed with ferrite and cementite in a lamellar formation is called pearlite [7].
It has gotten its name from the visual effects when etched and observed in a microscope. Most of the pearlite is ferrite, the cementite is therefore acting as dispersion strengthening particles.
The transformation from austenite to pearlite starts almost exclusively in grain boundaries.
The grains size, and thus the amount of grain boundaries, is therefore an important property to pearlite formation [8]. Nucleation of pearlite can also occur in dislocations (Section 2.2.1.3), which is why plastic deformation is beneficial for the transformation to pearlite.
Martensite
When quenching a steel from its austenite phase the structure obtained is the hard and brittle martensite. The cooling rate in quenching is so high that diffusion of carbon atoms cannot occur [11]. The composition in martensite is therefore exactly the same as in the former austenite. The carbon atoms get trapped in the BCC structure giving martensite its BCT structure. The interstitially soluted carbon atoms enables the carbon content in martensite to greatly exceed the limit for solid solution. A higher content of carbon generates harder and less ductile martensite, due to the larger matrix disturbance.
Bainite
At temperatures between pearlite and martensite formation several different microstructures form, called bainite. There is a distinct line of bainite transformation in low-alloy steels, while the change from pearlite to bainite, and bainite to martensite is more diffuse in medium- and high-carbon steels. A sufficiently high alloy content can completely arrest the formation of bainite [11].
Six different morphologies of bainite are found. In high- and medium-carbon steels, bainite is a combination of ferrite and cementite, but has unlike pearlite a non-lamellar structure. There are two morphologies of ferrite-cementite bainite; upper and lower bainite. Low-carbon steels form bainite without any cementite. This bainite is composed of ferrite and retained austenite.
The bainite in low-carbon steels is considered as a form of ferrite, due to the lack of cementite.
2.2.1.3 Dislocations
Dislocations are line defects in the crystal lattice of a material, that can appear in the lattice after solidification or from plastic deformation. Dislocations themselves, and grain boundaries are effective hinders to dislocation motion. Ductility, a measure of the ability to deform without break, is highly dependent on dislocation propagation, and therefore also on dislocation density and the size of grain boundaries in the material. There are three types of dislocations; edge dislocation, screw dislocations, and mixed dislocations [7].
2.2.1.4 Deformation
Deformation of a material can occur in different ways. There are two main principles of deformations that occur in almost all materials before break, although the proportion of each part differs.
Figure 12: A schematic diagram of a stress-
Elastic Deformation The first part of a stress-strain curve (Figure 12) is the elastic part.
The elastic deformation shows as the linear fraction of the curve. This type of deformation is not permanent and will regress when unloading, also called elastic springback. The springback is the elongation of the material in the elastic area [13].
Plastic deformation A higher load, resulting in a higher stress, makes the deformation transform from elastic to plastic strain. That happens at the end of the the linear part of the curve when the stress reaches the yield stress of the material. The plastic deformation is permanent deformation that will remain even after unloading. There is elastic springback when unloading even when plastic deformation has occured, making it an important property to consider in metal forming [13]. The plastic deformation occurs in the non-linear region of the stress strain curve between the yield stress and the tensile stress, Figure 12.
Cold working and strain hardening
Figure 13: Rolling of a metal sheet and the change of grain geometry [7].
The principle of forming a metal by plastic deformation below its recrystallization tem- perature is called cold work. The rolling process in the manufacturing of metal sheets are one form of cold working. In rolling, the grains gain anisotropic behavior by elon- gation through the rolling process, demon- strated in Figure 13.
The principle of strain hardening is based on that the number of dislocations is ele- vated during plastic deformation. When the amount of dislocations is at a certain level, they start to interact with each other, pre- venting further motion. Interacting disloca- tions increases the strength, but it also de-
creases the ductility [7]. Not all materials can be strain hardened, but metals are one of those that can. The criteria of strain hardening is that the material needs to be of crystalline structure and also non-brittle.
Hot forming
Another common way to form steel is by hot forming. Hot forming is when the steel is heated up before deformation. The plastic deformation initiates recrystallization which creates equiaxed grains instead of elongated as in cold working. Hot rolling sheet metal is a substitute to rolling in its cold state that produces a softer and more ductile metal sheet that is suitable for subsequent cold working. One can obtain different grain sizes by controlling the temperature of the hot rolling [11].
2.2.1.5 Dispersion strengthening
A dispersion-strengthened material consists of a matrix of a relatively soft and ductile phase, and small particles of a harder phase within the matrix. The strengthening effect comes from the stopping of dislocation motion that the dispersion soluted particles provide [7].
2.2.1.6 Annealing
Cold working can introduce both wanted and unwanted behavior to the material. Annealing is a heat treatment that is used to remove some or all of those behavior. Annealing can be done in three different levels.
Recovery
Cold worked material contains a lot of dislocations and some residual stresses. The first stage of annealing is heating to a level where dislocations dissolve and reorganize to form polygonised sub-grain boundaries [8], Figure 14. Rearrangement of dislocations lowers internal stresses in the material, but mechanical properties are maintained as the dislocation density remains unchanged.
Figure 14: Illustration of rearranged disloca- tions as sub-grain boundaries [7].
Recrystallization
At a specific temperature, nucleation of new grains is initiated in the sub-grain boundaries formed by the dislocations in the recovery stage, creating a new polygonal grain structure. This temperature is called the recrystallization temperature and is dependent on several different properties. A high amount of the dislocations is removed giving the material a lower strength and higher ductility than in the cold worked state [7].
The main things that the temperature of deformation depends on is [14]:
• Temperature and time of annealing
• Degree of cold work in the metal
• Amount of alloying elements in the metal
• Grain size before cold work
• Temperature during the deformation
The recrystallization can be plotted against time for a certain amount of strain and tempera- ture. Figure 15 shows an example of a recrystallization diagram.
Figure 15: Recrystallization diagrams for a mi- croalloyed steel at 20 and 35% deformation [15].
Grain growth
Even higher temperatures benefits grain growth. Grain growth is present in all polycrystalline materials at an adequate temperature. The grain growth phenomena is generally something that one wants to avoid as bigger grains are equal to less grain boundaries. Grain boundaries are, like other dislocations, a hinder for dislocation motion, and thus affects the ductility and
2.2.1.7 Normalizing
The heat treatment used to give the steel a fine and uniform structure is called normalizing.
In normalizing the steel is heated above the austenitizing temperature, to an even higher temperature than that of annealing. The steel is then cooled in air which consequence in higher cooling rates than those used in annealing. The structure obtained due to the cooling rate is fine ferrite grains and a small spacing between the lamellas in the pearlite [11].
2.2.2 Welding
Joining two parts of metal can be done in several ways, one of them are fusion-welding. In fusion welding a minor area of both parts are heated above the melting point and the two parts are joined in the following solidification [7]. The welding heat that melts the material in the fusion zone also affects the surrounding area, known as the heat affected zone (HAZ), similar to Figure 16.
2.2.2.1 HAZ
Figure 16: Different zones in a weld; fusion zone (FZ) with columnar grains, heat affected zone (HAZ) with equiaxed grains of various size, and unaffected cold worked zone (CWZ) with elongated grains [7].
There will always be a heat affected zone in a fusion weld. The size and structure of the HAZ is influenced by heating rate, cooling rate, and the peak temperature. A rapid en- ergy input will quickly raise the temperature in the material which will limit how far the heat is able to spread and results in less en- ergy needed to melt the material, thus the size of the HAZ is reduced. A high peak tem- perature contains more energy which will en- large the HAZ and keep the heat in the ma- terial for a longer time, i.e. affect the cool- ing rate. Factors controlling cooling rate are, amongst others, material thickness, thermal
conductivity, and temperature of both the specimen and ambient air (or other cooling media) [16]. The microstructure of the HAZ depends on peak temperature at the spot and the time spent in the temperature range (i.e. cooling rate), as discussed in Section 2.2.1.2.
2.2.2.2 Resistance welding
Heat generation in resistance welding comes from two sources of the same origin; current flow.
The pieces that will be joined are clamped with electrodes and displaced until contact occurs in the weld interface. The resulting current flow induces a joule heating of the two parts, in addition to the higher temperature in the interface that comes from the electrical resistance in the interface. Resistance welding is a very heat effective welding method compared to regular gas or arc welding as the welded metal itself is the heating source, and the heating rate is therefore very high [16].
Flash butt welding
FBW, that is conducted at Ringsvets, can be divided into five steps. In the starting point of the welding a voltage is applied to the pieces before the surfaces are in contact (Figure 17(a)).
The two pieces are then displaced towards each other, which generates a flashing between the two butts as the current flows through the points closest to each other (Figure 17(b)). The flashing burns off oxides and melts protuberances, resulting in a temperature rise of surround- ing material. With continuously displaced ends more and more of the butting areas will be burnt and melted (Figure 17(c)). As enough volume near the interface is above the plastic temperature, stage two is initiated. The pieces are further displaced towards each other, be- yond the contact point with high force, driving molten oxides and bulk material out of the interface [17], (Figure 17(d)). Subsequent removal of current enables the joint to cool down and end up as a uniform joint (Figure 17(e)).
Figure 17: Sketch of the flash butt weld process with: (a) applied current, (b) displaced ends
The two different stages of FBW influence the material in different ways. Heating and melting material are somewhat equivalent to other welding methods; there is a fusion zone, a heat affected zone of various grain sizes and structures, and the unaffected bulk. What differs is the second stage of FBW where the material is upset and plastically deformed. Most of the molten material is pushed out in the upset, leaving the fusion zone very small. Section 2.2.1.3 describes dislocations and that plastic deformation is a source of those. A high amount of dislocations is introduced in the plastically deformed area when applying the butting force, Figure 18(c)-(e) [19].
Figure 18: The change of microstructure during the flash weld process [19].
2.2.3 Induction heating
Figure 19: An induction coil (orange) with ap- plied current (yellow arrows), induced magnetic field (black lines), and the resulting eddy cur- rent (black arrows) [20].
Induction heating is somewhat similar to the heating in resistance welding. The tempera- ture raise comes from a joule heating of the part, but with an eddy current as the source.
An eddy current is a current induced by a sur- rounding magnetic field, according to Fara- days law. By applying an alternating current to an induction coil, one creates an alternat- ing magnetic field. By exposing the subject- to-be-heated to the magnetic field, the result- ing eddy current flow in the subject elevates the temperature due to the electrical resis- tance [21], see Figure 19.
The induction heating process is a very effec- tive heating process as it, just like resistance welding, uses the material itself as the heat-
ing source. The high efficiency allows the heating to be local, with a very small area nearby affected by the heating. Induction heating is a heating method of high control and precision.
When set up and calibrated according to the heat treatment requested, the repeatability of the process is excellent making it suitable in automated industry lines [22].
Most conductive materials can be induction heated, but magnetic materials are easier to heat than non-magnetic materials. That is because magnetic materials produce a hysteresis effect that helps heating the material, in addition to the eddy current heat. The hysteresis effect stops working when the material reaches its Curie temperature, where the material stops being magnetic [20].
2.2.4 Shaping
Shaping is an orthogonal machining process where a cutting tool moves over a surface in a reciprocal motion. The cutting tool digs into the material in the forward motion removing material as chips. It is defined as an orthogonal machining process because the velocity vector of the tool is perpendicular to the cutting edge of the tool.
2.2.4.1 Chip formation
Figure 20: Close-up on the chip formation pro- cess [23].
The formation of chips involves plastic defor- mation and shearing of the cut material be- tween the cutting edge and the upper surface of the chip where it deflects away from the surface [23], see Figure 20.
The cutting conditions and the geometry and size of the chips are highly interrelated. The chips are often used in machining as a mea- sure of how accurate used parameters are.
Material properties, machining parameters, and geometries are all affecting chip behav- ior.
The thickness of the chips is of course reliant on the cutting depth, but also the grade of
deformation that occurs in the chip. The grade of deformation is denoted r and is defined as the ratio between the thickness of the undeformed chip (t1), and the deformed one (t2)[23]:
r = t1
t2 (1)
The material is assumed to be incompressible, ∆V = 0, ensuing that the deformation in the cutting direction is directly linked to the r -ratio. A higher deformation in the cutting direction forces the thickness to be greater. The shear plane angle (φ) gives an indication of the cutting direction deformation and is defined by the tool rake angle (α), cutting speed, and t1. A shallow cutting depth, low cutting speed, and a low α minimizes the shear plane angle and thus maximizes the deformation. A high deformation also elevates the cutting force.
As there are plastic deformation involved, there will also be deformed grains in cutting. Most of the deformed material is removed from the bulk in the chip, but there are also some deformed grains in the surface of the shaped object.
2.2.5 Friction and wear
Two surfaces sliding against each other will, inevitably, be subjected to both friction and wear.
The underlying mechanism for friction and wear is highly dependent on the basis of the sliding [24].
2.2.5.1 Abrasive wear
Wear by hard particles between two surfaces is called abrasive wear, and it can be divided into both two- and three-body abrasion, Figure 21. Two body abrasion is when hard particles are fixed on one surface and sliding against the other, and three body abrasion is when the hard particles are free to either roll or slide between the surfaces [24].
Figure 21: The two types of abrasive wear [24].
Abrasive wear can occur both by plastic deformation and by brittle fracture. The two types of wear differs in the hardness and ductility of the bulk material, plastic deformation occurs in soft and ductile materials whereas brittle fracture occurs in hard and brittle materials.
3 Method
The method chapter is divided into three parts. One part with the processes in theory, one with the processes in practice, and the third part were them both are compared.
3.1 Influence of processes in theory
As seen in Section 2.2, the microstructure and mechanical properties of a steel are greatly affected by physical circumstances. This section will describe how the rings theoretically will change due to the operations, with a base in the literature study.
3.1.1 Reference
A hot rolled sheet metal would have polygonal grain structure. The grain size depends on deformation and hot rolling temperature.
3.1.2 Rolled
The rolled metal sheet probably has a higher hardness than the flat sheet. The hardness difference depends on how deformed the sheet is. A thick thickness and a small radius forces the deformation on the outer and inner radius to be high, and it is probably at those areas where the rolled ring is the hardest. The hardness elevation comes from deformed grains and dislocation density, as stated in Section 2.2.1.4.
3.1.3 Welding
Resistance welding is a welding method of high heating rate resulting in a relatively small HAZ.
The weld joint contains of small grains near the unaffected material and bigger grains near the interface. There is also a high dislocation density near the interface. The upset material makes the thickness non-uniform with a bead of oxide and previously molten material.
3.1.4 Brushing
The brushing operation is done in a temperature above room temperature, where the hardness and yield strength of the material are lower than in lower temperatures. The metal wires on the brushes are relatively stiff and hard making the material removal akin to abrasive wear. The abrasive wear removes material and induces some plastic deformation into the material. The result is probably very equal to the previous stage with a small fraction of plastic deformation.
The material removal is the most at the sharp edges making the remaining bead smooth again.
3.1.5 Forging
The forging operation is straight forward plastic deformation in a cold state, cold working.
Elastic springback needs to be considered and compensated for to obtain the correct thickness.
The material is deformed in the area of the upset resulting in elongated grains, similar to those in Figure 13. Assuming that the material is incompressible implies that the volume of the material that is deformed and displaced from the thickness will be added to the circumference, and/or the width (axial direction) of the ring. The cold working will, as described in Section 2.2.1.4, induce a lot of dislocations to the material in the deformed area.
The result of the forging is a ring with uniform thickness, and a little bit larger circumference.
The grains will be similar to those after welding but a bit deformed. The ring will also contain a lot of dislocations and some residual stresses.
3.1.6 Heat treatment
Previously performed operations provides a microstructure dissimilar to the base material.
The heat treatment would optimally regain the size and geometry of the grains in the rolled condition, and also present a microstructure of the same kind. Most of these properties can be obtained with a heating procedure.
The heat treatment would give the ring nearly equal mechanical properties through the whole ring. There will be some difference in grain size through the weld zone due to the temperature gradient. A thoroughly controlled cooling allows the material to be of uniform microstructure.
3.1.6.1 Heating temperature
Heating the ring above the recrystallization temperature will dissolve dislocations and nucleate new grains. The desired microstructure will be attained with sufficient grain growth, viz, a suitable temperature gradient in an appropriate time. As the material is a very low alloyed steel there is not much restriction in cooling rate. Air cooling would be sufficient to obtain the desired ferritic structure.
3.1.7 Calibration
With all operations finished to eliminate the properties induced by previous processes, the resulting ring would be in good enough shape to be called homogeneous in the desired mea- surements. The ring should show a consistent level of average grain size through the whole ring with equiaxed grains all through. The microstructure should be equal throughout the sample with some dislocations and residual stresses induced by the last operation.
Again, the material is assumed to be incompressible. The change in the circuit of the ring will also change the thickness of the ring. Attention must be paid so that the thickness ends up in the interval of what the customer ordered. The strain must also be limited to not exceed the amount possible before break (shear stress below tensile strength).
3.2 Experimentals
Practical tests were executed in the metallurgical laboratory at Karlstad University. Samples were collected from various stages in the production at Ringsvets. This section will describe the samples and the analyses along with the results obtained.
3.2.1 Introduction
The theory states that several processes in the making of the rings have a large impact on the mechanical and metallurgical properties. Material examination is therefore done to see in which way the processes affects the material. The hardness of a material is a great way to judge mechanical properties as it is defined as the resistance to plastic deformation, which is a major property deciding the mechanical behavior of a material [25].
3.2.2 Purpose and goal
The purpose of the examination is to determine how the current processes affects the material, in microstructure and hardness.
The goal is to obtain pictures of the weld area in a large magnification to be able to distinguish the phase of the microstructure and the grain shape. The grain size should be measured, and hardness measurements should be done to complete the information.
3.2.3 Material and Measurements
Figure 22: The measurements of the examined ring.
The exact material of the rings are classified, as men- tioned in Section 2.1.1, but they are made of a very low alloyed steel. Ringsvets produces rings in a variety of dimension. The tested rings had the measurements of diameter = 352 mm, thickness = 5 mm, and width
= 75 mm, illustrated in Figure 22.
3.2.4 Method
Figure 23: The six samples collected at Ringsvets. From the left; reference, welded, shaped and brushed, forged, heat treated, and calibrated.
Tests were done on samples in the stages of the produc- tion line that influences the microstructure the most, which where concluded to be; welding, forging, and heat treatment. Samples were therefore collected be- fore and after those processes to be able to make a conclusion of the influence of the operations. A sample of a finished ring (after calibration) were also collected along with reference samples of as-delivered sheet metal and ring formed rolled sheet metal. The samples can be seen in Figure 23.
An attempt to measure the temperature in each stage was made. A hand-held IR-thermometer was used in the effort to get the temperature peak and cooling rate on paper. The thermometer was used simultaneously with the line in full production.
Ringsvets believe that some processes are not perfect and produces a differing result through the width of the rings ( ¨Oholm J, oral communication, januari 25th, 2019). The width was therefore cut in three places, with a being furthest from the machine, b in the middle, and c closest to the machine, Figure 24. The samples were hot mounted, looking in the axialdirec- tion of the ring. That is to make them easier to handle, and making them fit the preparation machine [26], Figure 25. The hot mounted samples were then prepared in a standard metal- lurgical way [27].
Figure 24: The samples were sectioned in three places; a, b, and c.
Figure 25: Eighteen of the hot mounted specimens.
Table 1: The the samples and their associated num- bers
No. Sample 1. Reference 3. Welded
4. Shaped and Brushed 5. Forged
6. Heat Treated 7. Calibrated A unique name was given to each sample, to be able to separate
them from each other. The names have the format of XYZ. X is after which process the sample is collected (Table 1), Y is where on the width the cut is made (Figure 24), and Z is if the sample is taken in the middle of the weld (Z=1), or outside the HAZ (Z=2). The pictures are given a name with an additional U and V in it, XYZ-U-V. The U denotes the magnification of the picture, and V is the position in each sample at which the picture is taken, according to the figures associated to each sample.
The polished cylinders were etched and put in a microscope. Pictures were taken at sites of interest, which differed depending on the previous operation. The locations of the pic- tures are described later.
Samples were collected in and outside the weld zone. Pictures were taken on the outer areas in two spots as shown in Figure 26.
Figure 27: The grain size measurement points.
Grain size is one property that can influence the mechanical behavior of a material, why it seemed interesting to make grain size mea- surements on some of the specimens. The grain size was measured in three different places of each specimen, as illustrated in Fig- ure 27. The grain sizes were measured with Abrams three-ring procedure [28] in three spots close to each other, which combined produced an average value.
Figure 28: The three directions of the hardness testing.
The last step of the examination was hard- ness testing. The hardness test method was the Vickers method [29, 30]. The load was 25 gf with tests done in both the circumferential direction, called the 1-direction, and in some cases also the radial direction of the rings, the 23-direction in this test, as seen in Figure 28.
As the thickness of the rings might differ, the 23-test was divided into two. Two hardness tests were done with the starting point at the outer and inner edge (the 2 respective the 3- direction), and then put together in one dia- gram to get a feeling of the hardness profile through the thickness.
3.2.4.1 Reference and rolling
Figure 29: The only spot where pictures were taken in the reference and rolled sample.
Pictures were taken of the as-delivered sheet metal and the rolled, but not welded, ring.
Pictures of both samples were taken in just one spot as the microstructure is probably equal through the whole sample as seen in Figure 29.
3.2.4.2 Welding
Figure 30: Points where pictures are taken after the welding operation.
In the welded stage the microstructure would have changed due to both heating and de- formation (Section 2.2.2). The number of spots where the microstructure was examined is therefore more than in the reference case, as seen in Figure 30.
3.2.4.3 Shaping and brushing
Figure 31: Pictures were taken in three loca- tions after shaping and brushing.
The change in microstructure is probably not that much in the brushing stage, but it is of high interest to compare the sample before and after forging, which is why the sample was taken after brushing. The difference from the post-welding sample is in the area of the bead (Section 2.2.4), and the pictures were taken in and around that area, Figure 31.
3.2.4.4 Forging
Figure 32: The picture shows where images where captured in the forged specimen.
Forging is done by massive plastic deforma- tion and changes the grains a lot in the de- formed area (Section 2.2.1.4), hence the cap- tured spots seen in Figure 32.
3.2.4.5 Heat treatment
Figure 33: Heat treating gave seven spots to take pictures at.
The heat treatment is the third process that changes the microstructure a lot, and is done to do so. The heat treatment is done to neu- tralize the earlier changes, and pictures are for that reason taken all over the sample, Fig- ure 33.
3.2.4.6 Calibration
Plastic deformation also occurs in the calibration stage, which is why it is interesting to examine the influence of calibration in the weld area, Figure 34.
The goal with all post welding operations is to make the ring uniform. Confirmation of the function of the processes was made by comparing the finished weld area with the rest of the ring, unaffected from the welding, forging, and heat treatment, Figure [35].
Figure 34: Pictures were taken in six points of the finished ring.
Figure 35: A finished but unaffected sam- ple, welding wise, got three points of inter- est.
4 Results
Relevant results of the material examination are presented in pictures of the microstructure with completing comments, tables of grain size, and graphs of the hardness through the sample.
The complete results from the examination can be seen in Appendix A.
What can be seen in the pictures of the microstructure is that the samples were not polished in a truly sufficient way. They have also been etched a little bit too much resulting in very corroded grain boundaries and, in some cases, also corroded grains. However, the conclusion was that the samples could be used to examine the microstructure despite those flaws.
The temperature measurement gave inconclusive results that were hard to explain. The mea- surements are therefore excluded from the study. There are no evidence of that the phase of the microstructure should have changed through the processes which makes the only mi- crostructural change the grain size and geometry. The material had the same phase as the reference at all stages.
4.1 Reference
Microstructure
What can be seen in Figure 36 is that the grains were edgy and of different shapes in various sizes, mostly medium sized. There were smaller grains along the edges of the sheet. The phase of the grains was almost pure ferrite.
Figure 36: Pictures in different magnifications of the reference specimen in spot 1.
Hardness
Figure 37 shows the hardness of the as-delivered sheet metal in the ring direction. The hardness of the sheet was a steady average around just above 150 HV.
Figure 37: The hardness profile of the reference specimen in the ring direction, from the center of the weld and outwards.
4.2 Rolling
Figure 38: The hardness profile of the rolled specimen in the radial direction, from the outer radius and inwards.
Microstructure
The grain size in the rolled condition was smaller further from the edge than in the raw sheet. Otherwise it was similar to before, with various types of grain geometries.
Hardness
Rolling the sheet elevated the hardness in general with a minimum holding the initial hardness in the center and a higher hardness closer to the inner and outer radiuses of the ring, Figure 38.
4.3 Welding
Microstructure
There were similar grains to the reference in both position 1 and 2 after welding, although a little bit smaller grains in position 2. Position 1 after forging and position 7 after heat treatment did also have the same microstructure, which is why the images are not included in the report.
In Figure 39 one can see that the grains in position 3 were of very mixed geometry with smaller and bigger grains in both equiaxed and more random conditions.
Figure 39: Pictures in different magnifications of the welded specimen in spot 3.
The grains in the bead were even more irregular and needle like. Both in the upper and in the lower bead, Figure 40 and 41.
Figure 40: Pictures in different magnifica- tions of the reference welded in spot 4.
Figure 41: Pictures in different magnifica- tions of the welded specimen in spot 5.
Figure 42: Pictures in different magnifications of the welded specimen in spot 6.
The outer part of the bead had a more sym- metrical geometry but there were also some needle like grains. Figure 42 also shows that there was a lot of slag and inclusions.
Figure 43: Pictures in different magnifications of the welded specimen outside the weld.
The images of the sample taken outside the weld proves that there was no considerable change of grain structure longer from the weld, Figure 43. Other pictures from corre- sponding samples will therefore be excluded from the report.
Figure 44: The grain size in the welded speci- men.
Grain Size The diagram of the grain size after welding seen in Figure 44 shows that the grains were pretty equal in size with a very small valley a couple of millimetres outwards from the center of the weld.
Figure 45: The hardness profile of the welded specimen in the ring direction, from the center of the weld and outwards.
Hardness
Looking at Figure 45, one can see that the hardness in the center of the weld was very similar to the reference samples. One can also see that there is a very slight inclination of the graph further from the weld area. All three graphs show the same behavior with a hardness of ca 155 HV inclining to ca 165 HV in average at 7-8 mm from the center of the weld.
4.4 Shaping and brushing
Microstructure
Position 1 just outside the weld had a somewhat similar structure to the reference sample, see Figure 46.
Figure 46: Pictures in different magnifications of the shaped and brushed specimen in spot 1.
At the edge of the remaining bead there was the same geometry of the grains, but they were, as one can see in Figure 47, a little bit bigger.
At the peak of the bead there were grains that are quite deformed. Oblong and needle like grains were present near the edge, Figure 48
Figure 47: Pictures in different magnifica- tions of the shaped and brushed specimen in spot 2.
Figure 48: Pictures in different magnifica- tions of the shaped and brushed specimen in spot 3.
Hardness
Shaping and brushing the weld appears to have risen the hardness near the center making the average hardness very static at just below 170 HV, aside a very small minimum at just over 2 mm from the center. All three samples had the same behavior, which can be seen in Figure 49.
Figure 49: The hardness profile of the shaped and brushed specimen in the ring direction, from the center of the weld and outwards.
4.5 Forging
Figure 50: Pictures in different magnifica- tions of the forged specimen in spot 2.
Microstructure
The forging produced deformed grains through the whole thickness. The grains in the center (Figure 51) were deformed more severe than those on the outer (Figure 50) and inner radius (Figure 52).
Figure 51: Pictures in different magnifica- tions of the forged specimen in spot 3.
Figure 52: Pictures in different magnifica- tions of the forged specimen in spot 4.
Grain Size
The forging made a big impact in the appearance of the grains. As seen in Figure 53, the grains were small in the center with slightly bigger grains some distance from the center. There are ambiguous results further from the center. Two of the graphs shows a decrease in grain size while the third one shows an increase in grain size.
The forging made a big impact in the appearance of the grains. As seen in Figure 53, the grains were small in the center with slightly bigger grains some distance from the center. There are ambiguous results further from the center. Two of the graphs shows a decrease in grain size while the third one shows an increase in grain size.
Figure 53: The grain size of the forged speci- men.
Hardness
The graphs in Figure 54 demonstrates that the forging operation made the weld very hard while the points at 4 mm and further from the weld had approximately the same hardness as before.
The hardness test through the thickness in Figure 55 displays a hardness profile well above the earlier ones with a maximum in the middle of the thickness. There was a small deviation between the inner and the outer ra- dius where the outer one was the hardest.
Figure 54: The hardness profile of the forged specimen in the ring direction, from the center of the weld and outwards.
Figure 55: The hardness profile of the welded specimen in the radial direction, from the outer radius and inwards.
4.6 Heat treatment
Microstructure
The grains in the lower outside area of the HAZ were still deformed after heat treatment, with the center of the ring being the most deformed, Figure 56. The figure shows that the outer areas also had deformed grains, although a little bit less deformed than the center.
Figure 56: Pictures in different magnifications of the heat treated specimen at section a, b, and c in spot 1.
Figure 57 displays the upper area closer to the center with equiaxed grains in all three areas.
Figure 57: Pictures in different magnifications of the heat treated specimen at section a, b, and c in spot 2.
The grains in Figure 58 are similar to those in Figure 57 with equiaxed grains. The outer areas have the same grain size with smaller grains in the central one.
Figure 58: Pictures in different magnifications of the heat
In position 4, Figure 59, the grains were equiaxed with mixed dimensions.
Figure 59: Pictures in different magnifications of the heat treated specimen at section a, b, and c in spot 4.
The upper outer part of the HAZ had equiaxed grains in the two outward sections and a bit deformed grains in the section closest to the machine, Figure 60.
Figure 60: Pictures in different magnifications of the heat
There were equiaxed grains at the edge outside the HAZ in general, but the area the longest from the machine had bigger elongated grains at the edge, as can be seen in Figure 61.
Figure 61: Pictures in different magnifications of the heat treated specimen at section a, b, and c in spot 6.
Hardness The heat treatment affected the grains in the way such that there were bigger grains found in the center of the weld with smaller grains further from the center, Figure 62.
Figure 62: The grain size in the heat treated specimen.
Hardness
The heat treatment lowered the hardness to the same level as the reference sample, from the center to about 2 mm outwards. Further from the center showed the same behavior as in the forging stage, as seen in Figure 63.
Figure 63: The hardness profile of the heat treated weld area in the ring direction, from the center of the weld and outwards.
There was not a completely equal hardness through the thickness after the heat treatment, at least not in the center of the thickness. Figure 64 shows that the edges were harder than the middle with the inner radius the hardest of the two.
Figure 64: The hardness profile of the heat treated specimen in the radial direction, from the outer radius and inwards.
4.7 Calibration
Figure 65: Pictures in different magnifications of the fin- ished ring at section a, b, and c in spot 1.
Microstructure
The grains in position 1 and 2 seen, in Figure 65 and Figure 66 respec- tively, differed in the sample taken furthest from the machine (7a1).
The two closest to the machine have quite small equiaxed grains while the outer one had bigger elongated grains at the edge.
Figure 66: Pictures in different magnifications of the fin- ished ring at section a, b, and c in spot 2.
Figure 67: Pictures in different magnifications of the fin- ished ring at section a, b, and c in spot 3.
There was also a difference in the microstructure in the lower part of the HAZ in the specimen. One can see in Figure 67 that outer areas of the specimen had equiaxed grains with a small grain size, and the cen- ter of the part had angular irregular grains.
The lower part of the weld still had deformed grains with the middle area of the part the most deformed, shown in Figure 59.
Figure 68: Pictures in different magnifications of the fin- ished ring at section a, b, and c in spot 4.
Figure 69: Pictures in different magnifications of the fin- ished ring at section a, b, and c in spot 5.
The low magnification image in Fig- ure 69 shows that there was a dis- tinct line between the HAZ and the area further from the weld.
The HAZ had relatively big grains and the unaffected area had small grains. Furthermore, the grains in 7a1 and 7c1, Figure 69, shows equiaxed grains while the center (7b1) had angular, deformed grains.
Figure 70: Pictures in different magnifications of the finished ring in spot 6.
The upper area of the weld showed the same microstructure in all three spots examined, apart from the rest of the calibrated piece.
They all had the structure shown in Figure 70, with grains grown in random directions.
Hardness
Figure 71: The hardness profile of the finished product in the ring direction, from the center of the weld and outwards.
The hardness in Figure 71 is the hardness of the finished ring. It appears that there was a slightly higher hardness in general around the weld than in the ring unaffected by the processes, apart from rolling and calibration.
There was, though, an area with a relatively high hardness peak between ca 1.5 and 3.5 mm from the center of the weld.
5 Discussion
In all practical tests and measurements there is an uncertainty of how accurate the results are.
The testing done in this work is no exception. The normal thing to do in practical testing is to do many equal tests, and from that take an average value that represents the population.
This work is time limited and such many tests could not be made. This examination is done on only one sample per process, one can therefore not be sure that the results are significant to all the manufactured rings. Though, the idea in an automated production line is that all produced products should be in the same condition, making it a reasonable assumption that the results are somewhat representative in this case.
One thing that can confirm the results of this test is the test that was made earlier at Ringsvets.
The results in that test were not completely equal, but very similar to the results obtained in this work. The microstructure seems to be very alike, but the hardness differs. The reason why there is some difference may be the geometry of the rings, which is not specified in the examination done before. It might also be a difference in external conditions that alters the result. One difference that most likely changes the outcome is how the examination is performed, and that is confirmed to differ between the earlier and the present examination.
5.1 Microstructure and preparation
The margin of error in the captured images are mostly human errors. The pictures are meant to be taken at the same point in all corresponding samples. As the spots are chosen by the person taking the pictures, they can be a bit off between the different specimens, and thus, making it impossible to compare if they are not taken in the same area.
There is not much wrong one can do in the cutting and hot mounting, except from enabling the heat getting too high and alter the microstructure. The polishing, however, is a critical part of the preparation where it is very important to use the correct method and equipment. If the abrasives are used with a too big grit difference, or not polished for a sufficiently long time, the scratches from the previous stage might not be eliminated. Similarly, if the equipment is not cleaned enough between each stage, abrasive particles can be left on the surface and scratch it in later stages. The scratches would then be too severe for the later stages to eliminate and the result would thereby be of deficient quality.
Etching the sample is also a stage where a good result can be ruined. By inadequate etching of the sample the microscope might not be showing what is wanted. By etching the sample too much one can destroy the surface resulting in, like inadequate etching, inexplicit indications.
In Abrams method of grain size measuring one marks all spots where a ring cuts through a grain boundary. As that method involves the perception of what is a grain boundary, the measurement error can be severe. One should therefore be a little bit restrictive with how much one thrusts the results. The measurements are taken in three spots creating an average to reduce the error, but there is always a possibility that the value is incorrect.
5.2 Hardness tests
When doing a microhardness test the diamond tip can hit grains and inclusions with harder or softer structure. The hardness can therefore, as seen in the results, vary a lot between spots close to each other. One can therefore not depend on one single spot to evaluate the hardness.
The trend will, however, be pretty accurate to the actual hardness when doing a hardness profile, as done in this case.
The method of Vickers hardness testing involves a person in the work of measuring the hard- ness. That means that the person doing the tests plays a big role in the result of the measure- ment. Different people have different perceptions of where the edges of the indent are which can make the measurement differ on the same sample. The measurements in this work are done by the same person, minimizing the fault of perception. The exact value might not be, or is most likely not, completely accurate, but the possible fault is the same in all tests (as equal as it can be, considered the human error).
A problem related to the human error is that the samples where etched when the hardness tests were executed. The consequence of that can be that the outlines of the indentation might be hard to distinguish from grain boundaries and other irregularities.
Apart from the human error there is also a possible source of error in the machine. The hardness value can deviate from reality if the machine is broken or calibrated incorrectly.
Even if the test was perfect, one cannot be sure to have the correct result as there is some room for error in the compilation of data. The tests are done with an old computer without USB ports and internet connection. The results are in this case transferred from the test computer to another computer via a floppy disc. The data can then be transferred to, for example, a USB memory before it is compiled on a third computer. Another consequence of using an old computer is that the result file is not compatible with any program available in the computer used to compile the data. The results are therefore manually transferred from the result file to a program in which graphs can be made. It would not be a too big surprise if a couple of values has been mixed up somewhere on the way.
