A study on the shape correction for stamped product of high strength steels by laser irradiation

LIST OF TABLES Table 1 Chemical composition of HSS material………...………73 Table 2 Preparation of Specimen to Investigate the Effects of Size……...73 Table 3 Process Parameters for Investigat

M.S Thesis

A STUDY ON THE SHAPE CORRECTION FOR STAMPED PRODUCT OF HIGH STRENGTH STEELS BY LASER

IRRADIATION

Thanh Hai Nguyen

The Graduate School YEUNGNAM UNIVERSITY

Department of Mechanical Engineering Mechanical Engineering Major

Advisor: Prof Hyun Bo Shim

ABSTRACT

In automotive industries, use of high strength steels (HSS) is rapidly increasing since high strength steels can meet the requirement of automotives, such as light weight of the structure, low fuel consumption, and safety of passengers, etc Most automotive parts, including bumper part, can be manufactured by sheet metal forming, where stamping process is the

typical technology The stamping of high strength steels makes difficult to develop the stamping dies due to low formability and large springback compared to low strength steels As a result, it demands longer period of product development than those of the stamping of conventional steels Besides, the basic purpose of manufacturing is to create the products having the severe competition in the automotive market Therefore, shortening the period of product development and reducing cost are strongly demanded

In order to meet both the requirement of automotive industries and the difficulties of stamping processes, a new heuristic method which is combined use of stamping and laser irradiation has been proposed in this study

To verify the method, corresponding experiment and a series of numerical simulations, irradiation process, stamping process and springback were carried out Springback amount can be transferred into certain angles Based on the trend of springback angles, the desirable laser bending angles are definitely obtained Moreover, the condition of laser irradiation in order

to achieve the desirable bending angles can be derived based on the simulation analysis of laser irradiation using flat plate Applying the laser irradiation patterns, the springback angles were significantly compensated within acceptable tolerance

This work still remains some things being done in the future such as finding the real temperature dependent materials of high strength steels, applying laser irradiation to correct springback in the real bumper parts, ect

TABLE OF CONTENTS

1 Introduction

1.1 Problem statement ……… 1

1.2 Springback correction methods……… 2

1.3 State of the Arts……… 5

1.4 Scope of Study………8

2 Simulation Analysis 2.1 Coupled thermo-mechanical analysis in LS-DYNA…….…… 10

2.2 Model development for laser bending of flat sheet……….…… 11

2.2.1 Geometry……… ………….……….………11

2.2.2 Mesh development 11

2.2.3 High strength steel material properties……….12

2.2.4 Laser beam heat source……… ………… …13

2.2.5 Thermal boundary conditions……….……….….…14

2.3 Temperature distributions……….……… … 15

2.4 Effects of sizes of specimen……….……… … 16

2.5 Effect of eccentric laser scanning lines……… …… 17

2.6 Effect of power and speed……….….……….… 18

3 Laser Irradiation Experiment of Flat Plate 3.1 Preparation of Experiment……….………….………… 20

3.2 Set up model for laser irradiation……….… ….….……22

3.3 Measurement of deformed shape……… ………24

3.4 Microstructure analysis……… ……… 25

3.5 Discussion……… ……… 25

3.5.1 Effect of width of specimens……… ……….25

3.5.2 Effect of length of specimens……… ………….… 27

3.5.3 Effect of eccentric laser scanning lines………… ….…….28

2.5.4 Effect of laser beam speed………… ……… …….…….29

3.5.5 Effect of line energy………… ……….…….30

4 Application to Correct Real Bumper Parts 4.1 Stamping and springback analysis………… … ………….… 31

4.2 Springback correction……… ……….….32

5 Conclusion and Future Work 5.1 Conclusion……….……… … 38

5.2 Future work……….……….….….…38

REFERENCES……….……… … 39

APPENDIX……….……… ……76

ABSTRACT……… …… 84

a) Change of Temperature after Laser Scanning

b) Temperature Distribution at the Section Cut Across Center Line

Fig 7 Effect of Length……… ……… 48 Fig 8 Effect of Thickness……… ……… 48 Fig 9 Effects of Scanning Position……… ……… 49 a) Bending Angle

b) Torsion Angle

Fig 10 Effect of Cooling Time……… ……….……….50 Fig 11 Effect of Power and Speed……… …………51

a) Scanning Speed 1m/min

b) Scanning Speed 2m/min

Fig 12 Mean Bending Angle……… ……….52 Fig 13 Model for Measurement Initial Shapes……… ……….53 a) Calculation Model for Measurement Initial Shape

b) Experiment Set-up for Measurement Initial Shapes

Fig 14 Measured Initial Deflection of Blank (200mm x 60mm)…… … 54 Fig 15 Laser Irradiation Path……… ……….55 a) along the Center Line

b) along the Eccentric Line

Fig 16 Experiment Set-up for Laser Irradiation………… ………56 Fig 17 Measurement of Deformed Shapes after Irradiation………… … 57 a) Scanning Line to Measure Deformed Shape

b) Experiment Set-up for Measurement of Deformed Shapes

Fig 18 SEM Photograph after Laser Irradiation (800 W, 1m/min)… … 58 a) SEM Micrograph at the Scanning Line

b) Microstructure around the Boundary between the Heat Affected Zone and Untransformed Zone

Fig 19 Effect of Specimen Width (Length 300mm)………… ………… 59 (a) Measured Deflection after Irradiation

(b) Bending Angle

Fig 20 Effect of Specimen Length (Width 60mm)……… ………60 Fig 21 Effect of Eccentricity (300mm x 60mm)……… ………… 61 (a) Bending Angle

(b) Torsion Angle

Fig 22 Effect of Scanning Speed……… ……… 62 Fig 23 Effect of Line Energy……… ……….63 Fig 24 Stamping of Bumper Shape Part……… ………64 (a) Part Shape

(b) Part Shape Parameters

(c) Stamping Die Set up

Fig 25 Deformation during Springback (BHF: 20KN)……… …….65 (a) X-displacement during Springback

(b) Z-displacement during Springback

Fig 26 Model for calculation springback angles……… ………… 66 a) Springback

b) Required Bending Angle

Fig 27 Required Bending Angle to Compensate Springback at Various Blank Holding Force……….67 Fig 28 Irradiation Bumper Part Model……… ……… 68 Fig 29 Application of Laser Bending to Compensate Springback (Blank Holding Force 20KN)……… 69 a) Finding Scanning Length

b) Bending Angle Interpolation

Fig 30 Springback Laser Scanning Patterns……… ……… 70 a) Scanning Pattern, Speed 1m/min

b) Scanning Pattern, Speed 2m/min

Fig 31: Springback Compensation……… ………71 Fig 32 Power Time Dependence……… ……… 72

LIST OF TABLES

Table 1 Chemical composition of HSS material……… ………73 Table 2 Preparation of Specimen to Investigate the Effects of Size…… 73 Table 3 Process Parameters for Investigating the Effects of Laser Beam Speed, Power and Line Energy… ……… 74 Table 4 Specimen Parameters for Investigating the Effects of Eccentric and Diagonal Laser Scanning Line……… ……… 74 Table 5 Parameters for Springback Correction by Simulation……….75

CHAPTER 1 INTRODUCTION

1.1 Problem statement

It is known that a holistically designed steel body structure that meets tough structural and crash criteria while weighing less and costing no more than typical vehicles in its class

ULSAB-AVC (Ultra Light Steel Auto Body-Advanced Vehicle Concepts) is the global steel initiative which offers steel solutions to meet society’s demands for safe, affordable, fuel efficient, environmentally responsible vehicles for the 21st century Since the wide applications of high strength steels (HSS) are keys to the realization of ULSAB, the use of HSS

is rapidly increasing in the automobile industries

Most automotive parts, including bumper part, are produced through the stamping process As far as the stamping is concerned, HSS makes difficult

to develop the stamping process due to its low formability and significant amount of springback compared to conventional steel

Despite the technical difficulties, the shortening of development period and cost reduction are also demanded, since the automobile industries competes with global challengers

Due to the environment surrounding the automobile industries, a hot press forming process, especially for the HSS, has been introduced The hot press forming is a stamping process of HSS, where simultaneous stamping

and heat treatment are carried out during the stamping by using the heated sheets up to 920 ℃ and controlled cooling at the die The advantages of hot press forming are such as complex forming, lower press tonnage and lower

or near zero springback, etc However the requirement of an extensive scale

of investment can be the main drawbacks of the hot press forming

Contrarily to the hot press forming, the cold stamping or conventional stamping is well developed and widely used process, except for the springback control of HSS

This study is a motivated combination of cold stamping and posterior shape correction by laser irradiation instead of the hot press forming Due to the limitation of low formability of the HSS, the major stamping process of HSS parts is forming, where bending is dominant to minimize surface strain, rather than drawing in which significant surface straining is involved Since the low formability problem can be ignored, remaining technical problem of HSS part stamping is to control springback The shape error caused by the springback can be corrected by the laser irradiation

1.2 Springback correction methods

In metal forming, a certain amount of springback is always accompanied

in greater or less degree, because of the elastic recovery of the material Especially in stamping processes, the springback causes shape error compared to the die shape It is influenced not only by the tensile and yield strength but also by the thickness and bending radius

Due to many advantages considered in advance, high strength steels are

used as dominated materials in automotive industries Because of larger yield strengths, parts made of HSS demonstrate more springback than parts made of mild steel Thus, understanding, prediction, control and reduction

of springback behavior have become very crucial in terms of decreasing development times

Numerous studies have attempted to determine the controlling factors in springback and find ways to overcome the drawback of using HSS in stamping processes Gang Liu [1] has proposed a method to use the variable blank holding force to eliminate the springback error in U–shape part The wrinkling limit and fracture limit in the forming process has been considered to obtain the blank holding force trajectory Nielen and his partners [2] developed an optimization method based on the die-optimality criterion for experimental approximations to minimize the difference between the simulation results and the intended design Using geometric variables such as critical locations where tool elevation should be modified, important tool radii and starting blank geometry, a set of stamping and springback models can be created automatically After first set of simulations, optimized values for each design variables are predicted The next set of simulations is automatically created using these variables This iterative process continues until every variable are determined within a specified tolerance or until a limiting number of iterative have been completed Jernberg [3], Dutton [4] and Lingbeek [5] have proposed a method based on the springback compensation tools This method is divided into two stages The first one is to create stamping dies based on the exactly designed shape The stamping products obtained by using these dies do not

completely coincide with the desired shape due to the springback appearance The second stage is to calculate the shape difference between the springback profile and desired profile The new stamping dies are repeatedly generated based on the difference The desired shapes are created exactly It is very easy to recognize that new compensation tools must be generated for each kind of parts Therefore, it took a long time for designing and manufacturing the compensation tools and huge cost was also needed

To overcome the limitation above, laser irradiation method was proposed in this study as a method to compensate the springback behavior The laser irradiation is already known as a method to form the sheet metal Many authors [6-10] have proposed a method to form sheet metals using laser irradiation They investigated the effects of processing parameters to the bending angles using short specimens by both simulation and experiments The bending angles were supposed as constant values along the longitudinal direction The effects of first scanning and multi scanning to the bending angles were also studied Bao [11] has investigated the effects

of edge to the bending angles The edge is found to be somewhat curved and the bending angles varied along the laser scanning path These phenomena are termed edge effects

Based on the survey of laser forming, it seems to be a promising method

to compensate the springback behavior after stamping of sheet metals for high strength steels

1.3 State of the Arts

Comparing to modifying tool geometry method, laser irradiation method had many advantages There were no tools used in this method, thus, time for designing and manufacturing tools were reduced The cost was also significantly reduced Besides, this method was more flexible and simpler than modifying tool geometry method Moreover, the size of specimens had

a widely changeable range while it was limited by the limitation of size of tools in modifying tool geometry Other hand, the equipments of laser irradiation method were so simple included laser machines

First and most important thing is to understand completely the laser irradiation process Laser bending process is a promising technology in manufacturing industries such as automobile, shipbuilding and airplane industries Three mechanisms are described that are generally considered to

be the most significant in demonstrating the laser forming behaviors

Temperature gradient mechanism

The temperature gradient mechanism (TGM) is the most widely reported laser bending mechanism, shown in Fig 1a) Due to the rapid heating of the surface by a laser beam and the slow heat conduction into the sheet, a thermal gradient into the material results a different thermal expansion through the thickness As the material is heated, initially the thermal expansion on the heated surface is greater than that on the cold surface Counter – bending occurs due to the bending moment created, causing in a

small amount of plastic tensile strain at the heated surface With continuing heating, the bending moment opposes the counter – bending away from the laser beam, and the mechanical properties of the material are reduced with temperature increase Once the thermal stress reaches the temperature – dependent flow stress, any additional thermal expansion is converted into a plastic compressive strain because free expansion is restricted by the surrounding material During cooling, the material contracts the upper layers again, and since it has been compressed There is a local shortening of the upper layers of the sheet, and a bending angle developed that bends the specimen towards the laser beam

Buckling mechanism

The second mechanism is buckling mechanism (BM) as shown in Fig 1b) In this case, the laser beam is usually much larger than the sheet thickness The diameter of the heated area is equal to the sheet thickness for the TGM, while this diameter is ten times the sheet thickness for the BM There is no temperature gradient through the sheet thickness The BM is activated by the laser parameters that do not yield a temperature gradient in the vertical direction Due to heating, thermal compressive stresses develop

in the sheet which results in large amount of thermo-elastic strain which in turn results in local thermo-elasto-plastic buckling of the material This buckle is generated along the moving direction of the laser beam scanning When the laser beam leaves the sheet surface, the buckle is generated across the whole sheet The direction of the bending angle is not defined by the

process itself as it is for the TGM The part can be made to bend in either the positive or the negative vertical directions depending on a number of factors including the process parameters, the prebending orientation of the sheet, the pre-exiting residual stresses, the direction in which any other elastic stresses are applied, internal stresses, and external or gravitational forces These parameters define the direction of the buckling in a complex manner However, it is possible to control the BM in such a way to make it reliable forming process

to an increase on overall thickness

Secondly, the shape difference between the shape after springback and the desired shape were transferred into certain springback angles Based on the method of calculation angles, the laser scan is irradiated at particular position The process parameters of laser irradiation such as power source, speed of laser beam, line energy can be carefully controlled to compensate

the springback angles

1.4 Scope of Study

The purpose of this study is to investigate and evaluate the effects of process parameters of laser irradiation, and then to demonstrate a methodology for springback correction It might be known that the laser irradiation can cause bending and the amount of bending angles can be controlled by controlling power and speed of laser beam, called process parameters Based on the measured shape errors, obtained by simulation, the irradiation patterns can be determined In order to verify the method, corresponding experiment and a series of numerical simulations, irradiation process, stamping process and springback were carried out Based on the results of experiment and simulation, optimal irradiation patterns were derived

The basic study methodology is to develop a computational thermal model for laser irradiation based on experiment data The properties of high strength materials were measured at the room temperature while other material properties of temperature dependence were assumed based on the trend of some published papers [6] The study is mainly focused on the experimental study to get a target angle over the longitudinal direction and the experimental findings are further investigated through a thermal modeling of laser irradiation to find out the effects of process parameters Controlling these parameters is to obtain the desired bending angles Consequently, applying the designed process parameters of laser irradiation

to the springback shape is to obtain the desired bending angles for compensating the springback problems

CHAPTER 2

SIMULATION ANALYSIS

2.1 Coupled thermo-mechanical analysis in LS-DYNA

LS-DYNA can solve steady state, transient and coupled thermo-mechanical analysis The options for setting the required type of analysis can be done in the *CONTROL_SOLUTION and

*CONTROL_THERMAL_SOLVER cards The laser bending process is nonlinear, thermal mechanical coupled problems The selection of the required solver is of high interesting in getting the analysis as linear or nonlinear and selecting four different solvers This work uses symmetric direct solver (ACTOL) which is a Gauss type profile solver Since the attempted problem is nonlinear, *CONTROL_THERMAL_NONLINEAR card is also used The default LS-DYNA analysis is explicit analysis and is often used for structural analysis However, transient thermal problems can

be solved by implicit analysis The implicit time integration is stable and hence larger time steps can be applied for thermal problems Thermal time step size can be set in the *CONTROL_THERMAL_TIMESTEP card and must be equal to structural time step size In laser irradiation, the specimen

is placed on the flatform which is modeled by using finite planar rigid walls

including in *RIGIDWALL_PLANAR_ID card

2.2 Model development for laser bending of flat sheet

2.2.1 Geometry

The geometric models are generated by CATIA software The specimen was modeled having the dimensions of 300x60x1.4mm3 It had four-side surfaces and two top and bottom surfaces

2.2.2 Mesh development

The mesh system of the specimen was created by HYPERMESH software Three-dimensional solid quad elements were used to mesh the sheet A total of 4800 elements (60 by 16 by 5 grid-points along the length, width and thickness, respectively) was generated with mesh density being fine at the laser scanning line and gradually increased in coarseness away from the laser irradiation line as shown in Fig 2 Along the thickness, bias options with biasing intensity of 20 were used to create the coarseness layers away from the top surface to the bottom surface Along the longitudinal direction, equal step, 5mm, between two consecutive nodes was set up Due to establish the thermal boundary conditions, sets of quadrilateral segments were generated on the top, bottom and four-side surfaces using the same nodal identifications which were used to create solid quad elements in advance In order to create platform, the simplest shell element was generated with only four nodes so that the specimen can

be placed on completely The laser beam was modeled as a diamond shape

with six nodes and eight shell elements The laser heat source was created from the lowest node of the diamond shape which was contacted directly to

the top surface of specimen

2.2.3 High strength steel material properties

The material used in experiment and simulation was high strength steel (HSS) Due to the difficult of investigating the exactly properties of materials which depend on temperature, the tensile test was only taken place

at the room temperature The chemical composition of HSS can be seen in Table 1 The HORIBA/EX-250 machine type was used to measure the chemical composition The measured carbon component was slightly higher than real component of the specimen due to the low accuracy of this machine The other components were measured exactly

Tensile test

The specimen for tensile test was designed as in Fig 3 a) The gauge length 50mm was used The thickness and width were 1.4 mm and 12.5 mm, respectively Room temperature was established at 160C The engineering stress – engineering strain and true stress – true strain relations were shown

in Fig 3 b) From true stress – true strain curve, yield stress and Young’s modulus were calculated at 421 MPa, 176.13 GPa, respectively The strength coefficient and hardening exponent were calculated at 956 and 0.182, respectively

Other parameters

To calculate the hardening modulus, the true stress – true strain curve was used A linear line connecting the yield point and the highest point on the curve was drawn This line had a slope to the true strain axis The hardening modulus must be the slope of this line and its value was 1188.8 MPa The properties at higher temperature were assumed base on the tensile test of real materials at room temperature and the trend of material properties described et al [6] The HSS material properties can be seen in Fig 4

2.2.4 Laser beam heat source

LS-DYNA uses a laser heat source model developed by J Goldak [12] The Goldak laser heat source model, shown in Fig 5, is based on the Gaussian distribution of power density in space A significant feature of the model is that a double ellipsoidal energy deposition profile is used Therefore, the size and shape of the laser heat source can easily model the deep penetration laser beam The Goldak model is described by the equations as follow:

Where Q is power density, W/m3, v is speed across the surface,

The laser shape parameters (a, b, cf, cr) are defined using

*BOUNDARY_THERMAL_WELD keyword in the LS-DYNA input file

2.2.5 Thermal boundary conditions

In general, temperature in a structure element can be calculated by solving the following heat transfer equation:

The specimen is freely placed on the platform The laser scanning was taken place on the top surface along the center line Therefore, the convective heat losses occurred on the top and four-side surfaces while conduction dissipation happened from the bottom surface to the platform

To express convective boundary conditions on free surfaces, equation (2.3) can be derived as follows:

convection coefficient, k is the thermal conductivity and T∞ is the ambient temperature In this present model, h was chosen to be 5 W/(m2.0C) using an ambient temperature of 230C for top and side surfaces of the specimen

Due to the relatively small difference between the process temperatures and ambient temperatures, the effect of heat dissipation causing by radiation is ignored

2.3 Temperature distributions

The TGM is based on the temperature difference between the top surface and the bottom surface along the thickness The change of temperature during laser irradiation, therefore, must be investigated The power 800W and speed 1m/min was used

Along thickness direction

Along the thickness, 5 layers were generated using bias option The temperature distribution along the thickness is shown in Fig 6 The maximum temperature 13290K occurred on the top surface and gradually decreased to the minimum 295.860K on the bottom surface Therefore, the difference was 10340K Due to the thermal boundary conditions, the laser beam moved forward making the temperature reduced significantly As first 0.1 sec passing through, the temperature on top surface is reduced 6680C After first 0.2 sec, the temperature on the top surface is dropped out 9360C The shape of temperature distribution along thickness can be seen visually

in Fig 6 b) It is clear that the temperature distribution did not contain the bottom surface Therefore, with these kinds of power and speed, the heat applied on the top surface can not affect the bottom surface

2.4 Effects of sizes of specimen

Three kinds of specimens were used with the length 100mm, 200mm and 300mm The width and thickness were 60mm and 1.4mm, respectively The effects of length of specimen can be seen in Fig 7 It describes that in the first 50 mm, the trend of bending angles was similar to one another, involving increasing region After this stage, the bending angles tend to decrease significantly Since the length is long enough, the steady region can

be seen in case 300 mm in length

The thickness of the specimen can affect the bending angles Therefore, its effect should be investigated Three kinds of thickness, 0.6mm, 1.0mm, and 1.4mm were used in simulation The length and width of specimens were 300mm and 60mm, respectively The Fig 8 shows the effects of thickness to the bending angles It can be seen that increasing the thickness, the bending angles increased strongly This can be explained that the bending angles were created based on the TGM The temperature on the top surface and the bottom surface differs a lot in the large thickness compared to the small thickness Besides, it is also known that the large difference temperature can cause the large bending angles

2.5 Effect of eccentric laser scanning lines

All cases above were taken place at the center of the specimens It is necessary to know the effects of other positions of laser scanning lines such

as eccentric lines In order to investigate the effects of laser irradiation of eccentric lines, three positions of the laser scanning lines were considered The positions of laser scanning line were located at the distance 10 mm and

15 mm from the center line of the specimen, respectively The power 800 W and speed 1 m/min were used The effects of eccentric laser scanning lines

to the bending angles can be seen in the Fig 9 a) It is clear that the total bending angles of the specimens which the laser scanning lines located near center are larger than those specimens which the laser scanning lines located far from the center The trend of angles in eccentric laser scanning lines can

be divided into three separated regions The angles increase rapidly and obtain the steady region before decrease significantly As moving the laser scanning line from 0 mm to 15 mm far from the center line, the bending angles are reduced slightly while the trend of angles still remains

Scanning at the eccentric lines can cause torsion angles which are defined as the different between the left angles and right angles of the scanning line The torsion angles can be seen in Fig 9 b) It illustrates clearly that scanning far from the center, the torsion angles are much larger than those near the center line Therefore, to avoid the torsion occur, the scanning line must be located at the center line

2.6 Effects of power and speed

In order to investigate the effects of power and speed, size of specimen, 200x60x1.4mm3, is used Same kind of mesh system was modeled as in the way of generating with size 300x60x1.4mm3 in advance The total of 3200 solid quad elements (40 by 16 by 5 grid-points along length, width and thickness, respectively) was created

In the simulation analysis, the range of power from 200 W to 800 W was concerned Two types of speed were used, 1 m/min and 2 m/min The cooling time is selected based on the effect of cooling time to the bending angles, Fig 10 It describes that increasing the cooling time from 6s to 10s, the bending angles increase slightly Increasing cooling time to 12s, the bending angles are similar to the bending angles created in case 10s Therefore, the cooling time for laser irradiation is selected at 10s for every case

The trends of bending angles along the longitudinal direction are illustrated in Fig 11 In both cases of speed, 1 m/min and 2 m/min, and power 200 W, the steady bending angles along the longitudinal direction can

be seen due to extremely small bending angles In case of 1 m/min, the trend

of angles is obtained the convex shape It increases on the first half, achieves the maximum at the center and then decreases on the second half

of specimen In case of 2 m/min, the angles gradually increase from the starting point to the ending point That means the bending angles increase as scanning time increases In addition, with the same power, the angles created by speed 2m/min are slightly smaller than those created by speed 1

m/min

In two cases above, the bending angles changes slightly along the longitudinal direction Therefore, the mean angles along the length can be derived while keeping the meaning The average bending angles according

to the power, thus, can be drawn in Fig 12 a) It is clear that the bending angles are proportional to the power Moreover, the angles increase significantly as increasing the speed in case of same power The relation between bending angles and line energy is also described in Fig 12 b) It has been known that line energy characterizes the change of both speed and power simultaneously It is calculated by the ratio of power and speed The figure shows that with same line energy, the bending angles can not be same

It depends on the speed The more speed is, the more bending angles are

CHAPTER 3 LASER IRRADIATION EXPERIMENT OF FLAT PLATE

3.1 Preparation of Experiment

Rectangular flat specimens have been prepared by laser cutting method from sheet The oil, grease, dust, rusts, etc on the surfaces of specimens were completely removed using acetone in order to maintain the consistent heat absorbability over all specimens Since the specimen has been cut from cold rolled steel coils, curvature also existed initially In order to improve the accuracy of the experiment, the shape of the specimen has been measured

The specimen was placed freely on the basement of 3D measurement machine in order to measure the initial shape, shown in Fig 13 b) The specimen was limited by the two clamps in order to maintain the same position for every case To measure the initial shapes, three left, center and right lines were remarked on the top surface The left and right lines were located in the distance 2 mm far from the nearest edges while the center line was located in the center of the specimen Coordinates at each node along the three lines were measured The number of nodes along the longitudinal direction in three lines must be equal The distance between two consecutive nodes on each line was 10 mm Processing the nodal coordinates, the initial shapes of each kind of specimens can be drawn with the mean values of selected specimens They can be seen from the Fig 14

The initial shapes of the specimens were not flat and thus, the different amount in vertical direction between left line and center line, right line and center line always occurred To characterize these differences, two angles,

α1i and α2i are defined in XZ plane which contains three nodes in left, center and right lines and normal to the specimen, as shown in Fig 13 a)

The initial angles α1 and α2 are calculated by the equation (2.1)

1 2

''''

i

i

CC ATAN

AC BB ATAN

AB

αα

Where CC’ the distance between the point on left line and its projection

on the basement, BB’ the distance between the point on right line and its projection on the basement

It must be noticed that these two angles are created due to the natural curvature of initial specimens after cut from rolled coil

All the initial shapes have concave profiles This can be explained that these specimens have the effects of original specimens which came from the rolled workpieces However, these effects will not affect the results since they will be compensated in calculation approach

In order to investigate the effects of sizes of specimens, many kinds of different specimens were designed to be irradiated To study the effects of length, the range of length sizes from 100 mm to 400 mm were created while keeping the width of specimens, 60 mm Moreover, different widths

of specimens from 50 mm to 150 mm while remaining the length, 300 mm, were conducted to observe the effects of width The effects of width and length of specimens must be conducted with same powers and speed, 800 W

and 1 m/min, respectively The specimen parameters for studying the effects

of sizes of specimens were described in Table 2

The laser irradiation process is mainly characterized by two factors, power and speed of laser beam According to Kim’s work [10], one more factor can be used, line energy which is defined by the ratio of laser power and speed of laser beam In this work, the effects of power, speed of laser beam and line energy to the bending angles were also studied In order to understand the effect of powers to the laser scanning, different powers were considered while speed of laser beam was kept constant Other hand, maintaining constant power and changing speed of laser beam is to study the effects of velocities of laser beam Both powers and speed changed, the effects of line energy were concerned The process parameters were designed in the Table 3 The same sizes of specimens, 300x60x1.4mm3, were used for every case

All cases above are conducted at the center line of the top surface of the specimen, Fig 15 a) It was necessary to know the effects of position of laser scanning line Therefore, the irradiation in eccentric lines was also taken place according to model in Fig 15 b) and the factors can be seen in Table 4 In this case, 800 W for the constant power and 1 m/min for constant laser beam speed were used The same sizes of specimens, 300x60x1.4mm3, were also used for every case

3.2 Set-up model for laser irradiation

Until this stage, the specimens were completely removed out of dust,

oil, rust, etc and ready for laser irradiation experiments Moreover, the initial shapes were also known Due to the lack of equipments and facilities, measuring the deformed shapes was conducted offline after laser irradiation definitely finished That means the deformed shape and bending angles due

to laser irradiation could not be measured directly during the laser irradiation experiment The bending angles could only be calculated through measurement the deformed shapes and the initial shapes The laser irradiation experiments were conducted on the laser machine type TrumPF Laser HL 1006 The maximum available power is 1000 W The time to increase the power from 0 to desired value is 0.5 ms The experimental set-up models for laser irradiation in experiments can be seen in the Fig 16.The specimens were placed freely on the basement of the laser scanning machine The materials, thus, can expand in every direction during laser experiments as much as possible The specimens were limited by two clamps fixed on the platform to locate exactly position of specimen for every experiment In the laser irradiation, the laser beam irradiated along the center or eccentric depending on what kinds of laser scan used The laser beam always irradiated consecutively the whole path from one end to the other Constant power and speed of laser beam were always used for each case During laser irradiation process, the specimens were deformed without any constraints The laser machine took 0.5 ms to reach the desired power Besides, the CO2 gas was also involved during the stage of power input to avoid making oxidization The focal diameter was always remained at constant values, 6.5 mm, on the top surface by keeping the invariant height between the laser head and the top surface of the specimen

3.3 Measurement of deformed shape

Due to lack of equipments and facilities, the deformed shapes were measured after finishing laser irradiation at different place Similar to the measurement initial shapes, the set-up model for measurement deformed shapes is shown in Fig 17 The deformed shapes were also measured in left, center and right lines as remarked in measurement initial shapes in advance The output of measurement approach was also coordinates at points which had the same quantities and distance between two consecutive points as setting up in measuring initial shapes As a result, the deformed shapes were easy to be drawn The bending angles were calculated based on the initial shape and deformed shape Difference from measurement initial shapes, the angles α1d and α2d are the bending angles due to laser irradiation plus the initial angles Therefore, the bending angles due to laser irradiation can be calculated as follows

no melting zone was involved The width and the depth of heated zone were measured at 3.5 mm and 0.58 mm, respectively

3.5 Discussion

Many kinds of different specimens were designed to be irradiated along the longitudinal position in order to investigate the effects of sizes of specimens Moreover, the center and eccentric lines were also investigated

3.5.1 Effect of width of specimens

The width of specimens can affect the deformed shapes, the bending angles and the torsion angles which are defined as the left angle minus the right angle To see those effects, different widths of specimens from 50 mm

to 150 mm while remaining the length, 300 mm, were conducted to observe the effects of width The same process parameters, power 800 W and speed

1 m/min were used

The effects of width of specimens to the curvature of deformed shape can be seen in Fig 19 a) To obtain the radius of curvature, two-end nodes and center node were supposed belonging to certain circle The radius of this circle is the radius of curvature The radius of center curve of deformed profiles in case of width 50 mm, 100 mm and 150 mm were 3904 mm, 3233

mm and 5685 mm, respectively Increasing the width from 50mm to 100mm, the curvature radius decreases slightly Width is continuously increased, the curvature radius increases again It can be explained when increases the width, the mass of materials at two sides of laser scanning line also increases It limited the displacements in vertical direction Thus, the curvature radius increases rapidly

Besides the effects to the curvature, the effects of width to the bending angles and the deformed shapes can be seen in the Fig 19 b) In case of 50

mm in width, from the starting point to 70mm, the bending angles increase rapidly from 1.020 to 1.760 and then remain the steady values along the length until 270mm As the width increases to the values 100mm, the bending angles also increase After that, if width is continuously increased

to 150 mm, the bending angles decrease slightly That means the bending angles get the stable trend at over 100 mm in width Besides, increasing the width, steady region was reduced and enlarged the proportional region from the starting In all cases, the difference between the maximum and minimum bending angle was nearly invariant, around 0.80

3.5.2 Effect of length of specimens

Other effects of sizes of specimens are the effects of length In order to investigate those effects, the range of length sizes from 100 mm to 400 mm was created while keeping the width of specimens, 60 mm Same power 800

W and laser speed 1 m/min were used The Fig 20 shows the effects of length of specimens It can be seen that from the starting point to 100 mm in length, the trend of angles was similar one another That means there is only increasing region while steady region is not included Increasing the length from 100 mm to 300 mm increases the magnitude of bending angles After that, increasing the length to 400 mm reduces rapidly the bending angles This region is called the first increasing regime Furthermore, in the range from 200 mm to 300 mm, increasing the length also increases the bending angles while the trend still remains The first increasing regions were similar each other However, the steady angles can not be seen in the length 100

mm and 200 mm but they can be seen in the length 300 mm from a distance

70 mm to 150 mm After the steady region, the bending angles decreases rapidly Increasing and decreasing regions are nearly symmetric through the steady region in case of 300 mm in length When the length reaches up to

400 mm, the trend of bending angles differs slightly in magnitude and trend The values decrease as increasing the length from 300 mm to 400 mm The slope-up also occurs in the starting region while the steady region could not

be seen The decreasing region at the other end also happens

3.5.3 Effect of eccentric laser scanning lines

All cases above were carried out at the center of the specimens It is necessary to know the effects of other positions of laser scanning lines such

as eccentric lines To see the effects of laser irradiation of eccentric lines, three positions of the laser scanning line were considered The positions of laser scanning line were located at the distance 15 mm and 20 mm from the center line of the specimen, respectively The power 800 W and speed 1 m/min were used The effects of eccentric laser scanning lines to the bending angles can be seen in the Fig 21 a) The total bending angles of the specimens which the laser scanning lines located near center are larger than those of specimens which the laser scanning lines located far from the center The trend of angles in eccentric laser scanning lines can be divided into three separated regions The angles increase slightly and then decreased gradually until 180mm and increased again As moving the laser scanning line from 15 mm to 20 mm far from the center line, the bending angles reduce rapidly while the trend of angles still remains When the laser scanning line reaches the center, the bending angles continue to increase and the trend changes slightly Instead of decreasing after increasing at the second region, the bending angles get nearly steady values, and then decreases at the third region

Scanning at the eccentric lines can cause torsion angles which are defined as the different between the left angles and right angles, Fig 21 b)

It is very clear that far from the center line, the torsion angles are much larger than those near the center line Therefore, to avoid the torsion occur,

the scanning line must be located at the center line

3.5.4 Effect of laser beam speed

To study the effects of laser beam speed, the same sizes of specimens 300x60x1.4mm3 were used The purpose to change the laser beam speed while keeping the power is to see the effects of speed Four cases of power 700W, 800W, 900W and 100W were taken place Each case, two kinds of speed were considered The results can be seen in the Fig 22 Bending angles are slightly different depending on the range of powers used In the range of power 700 W - 800 W, it is clear that increasing the speed also increases the bending angles In the case of speed 1 m/min, the trend of bending angles can be divided into three separate regions Between the proportional increasing region at the first and proportional decreasing region

at the third, the steady region is located In the case of speed 0.5m/min, the shape of bending angles looks like a sin shape with different amplitude The more power is, the less bending angles oscillate

Other hand, in the range of large power 900W – 1000W, increasing speed causes greatly change of the values of bending angles The bending angles vary slightly along the longitudinal direction It means that the steady region of angles can be created with high power and high speed The different angles between the small speed and large speed differed greatly In case of 900W, the gap was around 2.20 while 1.60 in case of 1000W

3.5.5 Effect of line energy

As mentioned before, line energy is defined by the ratio of laser power and speed of laser beam to characterize the change of speed and power simultaneously The effects of line energy can be seen in the Fig 23 It can

be seen that under 60 J/mm, increasing line energy will reduce the bending angles Over that value, increasing line energy can cause increasing angles

CHAPTER 4

APPLICATION TO CORRECT REAL BUMPER PARTS

4.1 Stamping and springback analysis

For stamping simulation, initial tool designs were prepared based on the proposed CAD model The model was set up using eta/DYNAFORM 5.6 The set-up of stamping can be seen in Fig 24 The die shape differs from the punch shape which contains binder The binders cover the whole blank at initial setting up due to hold the blank as much as possible

The rectangular blank, 300x150mm2, was used with 1.4 mm in thickness The mesh system is created by using fully integrated shell elements in which size of each element is 2 mm in width and 10 mm in length Number of points through thickness integration was set up at 9 points The total number of shell element of blank is 2280 elements

HSS material with Young’s modulus 176.13 MPa was selected The binder force was tested with variation values due to find the appropriate value that caused the smallest springback angles Speed of punch was chosen at 2 m/min The offset gap between die and punch was added more 10% of blank thickness, 1.54mm The strength coefficient and hardening exponents were set at 956 and 0.182, respectively Forming simulations were carried out using the single explicit solver The high strength steel material was modeled using material type 018,

*MAT_POWER_LAW_PLASTICITY in LS-DYNA code This is an

isotropic plasticity model with rate effects using a power law hardening rule For springback simulation, the “dynain file” method was used, where dynain file was created from the stamping simulation After springback simulation, the shape slightly changed The displacement during springback can be seen in Fig 25 It was easy to see that each node had been moved to new position after springback simulation The displacements along X direction show that the side walls at two-end regions tended to go inward the center line while the side walls at center tended to go outwards the center line The displacements along Z direction illustrate that the center region of the bumper shape tended to move downwards while the two-end regions tended to go upwards This phenomenon reduced significantly the initial radius of bumper shape

4.2 Springback correction

The model to investigate the springback amount can be seen in Fig 26 a) The left and right walls of bumper parts of each cross section made a certain angle with the Z axis after springback simulation Therefore, the springback amount can be characterized by two angles, α1 and α2, at two sides Correcting these angles to 00 will made the springback compensated These two angles can be calculated by the equation (4.1)