- Email: [email protected]
Holographic measurement of distortion during laser melting: Additive distortion from overlapping pulses
Optics and Laser Technology 100 (2018) 1–6
Contents lists available at ScienceDirect
Optics and Laser Technology journal homepage: www.elsevier.com/locate/optlastec
Full length article
Holographic measurement of distortion during laser melting: Additive distortion from overlapping pulses Peter Haglund a,b,⇑, Jan Frostevarg a, John Powell a, Ingemar Eriksson c, Alexander F.H. Kaplan a a
Department of Engineering Sciences and Mathematics, Luleå University of Technology, Sweden Westinghouse Electric Sweden AB, Västerås, Sweden c Dalco Elteknik AB, Östersund, Sweden b
a r t i c l e
i n f o
Article history: Received 1 March 2017 Received in revised form 30 June 2017 Accepted 26 September 2017
Keywords: Pulsed laser welding Laser surface melting Thermal distortion Laser forming Laser straightening Holography
a b s t r a c t Laser - material interactions such as welding, heat treatment and thermal bending generate thermal gradients which give rise to thermal stresses and strains which often result in a permanent distortion of the heated object. This paper investigates the thermal distortion response which results from pulsed laser surface melting of a stainless steel sheet. Pulsed holography has been used to accurately monitor, in real time, the out-of-plane distortion of stainless steel samples melted on one face by with both single and multiple laser pulses. It has been shown that surface melting by additional laser pulses increases the out of plane distortion of the sample without significantly increasing the melt depth. The distortion differences between the primary pulse and subsequent pulses has also been analysed for fully and partially overlapping laser pulses. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction When metal products are exposed to a temperature increase and melted locally, residual stresses and distortion are inevitable side effects of the heating and subsequent cooling process [1]. Most laser - material interactions involve some kind of permanent plastic thermal distortion, since the temperature is usually so high that the thermal stresses override the material tensile strength of the work-piece. This is an important point to consider when developing processes and designing components with respect to e.g. weld design, so an understanding of distortion is important. Thermal distortion is a complex physical process and the situation is even further complicated in the case where the laser interacts with material which is pre-stressed from mechanical processes such as cold working or thermal stresses, which is the case for most industrial applications. In cases where the distortion is unwanted, such as welding, the laser has established itself as a low heat input, low distortion alternative to traditional torch technologies but, nevertheless, some distortion will generally be apparent. On the other hand, in applications such as laser bending [2] and wire straightening [3] a distortion effect is the desired outcome. Most of the currently published research considers only continuous ⇑ Corresponding author at: Department of Engineering Sciences and Mathematics, Luleå University of Technology, Sweden. E-mail address: [email protected] (P. Haglund). https://doi.org/10.1016/j.optlastec.2017.09.053 0030-3992/Ó 2017 Elsevier Ltd. All rights reserved.
wave laser processing. In this paper the deformation of a sheet as a result of melting by individual laser pulses is studied empirically. 1.1. Distortion mechanisms Work by Zhou [4] divides distortion into in-plane shrinkage and out-of-plane bending of the component. Several researchers have used an approach where deformation is divided into three mechanisms, [5–9]: 1. Temperature gradient mechanism (TGM). 2. Buckling mechanism (BM). 3. Upsetting mechanism (UM). The buckling and upsetting mechanisms are important when the laser creates a local heating pattern where the temperature is fairly uniform through the thickness of the sheet. In both mechanisms the heating and cooling cycle experienced by the sheet causes plastic failure and eventual deflection of the material. In the case of buckling, the final out-of-plane deformation direction is generally determined by the previous residual stress condition of the material (rolling direction, etc.). In the upsetting process the material is thickened in the laser heated area (and often shortened in other directions). In the case of this experiment the temperature gradient distortion mechanism is dominant. As the name suggests this
2
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
mechanism depends upon a sharp temperature gradient from the surface to the underlying material. When the top surface of a metal plate is locally heated by a laser the material undergoes a cycle of distortion as follows: 1. Heating begins: the top surface of the plate expands locally causing the material to distort. The heated surface moves towards the source of heating (see Fig. 1b). 2. The thermal stresses exceed the yield stress of the material which is then plastically compressed in the heated zone. On cooling, this compressed zone causes the sheet to distort. During cooling the heated spot moves away from the source of heating (see Fig. 1c) and remains in this new position. This laser-material interaction can be complicated by the melting and solidification of the centre of the laser heated zone, but the basic principle remains the same. 1.2. Multiple consecutive distortion Edwardson et al. [10] has studied the cumulative distortion effects of multiple line laser melting and found that the bend angle increases for up to 60 scans. In work by Shi et al. [11] the elasticplastic deformation is further described and a new mechanism is proposed; the coupling mechanism, which is a combination of the TGM and the UM. In the work of Eriksson et al. [1] holography was used to monitor thermal distortion in real time to give information about the history of the deformation. In work by Dovc [12–14], an analytical model was developed and studied for a single pulse, this model confirms the results in [1]. Shi et al. [15] concluded that the deformation for a CW process is not purely 2D for the TGM case and state that the distortions are affected by the elastic properties of the plate. Pirch and Wisenbach [16] found that the TGM model does not explain the fact that the bending angle decreases with increasing amounts of irradiation. Several researchers e.g. [17–25] have used Finite Element simulations for modelling distortion resulting from laser irradiation processes. Accurate measurements are an important part of the validation process for such simulations. An important aspect is the influence of heat input and heating methods, which has been studied by Shi et al. [18]. In this work the penetration profile in combination with sheet thickness and hence the thermal field were chosen so that the out of plane deformation would be as large as possible with pulsed laser heating. In the current paper the transient time response of a steel blank sheet exposed to multiple laser pulses is analysed by holographic methods, showing how the distortion is additive if several pulses are overlapped or partially overlapped. These results will be of interest to researchers involved in both laser welding and applications like laser forming or straightening, where an in-depth knowledge of distortion is important. 2. Methodology In this empirical paper the sum of all stresses resulting in an out-of-plane distortion has been measured with time transient, i.e. contributions from TGM, BM and UM are all included. The
out-of-plane real time transient deformation has been measured with an accuracy that is a fraction of the wavelength of the green holographic laser, 532 nm. The experimental results presented here show precisely how distortion takes place during the heating-cooling cycle and this is a useful tool for understanding how distortion is created by superimposed laser pulses and partially overlapping pulses. The experimental setup is shown in Fig. 2 and is similar to that used in earlier work [1]. The imaged object was a 100 x 250 x 2.4 mm stainless steel 304L plate. To reduce the influence of residual stresses from manufacturing and cutting, the plate was annealed at 1100 °C and then stress reduced at 300 °C for 6 h prior to the experiment. The holographic results was recorded by an unmodified Redlake Motionpro X3 CCD camera with a frame rate of 1000 Hz and an exposure time of 997 ls, to reduce the need for synchronization. The camera recorded phase shift on an area of 24 mm x 32 mm, with a resolution of 600 x 800 pixels. A ROFIN-SINAR RSM 200D/ SHG, 532 nm wavelength Q-Switched green laser with an estimated pulse length of 200 ns and peak power of 9 kW, was used as illumination source for the holography. On the opposite side from the measuring system a laser spot weld was produced with a welding laser, HAAS 3006D Nd:YAG laser with a 600 lm fibre diameter and optical configuration of 180:180 mm. A peak power of 1005 W and a 100 ms long pulse resulted in a small, shallow spot weld on the plate surface. Measurements were performed for square shaped pulses with a pulse energy of 100 Joules. Observations were made of the effect of; a single pulse, five superimposed pulses and five pulses which overlapped by 50% (each surface weld was 1.3 mm wide and the laser spot was moved 0.65 mm between pulses). Further details of the laser parameters are summarized in Table 1. The shielding gas was a coaxial flow of Argon. Table 2 shows the chemical composition of the SS 304L Alloy. The plate was fixed along its bottom edge and there was an optical boundary condition that the bottom pixel-line of the camera was assumed to have zero movement to reduce scatter from vibrations. For each 1 ms interval the interferometric phase difference was calculated from the digital hologram through Fourier transforms [26]. The phase information of the reference light is encoded on the CCD as high spatial frequency content. The original image was Fourier transformed and all low frequency pixels were set to zero. Fig. 3 shows a Fourier transform image. The left quarter of the image contains all the low-pass phase information. After the inverse Fourier transform each pixel has a complex value related to the phase difference between the object light and the reference light. By calculating the complex conjugate of two consecutive frames the phase change of the object light is identified and the displacement of the surface can be calculated. In Fig. 3 the windowed area displayed on the left contains the phase information in the lower phase. The phase differences represent the out of plane (Z) movement of the plate, and if this movement is larger than k/2 (266 nm) the phase will wrap. This is then compensated for in Matlab by an unwrapping algorithm. The total phase shift was accumulated over time and converted to physical
Fig. 1. Distortion from localized surface melting with TGM as the dominant mechanism.
3
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 2. The experimental set-up.
Table 1 Laser parameters. Processing parameter
Value/type
Holography parameter
Value/type
Laser type Optical configuration Pulse peak power Beam diameter (at surface) Pulse duration
HAAS 3006D Nd:YAG 600 lm fibre diameter 180:180 mm optics 1005 W 0.6 mm 100 ms
Laser type Wavelength Peak power Pulse duration
ROFIN SINAR RSM 200D/SNG 532 nm 9 kW 200 ns
Table 2 Material composition for the 304L stainless steel used in this study. Element Content wt%
C 0.03
Mn 2.00
P 0.045
S 0.030
Si 0.75
Cr 18–20
Ni 8–12
N 0.10
Fe Bal
Fig. 3. The interesting phase information from spectral filtering framed to the left in the Fast Fourier Transformed image.
out of plane deformation, in µm. This information can then be used to give, for example, an out of plane deformation map with a time resolution of 1 ms, like the one presented in Fig. 4.
3. Results and discussion The information presented in Fig. 5 is the plate deformation (central to the weld spot) for different time steps through the
heating and subsequent cooling process. In Fig. 6 the maximum deformation of a plate which has been surface melted by a single laser pulse is plotted as a function of time. Fig. 6 shows that the material initially responds to the heat from the laser by a slight deformation (30 µm) towards the welding laser source as a result of surface heating and local expansion, this corresponds to Fig. 1b. Once the illumination is interrupted the distortion reverses in direction as a combination of the effects described in Fig. 1c, now moving away from the laser source, to
4
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 4. Final out of plane deformation.
Fig. 5. Deformation information corresponding to different time steps produced by the holographic technique (Selected lines only), solid lines are when laser is on and dotted lines describe deformation during the cooling process.
The solid line in Fig. 7 shows the maximum out of plane distortion for several laser pulses superimposed at the same position on the sample surface. This line shows that a similar distortion cycle to that seen in Fig. 6 takes place for every laser pulse. The individual effects of the secondary pulses differ from that of the primary pulse in two ways: 1. The initial ‘negative’ distortion is larger than it was for the primary pulse. 2. The ‘positive’ maximum and final distortions for all subsequent pulses are smaller than for the primary pulse.
Fig. 6. Deformation as a function of time of the backside of the plate at the weld spot position.
reach a maximum displacement of approximately 350 lm 80 ms later. As the sample cools, it reaches a final deformation of approximately 300 µm. This transient out of plane distortion follows the same principle as that reported by other researchers [12–14,27,28].
The reason for these differences is that the primary pulse heats up annealed, stress reduced material but subsequent pulses are acting upon material which contains residual stress from previous pulses. Thus the subsequent pulses initially release some of the radially contracting stresses on the top surface which causes the stress equilibrium to change and, as a consequence, the negative deflection is larger. Subsequent solidification, cooling and shrinkage is acting on pre-stressed material which is more resistant to further distortion, so the maximum and final distortions are smaller than they were for the first pulse.
5
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 7. Additive distortion as a function of time from a sequence of fully overlapping superimposed pulses.
The maximum negative and eventual positive distortions from each individual pulse are presented in Table 3. As the stresses build with increasing numbers of pulses, both the initial negative and the final positive distortions decrease, but the inter-pulse reductions begin to plateau and subsequent pulses continue to contribute to the distortion process. It is interesting to note that, although the distortion created by superimposed pulses is additive, there is very little increase in the depth or volume of melt created by several pulses as compared to one, as long as the material is allowed to solidify and approach room temperature between pulses. Fig. 8 compares cross sections through the middle of the weld after one and five pulses. A single pulse melted the material to a depth of 0.46 mm and the addition of another four pulses increased this by only 11% to 0.51 mm. One should keep in mind that the surface conditions for subsequent pulses are different compared to the initial surface which had an as rolled surface and hence absorptivity is affected. From Figs. 7 and 8 one can see that repeated pulses can be a powerful distortion tool with minimal surface melting. In this case five pulses have resulted in a total deflection of over 600 microns with a melt depth equivalent to about 20% of the material thickness. A shallow melt depth is important to the deformation process because distortion will be optimised as a function of melt volume and the distance from the centre of gravity of the melt to the centre line of the sample, (the TGM principle is dominant). The dotted line in Fig. 7 presents the results of multiple pulses creating shallow weld pools which overlap by 50% of their width. The results given in Fig. 6 have been included in Fig. 8 to facilitate direct comparison. The maximum negative and eventual positive distortions from each individual (partially overlapping) pulse are presented in Table 4. It should be noted that for the 50% overlap experiment
Fig. 8. A comparison of the melt cross section created by a single pulse (left) and five totally overlapping pulses (right).
the position of maximum distortion is not stationary on the material surface as it is for the 100% overlap pulses i.e. the position of maximum distortion moves along the line of overlapping melt zones. A comparison of Tables 3 and 4 and a review of Fig. 7 reveal the difference between overlapping and partially overlapping pulses: For partially overlapping pulses the negative distortion of the plate is less and the positive distortion is greater for pulses 2–5 than they are in the overlapping case. This is clearly because the pulses after the first pulse are interacting with both pre-melted material and un pre-melted material. The result is therefore a
Table 3 Distortion resulting from five superimposed pulses. Pulse No.
1
2
3
4
5
Negative distortion [lm] Positive final distortion [lm] Cumulative distortion [lm]
42 310 310
167 90 400
155 80 480
155 69 549
131 66 615
6
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Table 4 Distortion resulting from five half-overlapping pulses. Pulse No.
1
2
3
4
5
Negative distortion [lm] Positive final distortion [lm] Cumulative distortion [lm]
42 310 310
95 170 480
95 170 648
120 155 805
109 120 925
mix between the conditions of the first pulse and a fully overlapping pulse. These results differ from those in the line scan work of Edwardson et al. [29], using strain gauges, in that in the present work the initial out of plane distortions are larger for the first pulse. This demonstrates a difference in the 3D stress state surrounding a pulse compared to the more 2D state surrounding a line scan. 4. Conclusions Using a pulsed holographic measurement technique the deformation experienced by a stainless steel plate surface melted by single and multiple pulses of laser light has been revealed in real time with an accuracy of the order of hundreds of nanometres. The results show an initial out of plane deflection towards the source of heating, which is followed by a deflection in the opposite direction during solidification and cooling. The eventual distortion level of a surface melted sheet has been shown to be substantially affected by its pre-existing stress state. If overlapping or partially overlapping pulses are used to repeatedly melt the material surface the material continues to deform with each pulse. This increase in deformation is not accompanied by substantial increases in melt depth if the surface welds are allowed to solidify and cool between pulses.
References [1] I. Eriksson, P. Haglund, J. Powell, M. Sjödahl, A.F.H. Kaplan, Holographic measurement of thermal distortion during laser spot welding, Opt. Eng. 51 (2012) 30501, https://doi.org/10.1117/1.OE.51.3.030501. [2] F. Vollertsen, I. Komel, R. Kals, The laser bending of steel foils for microparts by the buckling mechanism-a model, Model. Simul. Mater. Sci. Eng. 3 (1995) 107– 119, https://doi.org/10.1088/0965-0393/3/1/009. [3] V. Paunoiu, E.A. Squeo, F. Quadrini, C. Gheorghies, D. Nicoara, Laser bending of stainless steel sheet metals, Int. J. Mater. Form. 1 (2008) 1371–1374, https:// doi.org/10.1007/s12289-008-0119-8. [4] J. Zhou, H. Shen, X. Yu, J. Hu, Z. Yao, On the competition between in-plane and out-of-plane deformations in laser thermal adjustment, Opt. Laser Technol. 45 (2013) 689–696, https://doi.org/10.1016/j.optlastec.2012.05.010. [5] P. Carlone, G.S. Palazzo, R. Pasquino, Inverse analysis of the laser forming process by computational modelling and methods, Comput. Math. with Appl. 55 (2008) 2018–2032, https://doi.org/10.1016/j.camwa.2007.09.003. [6] A.A. Petrov, A.D. Kochetov, Pulsed laser-induced forming of thin sheet metal, in: V.P. Veiko, T.A. Vartanyan (Eds.), International Society for Optics and Photonics, 2010: p. 799609. doi:10.1117/12.889133. [7] H. Shen, F. Vollertsen, Modelling of laser forming – An review, Comput. Mater. Sci. 46 (2009) 834–840, https://doi.org/10.1016/j.commatsci.2009.04.022. [8] Y. Shi, P. Yi, Y. Liu, Numerical investigation of temperature field of different mechanisms in laser forming, Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 226 (2012) 2118–2125, https://doi.org/10.1177/0954406211430884.
[9] Z. Hu, R. Kovacevic, M. Labudovic, Experimental and numerical modeling of buckling instability of laser sheet forming, Int. J. Mach. Tools Manuf. 42 (2002) 1427–1439, https://doi.org/10.1016/S0890-6955(02)00075-5. [10] S.P. Edwardson, J. Griffiths, G. Dearden, K.G. Watkins, Temperature gradient mechanism: overview of the multiple pass controlling factors, Phys. Procedia. 5 (2010) 53–63, https://doi.org/10.1016/j.phpro.2010.08.122. [11] Y. Shi, Z. Yao, H. Shen, J. Hu, Research on the mechanisms of laser forming for the metal plate, Int. J. Mach. Tools. (2006). [12] M. Dovc, J. Mozina, F. Kosel, An analytical model for time-dependent deformation of a circular plate illuminated by a laser pulse, J. Phys. D. Appl. Phys. 31 (1998) 65–73, https://doi.org/10.1088/0022-3727/31/1/010. [13] M. Dovcˇ, J. Mozˇina, F. Kosel, Optodynamic deformation of a metal plate illuminated by a gaussian laser pulse, J. Therm. Stress. 24 (2001) 1043–1062, https://doi.org/10.1080/01495730152620050. [14] M. Dovc, J. Mozel, F. Kosel, Pulsed Laser Bending of a Plate as an Optodynamic Process Plates, in: M. Bamberg (Ed.), LANE Vol. 2, Erlangen, Germany, 1997: pp. 421–430. [15] Y. Shi, H. Shen, Z. Yao, J. Hu, Temperature gradient mechanism in laser forming of thin plates, Opt. Laser Technol. 39 (2007) 858–863, https://doi.org/10.1016/ j.optlastec.2005.12.006. [16] N. Pirch, K. Wissenbach, Mechanisms during laser bending, J. Laser Appl. 20 (2008) 135–139, https://doi.org/10.2351/1.2832404. [17] Y. Shi, H. Shen, Z. Yao, J. Hu, An analytical model based on the similarity in temperature distributions in laser forming, Opt. Lasers Eng. 45 (2007) 83–87, https://doi.org/10.1016/j.optlaseng.2006.04.006. [18] Y. Shi, Y. Liu, P. Yi, J. Hu, Effect of different heating methods on deformation of metal plate under upsetting mechanism in laser forming, Opt. Laser Technol. 44 (2012) 486–491, https://doi.org/10.1016/j.optlastec.2011.08.019. [19] K. Maji, D.K. Pratihar, A.K. Nath, Finite Element Analysis on Pulsed Laser Forming of Sheet Metal, in: 5th Int. 26th All India Manuf. Technol. Des. Res. Conf., IIT Guwahati, Assam. India, n.d. [20] K. Maji, R. Shukla, A.K. Nath, D.K. Pratihar, Finite element analysis and experimental investigations on laser bending of AISI304 stainless steel sheet, Procedia Eng. 64 (2013) 528–535, https://doi.org/10.1016/j. proeng.2013.09.127. [21] J. Hu, H. Xu, D. Dang, Modeling and reducing edge effects in laser bending, J. Mater. Process. Technol. 213 (2013) 1989–1996, https://doi.org/10.1016/j. jmatprotec.2013.05.007. [22] G. Chen, X. Xu, Experimental and 3D finite element studies of CW Laser forming of thin stainless steel sheets, J. Manuf. Sci. Eng. 123 (2001) 66, https:// doi.org/10.1115/1.1347036. [23] K. Kheloufi, E.H. Amara, E.-H. Amara, S. Boudjemai, D. Doumaz, Numerical Simulation Of Steel Plate Bending Process Using Stationary Laser Beam By A Coupled Finite Element Thermal Mechanical Analysis, in: AIP Conf. Proc., AIP, 2008: pp. 176–179. doi:10.1063/1.2999929. [24] S. Torabnia, A. Banazadeh, An analytical study on laser forming process of sheet metals, using new elasto-plastic temperature dependent material model, Adv. Mater. Res. 622–623 (2012) 569–574, https://doi.org/10.4028/ www.scientific.net/AMR.622-623.569. [25] J.H. Song, E.Z. Kim, S.J. Park, G.A. Lee, Numerical and experimental investigation on laser irradiated bending process of advanced high strength steel, Mater. Res. Innov. 17 (2013) s46–s52, https://doi.org/10.1179/ 1432891713Z.000000000303. [26] H.-S. Hsieh, J. Lin, Thermal–mechanical analysis on the transient deformation during pulsed laser forming, Int. J. Mach. Tools Manuf. 44 (2004) 191–199, https://doi.org/10.1016/j.ijmachtools.2003.10.003. [27] U. Schnars, J. Jueptner, Digital Holography, Springer Verlag, Berlin, Heidelberg, New York, 2005. [28] K. Lee, J. Lin, Transient deformation of thin metal sheets during pulsed laser forming, Opt, Laser Technol, 2002. [29] S.P. Edwardson, K.G. Watkins, G. Dearden, P. French, J. Magee, Strain gauge analysis of laser forming, J. Laser Appl. 15 (2003) 225–232, https://doi.org/ 10.2351/1.1620000.
Contents lists available at ScienceDirect
Optics and Laser Technology journal homepage: www.elsevier.com/locate/optlastec
Full length article
Holographic measurement of distortion during laser melting: Additive distortion from overlapping pulses Peter Haglund a,b,⇑, Jan Frostevarg a, John Powell a, Ingemar Eriksson c, Alexander F.H. Kaplan a a
Department of Engineering Sciences and Mathematics, Luleå University of Technology, Sweden Westinghouse Electric Sweden AB, Västerås, Sweden c Dalco Elteknik AB, Östersund, Sweden b
a r t i c l e
i n f o
Article history: Received 1 March 2017 Received in revised form 30 June 2017 Accepted 26 September 2017
Keywords: Pulsed laser welding Laser surface melting Thermal distortion Laser forming Laser straightening Holography
a b s t r a c t Laser - material interactions such as welding, heat treatment and thermal bending generate thermal gradients which give rise to thermal stresses and strains which often result in a permanent distortion of the heated object. This paper investigates the thermal distortion response which results from pulsed laser surface melting of a stainless steel sheet. Pulsed holography has been used to accurately monitor, in real time, the out-of-plane distortion of stainless steel samples melted on one face by with both single and multiple laser pulses. It has been shown that surface melting by additional laser pulses increases the out of plane distortion of the sample without significantly increasing the melt depth. The distortion differences between the primary pulse and subsequent pulses has also been analysed for fully and partially overlapping laser pulses. Ó 2017 Elsevier Ltd. All rights reserved.
1. Introduction When metal products are exposed to a temperature increase and melted locally, residual stresses and distortion are inevitable side effects of the heating and subsequent cooling process [1]. Most laser - material interactions involve some kind of permanent plastic thermal distortion, since the temperature is usually so high that the thermal stresses override the material tensile strength of the work-piece. This is an important point to consider when developing processes and designing components with respect to e.g. weld design, so an understanding of distortion is important. Thermal distortion is a complex physical process and the situation is even further complicated in the case where the laser interacts with material which is pre-stressed from mechanical processes such as cold working or thermal stresses, which is the case for most industrial applications. In cases where the distortion is unwanted, such as welding, the laser has established itself as a low heat input, low distortion alternative to traditional torch technologies but, nevertheless, some distortion will generally be apparent. On the other hand, in applications such as laser bending [2] and wire straightening [3] a distortion effect is the desired outcome. Most of the currently published research considers only continuous ⇑ Corresponding author at: Department of Engineering Sciences and Mathematics, Luleå University of Technology, Sweden. E-mail address: [email protected] (P. Haglund). https://doi.org/10.1016/j.optlastec.2017.09.053 0030-3992/Ó 2017 Elsevier Ltd. All rights reserved.
wave laser processing. In this paper the deformation of a sheet as a result of melting by individual laser pulses is studied empirically. 1.1. Distortion mechanisms Work by Zhou [4] divides distortion into in-plane shrinkage and out-of-plane bending of the component. Several researchers have used an approach where deformation is divided into three mechanisms, [5–9]: 1. Temperature gradient mechanism (TGM). 2. Buckling mechanism (BM). 3. Upsetting mechanism (UM). The buckling and upsetting mechanisms are important when the laser creates a local heating pattern where the temperature is fairly uniform through the thickness of the sheet. In both mechanisms the heating and cooling cycle experienced by the sheet causes plastic failure and eventual deflection of the material. In the case of buckling, the final out-of-plane deformation direction is generally determined by the previous residual stress condition of the material (rolling direction, etc.). In the upsetting process the material is thickened in the laser heated area (and often shortened in other directions). In the case of this experiment the temperature gradient distortion mechanism is dominant. As the name suggests this
2
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
mechanism depends upon a sharp temperature gradient from the surface to the underlying material. When the top surface of a metal plate is locally heated by a laser the material undergoes a cycle of distortion as follows: 1. Heating begins: the top surface of the plate expands locally causing the material to distort. The heated surface moves towards the source of heating (see Fig. 1b). 2. The thermal stresses exceed the yield stress of the material which is then plastically compressed in the heated zone. On cooling, this compressed zone causes the sheet to distort. During cooling the heated spot moves away from the source of heating (see Fig. 1c) and remains in this new position. This laser-material interaction can be complicated by the melting and solidification of the centre of the laser heated zone, but the basic principle remains the same. 1.2. Multiple consecutive distortion Edwardson et al. [10] has studied the cumulative distortion effects of multiple line laser melting and found that the bend angle increases for up to 60 scans. In work by Shi et al. [11] the elasticplastic deformation is further described and a new mechanism is proposed; the coupling mechanism, which is a combination of the TGM and the UM. In the work of Eriksson et al. [1] holography was used to monitor thermal distortion in real time to give information about the history of the deformation. In work by Dovc [12–14], an analytical model was developed and studied for a single pulse, this model confirms the results in [1]. Shi et al. [15] concluded that the deformation for a CW process is not purely 2D for the TGM case and state that the distortions are affected by the elastic properties of the plate. Pirch and Wisenbach [16] found that the TGM model does not explain the fact that the bending angle decreases with increasing amounts of irradiation. Several researchers e.g. [17–25] have used Finite Element simulations for modelling distortion resulting from laser irradiation processes. Accurate measurements are an important part of the validation process for such simulations. An important aspect is the influence of heat input and heating methods, which has been studied by Shi et al. [18]. In this work the penetration profile in combination with sheet thickness and hence the thermal field were chosen so that the out of plane deformation would be as large as possible with pulsed laser heating. In the current paper the transient time response of a steel blank sheet exposed to multiple laser pulses is analysed by holographic methods, showing how the distortion is additive if several pulses are overlapped or partially overlapped. These results will be of interest to researchers involved in both laser welding and applications like laser forming or straightening, where an in-depth knowledge of distortion is important. 2. Methodology In this empirical paper the sum of all stresses resulting in an out-of-plane distortion has been measured with time transient, i.e. contributions from TGM, BM and UM are all included. The
out-of-plane real time transient deformation has been measured with an accuracy that is a fraction of the wavelength of the green holographic laser, 532 nm. The experimental results presented here show precisely how distortion takes place during the heating-cooling cycle and this is a useful tool for understanding how distortion is created by superimposed laser pulses and partially overlapping pulses. The experimental setup is shown in Fig. 2 and is similar to that used in earlier work [1]. The imaged object was a 100 x 250 x 2.4 mm stainless steel 304L plate. To reduce the influence of residual stresses from manufacturing and cutting, the plate was annealed at 1100 °C and then stress reduced at 300 °C for 6 h prior to the experiment. The holographic results was recorded by an unmodified Redlake Motionpro X3 CCD camera with a frame rate of 1000 Hz and an exposure time of 997 ls, to reduce the need for synchronization. The camera recorded phase shift on an area of 24 mm x 32 mm, with a resolution of 600 x 800 pixels. A ROFIN-SINAR RSM 200D/ SHG, 532 nm wavelength Q-Switched green laser with an estimated pulse length of 200 ns and peak power of 9 kW, was used as illumination source for the holography. On the opposite side from the measuring system a laser spot weld was produced with a welding laser, HAAS 3006D Nd:YAG laser with a 600 lm fibre diameter and optical configuration of 180:180 mm. A peak power of 1005 W and a 100 ms long pulse resulted in a small, shallow spot weld on the plate surface. Measurements were performed for square shaped pulses with a pulse energy of 100 Joules. Observations were made of the effect of; a single pulse, five superimposed pulses and five pulses which overlapped by 50% (each surface weld was 1.3 mm wide and the laser spot was moved 0.65 mm between pulses). Further details of the laser parameters are summarized in Table 1. The shielding gas was a coaxial flow of Argon. Table 2 shows the chemical composition of the SS 304L Alloy. The plate was fixed along its bottom edge and there was an optical boundary condition that the bottom pixel-line of the camera was assumed to have zero movement to reduce scatter from vibrations. For each 1 ms interval the interferometric phase difference was calculated from the digital hologram through Fourier transforms [26]. The phase information of the reference light is encoded on the CCD as high spatial frequency content. The original image was Fourier transformed and all low frequency pixels were set to zero. Fig. 3 shows a Fourier transform image. The left quarter of the image contains all the low-pass phase information. After the inverse Fourier transform each pixel has a complex value related to the phase difference between the object light and the reference light. By calculating the complex conjugate of two consecutive frames the phase change of the object light is identified and the displacement of the surface can be calculated. In Fig. 3 the windowed area displayed on the left contains the phase information in the lower phase. The phase differences represent the out of plane (Z) movement of the plate, and if this movement is larger than k/2 (266 nm) the phase will wrap. This is then compensated for in Matlab by an unwrapping algorithm. The total phase shift was accumulated over time and converted to physical
Fig. 1. Distortion from localized surface melting with TGM as the dominant mechanism.
3
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 2. The experimental set-up.
Table 1 Laser parameters. Processing parameter
Value/type
Holography parameter
Value/type
Laser type Optical configuration Pulse peak power Beam diameter (at surface) Pulse duration
HAAS 3006D Nd:YAG 600 lm fibre diameter 180:180 mm optics 1005 W 0.6 mm 100 ms
Laser type Wavelength Peak power Pulse duration
ROFIN SINAR RSM 200D/SNG 532 nm 9 kW 200 ns
Table 2 Material composition for the 304L stainless steel used in this study. Element Content wt%
C 0.03
Mn 2.00
P 0.045
S 0.030
Si 0.75
Cr 18–20
Ni 8–12
N 0.10
Fe Bal
Fig. 3. The interesting phase information from spectral filtering framed to the left in the Fast Fourier Transformed image.
out of plane deformation, in µm. This information can then be used to give, for example, an out of plane deformation map with a time resolution of 1 ms, like the one presented in Fig. 4.
3. Results and discussion The information presented in Fig. 5 is the plate deformation (central to the weld spot) for different time steps through the
heating and subsequent cooling process. In Fig. 6 the maximum deformation of a plate which has been surface melted by a single laser pulse is plotted as a function of time. Fig. 6 shows that the material initially responds to the heat from the laser by a slight deformation (30 µm) towards the welding laser source as a result of surface heating and local expansion, this corresponds to Fig. 1b. Once the illumination is interrupted the distortion reverses in direction as a combination of the effects described in Fig. 1c, now moving away from the laser source, to
4
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 4. Final out of plane deformation.
Fig. 5. Deformation information corresponding to different time steps produced by the holographic technique (Selected lines only), solid lines are when laser is on and dotted lines describe deformation during the cooling process.
The solid line in Fig. 7 shows the maximum out of plane distortion for several laser pulses superimposed at the same position on the sample surface. This line shows that a similar distortion cycle to that seen in Fig. 6 takes place for every laser pulse. The individual effects of the secondary pulses differ from that of the primary pulse in two ways: 1. The initial ‘negative’ distortion is larger than it was for the primary pulse. 2. The ‘positive’ maximum and final distortions for all subsequent pulses are smaller than for the primary pulse.
Fig. 6. Deformation as a function of time of the backside of the plate at the weld spot position.
reach a maximum displacement of approximately 350 lm 80 ms later. As the sample cools, it reaches a final deformation of approximately 300 µm. This transient out of plane distortion follows the same principle as that reported by other researchers [12–14,27,28].
The reason for these differences is that the primary pulse heats up annealed, stress reduced material but subsequent pulses are acting upon material which contains residual stress from previous pulses. Thus the subsequent pulses initially release some of the radially contracting stresses on the top surface which causes the stress equilibrium to change and, as a consequence, the negative deflection is larger. Subsequent solidification, cooling and shrinkage is acting on pre-stressed material which is more resistant to further distortion, so the maximum and final distortions are smaller than they were for the first pulse.
5
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Fig. 7. Additive distortion as a function of time from a sequence of fully overlapping superimposed pulses.
The maximum negative and eventual positive distortions from each individual pulse are presented in Table 3. As the stresses build with increasing numbers of pulses, both the initial negative and the final positive distortions decrease, but the inter-pulse reductions begin to plateau and subsequent pulses continue to contribute to the distortion process. It is interesting to note that, although the distortion created by superimposed pulses is additive, there is very little increase in the depth or volume of melt created by several pulses as compared to one, as long as the material is allowed to solidify and approach room temperature between pulses. Fig. 8 compares cross sections through the middle of the weld after one and five pulses. A single pulse melted the material to a depth of 0.46 mm and the addition of another four pulses increased this by only 11% to 0.51 mm. One should keep in mind that the surface conditions for subsequent pulses are different compared to the initial surface which had an as rolled surface and hence absorptivity is affected. From Figs. 7 and 8 one can see that repeated pulses can be a powerful distortion tool with minimal surface melting. In this case five pulses have resulted in a total deflection of over 600 microns with a melt depth equivalent to about 20% of the material thickness. A shallow melt depth is important to the deformation process because distortion will be optimised as a function of melt volume and the distance from the centre of gravity of the melt to the centre line of the sample, (the TGM principle is dominant). The dotted line in Fig. 7 presents the results of multiple pulses creating shallow weld pools which overlap by 50% of their width. The results given in Fig. 6 have been included in Fig. 8 to facilitate direct comparison. The maximum negative and eventual positive distortions from each individual (partially overlapping) pulse are presented in Table 4. It should be noted that for the 50% overlap experiment
Fig. 8. A comparison of the melt cross section created by a single pulse (left) and five totally overlapping pulses (right).
the position of maximum distortion is not stationary on the material surface as it is for the 100% overlap pulses i.e. the position of maximum distortion moves along the line of overlapping melt zones. A comparison of Tables 3 and 4 and a review of Fig. 7 reveal the difference between overlapping and partially overlapping pulses: For partially overlapping pulses the negative distortion of the plate is less and the positive distortion is greater for pulses 2–5 than they are in the overlapping case. This is clearly because the pulses after the first pulse are interacting with both pre-melted material and un pre-melted material. The result is therefore a
Table 3 Distortion resulting from five superimposed pulses. Pulse No.
1
2
3
4
5
Negative distortion [lm] Positive final distortion [lm] Cumulative distortion [lm]
42 310 310
167 90 400
155 80 480
155 69 549
131 66 615
6
P. Haglund et al. / Optics and Laser Technology 100 (2018) 1–6
Table 4 Distortion resulting from five half-overlapping pulses. Pulse No.
1
2
3
4
5
Negative distortion [lm] Positive final distortion [lm] Cumulative distortion [lm]
42 310 310
95 170 480
95 170 648
120 155 805
109 120 925
mix between the conditions of the first pulse and a fully overlapping pulse. These results differ from those in the line scan work of Edwardson et al. [29], using strain gauges, in that in the present work the initial out of plane distortions are larger for the first pulse. This demonstrates a difference in the 3D stress state surrounding a pulse compared to the more 2D state surrounding a line scan. 4. Conclusions Using a pulsed holographic measurement technique the deformation experienced by a stainless steel plate surface melted by single and multiple pulses of laser light has been revealed in real time with an accuracy of the order of hundreds of nanometres. The results show an initial out of plane deflection towards the source of heating, which is followed by a deflection in the opposite direction during solidification and cooling. The eventual distortion level of a surface melted sheet has been shown to be substantially affected by its pre-existing stress state. If overlapping or partially overlapping pulses are used to repeatedly melt the material surface the material continues to deform with each pulse. This increase in deformation is not accompanied by substantial increases in melt depth if the surface welds are allowed to solidify and cool between pulses.
References [1] I. Eriksson, P. Haglund, J. Powell, M. Sjödahl, A.F.H. Kaplan, Holographic measurement of thermal distortion during laser spot welding, Opt. Eng. 51 (2012) 30501, https://doi.org/10.1117/1.OE.51.3.030501. [2] F. Vollertsen, I. Komel, R. Kals, The laser bending of steel foils for microparts by the buckling mechanism-a model, Model. Simul. Mater. Sci. Eng. 3 (1995) 107– 119, https://doi.org/10.1088/0965-0393/3/1/009. [3] V. Paunoiu, E.A. Squeo, F. Quadrini, C. Gheorghies, D. Nicoara, Laser bending of stainless steel sheet metals, Int. J. Mater. Form. 1 (2008) 1371–1374, https:// doi.org/10.1007/s12289-008-0119-8. [4] J. Zhou, H. Shen, X. Yu, J. Hu, Z. Yao, On the competition between in-plane and out-of-plane deformations in laser thermal adjustment, Opt. Laser Technol. 45 (2013) 689–696, https://doi.org/10.1016/j.optlastec.2012.05.010. [5] P. Carlone, G.S. Palazzo, R. Pasquino, Inverse analysis of the laser forming process by computational modelling and methods, Comput. Math. with Appl. 55 (2008) 2018–2032, https://doi.org/10.1016/j.camwa.2007.09.003. [6] A.A. Petrov, A.D. Kochetov, Pulsed laser-induced forming of thin sheet metal, in: V.P. Veiko, T.A. Vartanyan (Eds.), International Society for Optics and Photonics, 2010: p. 799609. doi:10.1117/12.889133. [7] H. Shen, F. Vollertsen, Modelling of laser forming – An review, Comput. Mater. Sci. 46 (2009) 834–840, https://doi.org/10.1016/j.commatsci.2009.04.022. [8] Y. Shi, P. Yi, Y. Liu, Numerical investigation of temperature field of different mechanisms in laser forming, Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 226 (2012) 2118–2125, https://doi.org/10.1177/0954406211430884.
[9] Z. Hu, R. Kovacevic, M. Labudovic, Experimental and numerical modeling of buckling instability of laser sheet forming, Int. J. Mach. Tools Manuf. 42 (2002) 1427–1439, https://doi.org/10.1016/S0890-6955(02)00075-5. [10] S.P. Edwardson, J. Griffiths, G. Dearden, K.G. Watkins, Temperature gradient mechanism: overview of the multiple pass controlling factors, Phys. Procedia. 5 (2010) 53–63, https://doi.org/10.1016/j.phpro.2010.08.122. [11] Y. Shi, Z. Yao, H. Shen, J. Hu, Research on the mechanisms of laser forming for the metal plate, Int. J. Mach. Tools. (2006). [12] M. Dovc, J. Mozina, F. Kosel, An analytical model for time-dependent deformation of a circular plate illuminated by a laser pulse, J. Phys. D. Appl. Phys. 31 (1998) 65–73, https://doi.org/10.1088/0022-3727/31/1/010. [13] M. Dovcˇ, J. Mozˇina, F. Kosel, Optodynamic deformation of a metal plate illuminated by a gaussian laser pulse, J. Therm. Stress. 24 (2001) 1043–1062, https://doi.org/10.1080/01495730152620050. [14] M. Dovc, J. Mozel, F. Kosel, Pulsed Laser Bending of a Plate as an Optodynamic Process Plates, in: M. Bamberg (Ed.), LANE Vol. 2, Erlangen, Germany, 1997: pp. 421–430. [15] Y. Shi, H. Shen, Z. Yao, J. Hu, Temperature gradient mechanism in laser forming of thin plates, Opt. Laser Technol. 39 (2007) 858–863, https://doi.org/10.1016/ j.optlastec.2005.12.006. [16] N. Pirch, K. Wissenbach, Mechanisms during laser bending, J. Laser Appl. 20 (2008) 135–139, https://doi.org/10.2351/1.2832404. [17] Y. Shi, H. Shen, Z. Yao, J. Hu, An analytical model based on the similarity in temperature distributions in laser forming, Opt. Lasers Eng. 45 (2007) 83–87, https://doi.org/10.1016/j.optlaseng.2006.04.006. [18] Y. Shi, Y. Liu, P. Yi, J. Hu, Effect of different heating methods on deformation of metal plate under upsetting mechanism in laser forming, Opt. Laser Technol. 44 (2012) 486–491, https://doi.org/10.1016/j.optlastec.2011.08.019. [19] K. Maji, D.K. Pratihar, A.K. Nath, Finite Element Analysis on Pulsed Laser Forming of Sheet Metal, in: 5th Int. 26th All India Manuf. Technol. Des. Res. Conf., IIT Guwahati, Assam. India, n.d. [20] K. Maji, R. Shukla, A.K. Nath, D.K. Pratihar, Finite element analysis and experimental investigations on laser bending of AISI304 stainless steel sheet, Procedia Eng. 64 (2013) 528–535, https://doi.org/10.1016/j. proeng.2013.09.127. [21] J. Hu, H. Xu, D. Dang, Modeling and reducing edge effects in laser bending, J. Mater. Process. Technol. 213 (2013) 1989–1996, https://doi.org/10.1016/j. jmatprotec.2013.05.007. [22] G. Chen, X. Xu, Experimental and 3D finite element studies of CW Laser forming of thin stainless steel sheets, J. Manuf. Sci. Eng. 123 (2001) 66, https:// doi.org/10.1115/1.1347036. [23] K. Kheloufi, E.H. Amara, E.-H. Amara, S. Boudjemai, D. Doumaz, Numerical Simulation Of Steel Plate Bending Process Using Stationary Laser Beam By A Coupled Finite Element Thermal Mechanical Analysis, in: AIP Conf. Proc., AIP, 2008: pp. 176–179. doi:10.1063/1.2999929. [24] S. Torabnia, A. Banazadeh, An analytical study on laser forming process of sheet metals, using new elasto-plastic temperature dependent material model, Adv. Mater. Res. 622–623 (2012) 569–574, https://doi.org/10.4028/ www.scientific.net/AMR.622-623.569. [25] J.H. Song, E.Z. Kim, S.J. Park, G.A. Lee, Numerical and experimental investigation on laser irradiated bending process of advanced high strength steel, Mater. Res. Innov. 17 (2013) s46–s52, https://doi.org/10.1179/ 1432891713Z.000000000303. [26] H.-S. Hsieh, J. Lin, Thermal–mechanical analysis on the transient deformation during pulsed laser forming, Int. J. Mach. Tools Manuf. 44 (2004) 191–199, https://doi.org/10.1016/j.ijmachtools.2003.10.003. [27] U. Schnars, J. Jueptner, Digital Holography, Springer Verlag, Berlin, Heidelberg, New York, 2005. [28] K. Lee, J. Lin, Transient deformation of thin metal sheets during pulsed laser forming, Opt, Laser Technol, 2002. [29] S.P. Edwardson, K.G. Watkins, G. Dearden, P. French, J. Magee, Strain gauge analysis of laser forming, J. Laser Appl. 15 (2003) 225–232, https://doi.org/ 10.2351/1.1620000.