Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum

G Model

CIRPJ-290; No. of Pages 10 CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

Contents lists available at ScienceDirect

CIRP Journal of Manufacturing Science and Technology journal homepage: www.elsevier.com/locate/cirpj

Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum Amber Shrivastava a, Manuela Krones b, Frank E. Pfefferkorn a,* a b

Department of Mechanical Engineering, University of Wisconsin-Madison, Madison, WI, USA Professorship of Factory Planning and Factory Management, Technische Universita¨t Chemnitz, Chemnitz, Germany

A R T I C L E I N F O

A B S T R A C T

Article history: Available online xxx

One of the advantages of friction stir welding (FSW) is reduced energy consumption as compared to arc welding processes. This advantage has been predicted and qualitatively established. However, a quantitative analysis based on energy measurements during the processes and how to equitably compare them is missing. The objective of this work is to quantitatively compare the energy consumption associated with the creation of full-penetration welds in aluminum 6061-T6 workpieces by FSW and gas metal arc welding (GMAW) processes. The workpiece thicknesses for the two processes (5-mm-thick for FSW and 7.1-mm-thick for GMAW) are chosen such that the maximum tensile force sustained by the joints during tensile testing is similar. This accounts for material saving due to the higher ultimate tensile strength resulting from FSW. The energy consumed for any pre-processes, the welding processes, and post-processes was measured. Finally, a life cycle assessment (LCA) approach was used to determine and compare the environmental impact of FSW and GMAW. For the welding parameters used in this study joining by FSW consumes 42% less energy as compared to GMAW and utilizes approximately 10% less material for the design criteria of similar maximum tensile force. This leads to approximately 31% less greenhouse gas emissions for FSW as compared to GMAW. Both, the lower energy consumption during FSW, and involved pre and post processes contributed in the overall energy reduction. ß 2014 CIRP.

Keywords: Energy consumption Environmental impact Life cycle assessment Friction stir welding Gas metal arc welding

Introduction Friction stir welding Friction stir welding (FSW) was invented by Wayne Thomas at The Welding Institute (TWI) in 1991 [1]. FSW is a metal joining process in which two or more components are plastically deformed and mechanically intermixed under mechanical pressure at elevated temperatures [2,3]. These joints are created below the solidus temperature of the workpiece material, which makes FSW a solid-state welding process. Fig. 1 shows a schematic of the FSW process for a butt weld. The process involves a non-consumable rotating FSW tool, with specifically designed probe (pin) and shoulder. The FSW tool is plunged with a downward force into the workpiece. Once the probe is completely inserted in the workpiece and the shoulder makes contact with its surface, the tool is traversed along the weld seam (butt welding) or defined path

* Corresponding author. Tel.: +1 608 263 2668. E-mail address: [email protected] (F.E. Pfefferkorn).

(lap welding, bead-on-plate, friction stir processing). The tool is retracted at the end of the weld. Initially, heat is generated due to friction between the tool and workpiece, which facilitates plastic deformation of the parent material (i.e., stirring). Once, the material is being plastically deformed in the stir zone, heat is generated by friction and heat dissipation due to plastic deformation. The plasticized material is mixed and extruded past the tool and finally, it is forged together in the wake of the tool. FSW as a metal joining process is gaining acceptance in industrial application as the joint qualities and the cost benefits are better understood. Most friction stir welds are currently made in aluminum and magnesium alloys; however, the application of FSW to dissimilar materials and higher melting temperature alloys (e.g., ferrous alloys) is increasing. Gas metal arc welding (GMAW) Gas metal arc welding (GMAW) was developed in the 1950s. It was formerly known as metal inert gas (MIG) welding [4]. It is a fusion welding process in which the workpieces melt and resolidify to make the joint. In GMAW, the heat required for melting

http://dx.doi.org/10.1016/j.cirpj.2014.10.001 1755-5817/ß 2014 CIRP.

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 2

A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

Fig. 1. Schematic of friction stir [butt] welding (FSW).

the workpieces is obtained from electrical energy. During welding, a consumable wire electrode is utilized which establishes the arc and melts in the process to feed additional material to the melt pool. The consumable wire electrode is continuously fed through a nozzle. The weld-area/melt-pool is shielded by an effectively inert atmosphere of argon, helium, carbon dioxide or various other gas mixtures. Fig. 2 shows a schematic of the GMAW process. GMAW is extensively used in the metal fabrication industry [4] and is suitable for welding both ferrous and non-ferrous metals. Energy consumption and process emissions It is believed in the welding community (research and industry) that less energy is consumed during FSW as compared to any fusion welding method. This is due to lower welding temperatures achieved during FSW and the solid-state nature of the process, i.e., no melting of the workpiece material. Lakshminarayanan et al. [5] estimated the heat inputs for GMAW and gas tungsten arc welding (GTAW) and compared them to the heat input for FSW. 6 mm thick aluminum 6061 plates were butt welded by these processes. The heat input was estimated only for the processes, and pre and postprocesses were not accounted in these calculations. The heat input for FSW was estimated according to Heurtier et al. [6]. The heat inputs for GMAW and GTAW were found to be 2 times and 1.5 times the heat input for FSW, respectively. Lakshminarayanan et

Fig. 2. Schematic of gas metal arc welding (GMAW) process [4].

al. [5] also found that the tensile strength of the FSW joints was 34% and 15% greater than the GMAW and GTAW joints, respectively. Prasad and Prasanna [7] studied the hardness and microstructure in the welded material for FSW and GMAW joints. It was revealed that the heat affected zone (HAZ) in FSW welds was narrower than in their GMAW welds: a result of the different heat input. There are aspects of FSW that in addition to lower welding temperatures could result in lower resource utilization, energy consumption, emissions, health hazards and environmental effects as compared with fusion welding processes (e.g., GMAW, GTAW, SMAW, etc). Balasubramanian [8] stated that more than 10,000,000 workers worldwide are employed full time as welders and a higher number of workers perform welding intermittently as part of their job. The common health disorders in full time workers due to the welding emissions include: irritation of the eyes, nose and throat, pulmonary edema, and Parkinson’s disease. Health hazards due to welding processes are mainly caused by particulate emissions in the breathing zone of the welder. Depending on the size of the particulates, their influence on the welder’s body may change. Therefore, particulates are described in categories according to their maximal size in mm. Pfefferkorn et al. [10] found that FSW leads to average emissions of PM 2.5 particulates of 0.018–0.029 mg/m3 for Al 6061-T6 and 0.015–0.022 mg/m3 for Al 5083-H111. Cole et al. [11] analyzed the rate of PM 5 particulates for GMAW of Al 6061 in the welder’s breathing zone and found an average of 12 mg/m3 for welding with Al 4043 wire and 14.1 mg/m3 for Al 5356 wire. Matczak and Gromiec [12] analyzed PM 0.8 particulate emissions while welding Al 5083 in industrial welding shops. Based on their results, the average emissions over an average 8-h-shift are 1 mg/m3 with a maximum of 3.6 mg/m3. These results suggest that particulate emissions from FSW of aluminum are orders of magnitude smaller than GMAW, which will result in significantly lower air handling and filtration requirements. Dawood et al. [9] measured the mechanical properties and gaseous emissions from FSW and GMAW of 3-mm-thick 1030 aluminum. The carbon monoxide and carbon dioxide emissions for GMAW were approximately 3.7 and 1.6 times greater than, respectively, the corresponding emissions during FSW. It was concluded that FSW is relatively green, environmentally-friendly and results in superior welding properties compared to GMAW, when welding the same thickness aluminum material. FSW of aluminum alloys does not require shielding gases or flux, and does not use filler material. There are no pre-processing operations required for FSW. Chamfering/edge-preparation of workpieces is not required for FSW, even for 50-mm-thick welds. Cleaning of edges is not required to create the joint. Friction stir welding has few, or no, post-processing requirements because of the lower temperatures experienced and lack of filler material. The only common post-process is associated with eliminating the exit hole created when the FSW tool is retracted. The lower weld zone temperatures result in little or no thermal distortion of the structure, therefore, little or no straightening is required. The lack of filler material results in a smooth weld surface that does not require grinding or machining. The fine microstructure produced in a friction stir weld and the lower amount of annealing/aging that occurs during the process results in mechanical properties that are often better than comparable fusion welds. This can reduce the need for post-welding heat treatment. The energy consumption associated with the common pre-processing, welding process, and post-processing steps of FSW and GMAW is qualitatively shown in Fig. 3. In industry, FSW is predominantly used for welding aluminum and magnesium alloys. The metallurgical developments

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

3

Energy assessment

Fig. 3. Qualitative energy flowchart of aluminum friction stir welding (FSW) and gas metal arc welding (GMAW) processes.

of FSW tool alloys have resulted in FSW of ferrous alloys. Lienert et al. [13] successfully friction stir welded hot-rolled AISI 1018 mild steel using molybdenum-based and tungsten based alloy tools. These welds failed in base metal regions during tensile testing, and demonstrated yield and ultimate tensile strengths comparable to those of the base metal. Defalco and Steel [14] discussed the advantages in productivity and cost savings for the pipeline industry that can come from friction stir welding of steel pipes. Fusion welding of dissimilar metals such as steel and aluminum has been regarded as nearly impossible for commercial purposes. Honda Motor Co., Ltd. recently announced its newly developed robotic technology for the continuous welding of steel and aluminum based on the FSW process and applied it to its flagship product [15]. It reduced the electricity consumption during the welding process by approximately 50% compared to GMAW. Mononen et al. [16] carried out a cost comparison of FSW and GMAW welded aluminum panels. The cost comparison was based on the production time, machine investment, patent licensing, consumables and tooling costs. It was discovered that GMAW welding costs were dominated by labor wages and machine investment and by a lesser degree by filler material costs. FSW costs were dominated by machine investment, patent licensing and labor wages. FSW was more economical than GMAW when the annual production volume was large enough (on the order of tens of km of weld per year). The choice of FSW instead of GMAW before the intersection point was justified by the FSW process advantages such as low distortion, high strength, low amount of welding defects and improved occupational health issues. Defalco [17] estimated that the cost per unit length for GMAW was approximately 1.6 times that of the FSW process. It was acknowledged that capital cost for FSW was higher, however, the cost per unit length was less due to fast welding speeds and low preparation costs.

Energy-related key performance indicators (KPI) are widely used tools to describe the energy consumption of manufacturing processes. They can be focused on energy consumption, environmental impact or financial figures [19]. Usually, normalized indicators are used, i.e., a ratio between an energy figure and a process-related figure [20]. Another indicator to assess and compare manufacturing processes is the process efficiency [21]. It describes the relation between the minimum energy required for the process and the total energy input. However, detailed information on temperatures and forces of a process needs to be available to calculate the minimum required energy. For the comparison of welding processes, a common indicator is the energy input divided by the weld length [22]. The energy input per volume of molten material or its reciprocal is also used [23]. The energy consumption can be determined either with publicly available data or with detailed power and time measurements [24]. However, FSW and GMAW do not only differ with respect to their energy consumption. The use of different resources, such as the material for the electrode, needs to be considered. Hence, a more holistic consideration in terms of the environmental effects is required. An overview on environmental assessment methods is given in [25]. The most commonly used method is the life cycle assessment (LCA) as described in ISO 14040 [18]. It evaluates the environmental effects of a product over its entire life cycle. A limitation to specific stages of a product’s life cycle is possible. As such, LCA analyses have been applied to manufacturing processes in general [26–28], but not yet to friction stir welding or gas-metal arc welding processes. Objective The objective of this work is to quantitatively compare the energy consumption and environmental impact associated with full-penetration butt-welding of aluminum 6061-T6 workpieces by FSW and GMAW processes. The energy consumed for any preprocesses and post-processes involved with the joining methods are also taken into account. The life cycle assessment (LCA) approach is used to determine and compare the environmental impact of FSW and GMAW. Methodology Life cycle assessment (LCA) An LCA provides a systematic perspective on the environmental impacts of a product over its life cycle [18]. According to ISO 14040 methodology, it is conducted in four steps: goal and scope definition, life cycle inventory analysis, life cycle impact assessment and life cycle interpretation. By describing the goal and scope, the products and processes are stated as the framework for the following study. A basic definition herein is the functional unit for the product to be analyzed. All energy and material flows are based on this functional unit. The second step, the life cycle inventory analysis, includes the determination of environmentally relevant inputs and outputs of the processes. The inputs and outputs are product and waste flows, which need to be linked to the resulting elementary flows from and to the environment. The results of the life cycle inventory are used to determine the environmental impacts in the third step of the LCA study. In the final step, the results of the inventory and impact analyses are discussed in order to draw conclusions toward the initial goal. The goal of this LCA study is to analyze the environmental effects of FSW and GMAW when manufacturing a butt weld for

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

4

aluminum 6061-T6. The results will help researchers and manufacturing engineers to understand the environmental effects that are associated with the FSW process. Therefore, the potential for FSW to help us move toward more sustainable manufacturing will be elaborated in this study. The scope of the life cycle assessment (LCA) describes the product systems to be studied, their system boundaries and the functional unit. A product system is the collection of unit processes that are necessary to create the functional product. The product systems to be compared in this study are 6061-T6 aluminum workpieces that are welded via FSW and GMAW. The functional unit is used as the basis for a comparison between two product systems. For this study, the reference (e.g., structural design criteria) was defined as the maximum tensile force sustained by the welds. Due to the differences in ultimate tensile strength after welding by FSW and GMAW, the thicknesses of the workpieces were different (Section ‘‘Determination of workpiece thicknesses’’). The second criterion for the functional unit was to achieve the same functional weld length for both product systems. The FSW process leaves an exit hole where the tool is retracted from the workpiece. Hence, the initial material for FSW is longer than the functional weld length and the exit hole was cut off after the welding (post-processing). The GMAW process required pre-processing in terms of the edge preparation (see Section ‘‘Workpiece lengths and width’’). Therefore, four unit processes were considered: the FSW process and post-process cutting off the exit hole were part of the FSW product system; and the pre-process groove milling and GMAW process were part of the GMAW product system. Additionally, the production of raw aluminum and its environmental impacts were considered. However, since the scope of the study focuses on the defined welding product systems, other related processes were excluded from the system boundary, such as other phases of the production cycle (e.g., handling, fixturing) and equipment-related processes (e.g., maintenance). The process flow diagrams for both product systems show the modeled input and output flows (Figs. 4 and 5).

The compilation of the life cycle inventory and the life cycle impact assessment was performed with openLCA software [29] and the ecoinvent V3 database [30]. The data for the life cycle inventory originates from primary and secondary data. The inputs to the manufacturing unit processes were measured or calculated. The measured values for the experiments included the electrical power, the input masses, the shielding gas flow rate and the process time. The electricity and shielding gas consumption was calculated by means of these values. The consumption of electrode wire is calculated with the set feed rate and the process time. The gaseous emissions of GMAW depend on a variety of process parameters [31]. However, no reference values for the used process parameters were available, which is why these were not included quantitatively. The inventory data for up-stream processes as shown in Figs. 4 and 5 was acquired from the ecoinvent V3 database. It should be noted that the process of producing raw aluminum in ecoinvent does not include tempering, alloying and rolling. The effects of shielding gas production are considered by the values given in [32] because ecoinvent does not provide data for this material. The environmental impacts for each product system were calculated using the ‘‘Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts (TRACI)’’ version 2.1 of the U. S. Environmental Protection Agency [33]. Determination of workpiece thicknesses One design consideration is the maximum tensile force that can be applied during service without causing the welded assembly to fail. The welded assemblies in this study will be deemed comparable based on this design consideration. The FSW and GMAW joints should withstand identical maximum tensile forces, within a tolerance of 5%. The UTS for the as-received 6061-T6 (i.e., base metal) was measured to be 310 MPa (45 ksi). Prior research by Fehrenbacher et al. [34] showed that with the welding parameters chosen in this study FSW of 6061-T6 produces a joint with an average

Al 6061 recycling

Al 6061 production

cutt ing waste aluminum plate

welded aluminum plate

final welded product (FSW)

FSW

Cutt ing

electricity

electricity

Fig. 4. Process flow diagram for FSW product system.

Al 6061 recycling

Al 6061 production

scrap

aluminum plate

final welded product(GMAW)

grooved aluminum plate

Grooving electricity

GMAW electricity

electrode

shielding gas

Al 4043 production Shielding gas production

Fig. 5. Process flow diagram for GMAW product system.

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

5

Fig. 6. Images of final welded assemblies: (a) friction stir weld (FSW) and (b) gas metal arc weld (GMAW).

UTS = 236 MPa (34 ksi): 76% of the base metal UTS. The UTS of an aluminum 6061-T6 GMAW joint in its as-welded state is expected to be 165 MPa (24 ksi) [35]: 53% of the base metal UTS. Because these welding processes produce joints with different ultimate tensile strength (UTS), different thickness materials will have to be used in order to create assemblies that can withstand the same maximum tensile force. FSW was considered the benchmark process in this study. Full-penetration FSW seam welds were carried out in 5-mm-thick 6061-T6 samples. The tensile test samples that were used to measure the maximum tensile force were 25.4-mm-wide. Therefore, the maximum tensile force that the FSW weld can withstand was expected to be 30 kN = (0.005 m  0.0254 m)  236  106 Pa. In order to withstand the 30 kN maximum tensile force the GMAW joint must be 0.0071 m [thick] = 30 kN/(0.0254 m  165  106 Pa).

gas metal arc welds. Table 1 shows the details of the equipment and parameters used for pre-processing and welding with GMAW. Welding Three full-penetration welds were created with each process: FSW and GMAW. The FSW tool used in the experiments is made of H13 tool steel with a concave shoulder and a threaded, conical probe with three flats. The tool shoulder diameter is 15 mm, the probe (pin) diameter tapers from 7.0 mm to 5.0 mm and the probe length is 4.7 mm. The friction stir welds required only one pass to create a 5-mm-thick full-penetration weld in 6061-T6 aluminum.

Workpiece lengths and width The width of the workpieces was 102 mm (4 in.). For each test, two workpieces were butt-welded together to create an assembly that is 204 mm (8 in.) wide (Fig. 6). The length of the final welded assemblies was 152 mm (6 in.) as shown in Fig. 6. In the case of GMAW this meant depositing a 152-mm-long weld bead (Fig. 6b). FSW leaves an exit hole where the tool is retracted from the workpiece. It is common to have excess material on either end of the weld or tack on ‘‘run-on’’ and ‘‘run-off’’ tabs to the assembly. In either case, the ‘‘tabs’’ or excess material is often cut off after FSW. In this study, a 177 mm (7 in.) long friction stir weld is produced in 203 mm (8 in.) long workpieces after which 25.4 mm are cut off from either end (Fig. 7). Experimental methods Fig. 7. Schematic of the dimensions of the FSW assembly.

The wall-plug energy consumption for all processes was measured with an industrial power analyzer designed for field use (Fluke 435-II). The power analyzer was always connected to, and removed from, the equipment by a certified electrician trained in its use. All processes were recorded on video with a time stamp. The processing time was determined from a replay of the video. Pre-processes There was no pre-processing required for the FSW workpieces. The edge preparation for the GMAW workpieces was carried out in accordance with the American Welding Society (AWS) standard D1.2 [35]. A V-groove with 608 groove angle between the two workpieces was created for GMAW (Fig. 8). This was carried out by milling the edge of each workpiece at 308 from vertical in one pass. Three pairs of workpieces (6 in total) were prepared for three

Fig. 8. Schematic of GMAW workpieces with V-groove.

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

6

Table 1 Equipment and operating parameters used for GMAW. Gas Metal Arc Welding (GMAW) 6061-T6 Aluminum (initial workpiece dimensions: 102 mm  152 mm  7.1 mm) Pre-processing (edge preparation) Machine type Manufacturer Model Rated spindle power Spindle speed Feed rate Max. cutting depth Cutting length

Welding CNC mill HAAS TM-1 5.6 kW 1800 rpm 100 mm/min 4.25 mm 152 mm

Post-processing

Machine type Manufacturer Model Weld length Filler material Filler wire diameter Wire feed rate Shielding gas

GMAW welder Miller Electric Millermatic 350 P 152 mm Al 4043 0.89 mm 15.3 m/min 100% Argon

N/A

Table 2 Equipment and parameters used for FSW. Friction stir welding (FSW) 6061-T6 aluminum (initial workpiece dimensions: 102 mm  203 mm  5 mm) Pre-processing

Welding

N/A

Machine type Manufacturer Model Rated spindle power Spindle speed Weld speed

Post-processing (Cutting) CNC mill HAAS TM-1 5.6 kW 1100 rpm 400 mm/min

Different process parameters (e.g., spindle speed and feed rate for FSW) result in a variation of energy consumption and weld quality in terms of UTS. As mentioned in Section ‘‘Determination of workpiece thicknesses’’, FSW was carried out in accordance with the parameters from [34] that resulted in the strongest weld (Table 2). GMAW was carried out manually in spray-arc (pulse) mode and required two passes for the creation of a full penetration weld in 7.1-mm-thick 6061-T6 aluminum. Table 1 shows the details of the equipment used and process parameters for GMAW. In addition to the wall-plug energy consumption and process time of GMAW, the shielding gas flow rate from the flow meter, shielding gas pressure from the pressure gauge, and voltage and wire feed rate from the welding machine were also recorded. These additional measurements enabled the calculation of the total volume of shielding gas and mass of filler material that were used. Post-processes There was no post-processing required for the GMAW joints. As described in Section ‘‘Workpiece lengths and width’’, 25.4 mm strips were discarded from the assembly after FSW in order to remove the beginning and end of the joint (Fig. 7). This required two cuts for each FSW joint. A semi-automatic drop saw was used for cutting these sections. No coolant was used during the cutting process. Table 2 shows the details of machines used and operating parameters for FSW. Fig. 5 shows the schematic of the sections removed by cutting from FSW joints. Tensile testing A tensile testing machine with a 90.7 kN (20 klbs) capacity servo hydraulic load frame and load cell of the same capacity (MTS model no. 661.21A-03) was used to measure the maximum tensile strength (peak load) for all welds. The tensile testing specimens were cut from the middle section of each joint following AWS standard D1.2 [35]. The width of each tensile testing specimen was 25.4 mm (1 in.) and the thickness was the same as the original workpiece, i.e., 5 mm for FSW joints and 7.1 mm for GMAW joints.

Machine type Manufacturer Model Rated saw blade motor power Rated power hydraulic unit Cutting length

Drop saw COSEN SH-460 M 1.5 kW 0.18 kW 204 mm  204 mm per welded assembly

Results and discussion Material saving Figs. 9 and 10 show the maximum tensile force for the three FSW and three GMAW joints, respectively. In both groups of

Fig. 9. Maximum tensile force for FSW joints (‘HAZ’ and ‘Weld’ indicate where the weld fractured).

Fig. 10. Maximum tensile force for GMAW joints (‘HAZ’ and ‘Weld’ indicate where the weld fractured).

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

7

Table 3 Key inputs and outputs in life cycle inventories for manufacturing processes. FSW product system

GMAW product system

Input

Output

Pre-processing

N/A

Welding

Al 6061

530 g

Energy

127 kWs

Welded plate Energy

530 g 48 kWs

Post-processing

Input

Welded plate

530 g

Final product Waste

130 g

Table 4 Results for process duration, average power and energy consumption for FSW and GMAW product systems. FSW product system

Process duration (s) Energy consumption (kWs)

Output

Al 6061 Energy

588 g 122 kWs

Grooved plate Waste

Grooved plate Energy Electrode Shielding gas

572 g 181 kWs 14 g 2.7 L

Final product

572 g 16 g

N/A

GMAW product system

of the FSW joint and 18 g of material from the chamfer prior to creating the GMAW joint. This indicates that FSW results in a net saving of 170 g of material against GMAW for the size workpiece and design criteria chosen in this study. Energy consumption

FSW

Cutting

Groove milling

GMAW

60

44

217

34

127

48

122

181

welds (FSW and GMAW), two out of three specimens fractured in the heat affected zone and one specimen fractured in the weld zone. The average maximum tensile force for FSW and GMAW joints was 30.9 kN (6.8 klbs) and 31.6 kN (7 klbs), respectively. The difference between the average maximum tensile forces of the FSW and GMAW joints is 0.7 kN: less than the 5% tolerance level established for this comparison. This justifies the smaller workpiece thickness for FSW as compared to GMAW. The initial masses (before pre-processing) of a pair of FSW and a pair of GMAW workpieces were measured to be 530 g and 588 g, respectively (Table 3). Despite the added length of the FSW workpieces because of sacrificial material at each end they represent reduction/saving in material of 58 g. This corresponds to approximately 10% reduction in material consumption for FSW as compared with GMAW in this scenario. It should be noted that the sacrificial material at each end of the friction stir weld remains the same for different length welds. Hence, if 1-m-long welds of 6061-T6 aluminum were being compared, instead of 152-mm-long welds, the material savings of FSW would be approximately 1017 g (26%) as compared with GMAW. Therefore, the saving in material will be higher for longer weld lengths with FSW against GMAW. Table 3 shows that 130 g of material was discarded from the ends

The energy consumption for each manufacturing process was calculated from the measured instantaneous electrical power as a function of time (Table 4). The transient power levels associated with each process are depicted in Figs. 11 and 12. These are presented for one pass of each process for a representative workpiece: i.e., cutting one groove, depositing one GMAW bead, etc. As shown in Fig. 11(a), energy was consumed at a rate of 460 W (initial idle power) before the FSW tool made contact with the workpiece, and 890 W after the FSW tool disengaged from the workpiece (final idle power). During this period, energy was consumed for spindle rotation, axes movement and standby functions. The standby functions consume energy at the rate of standby power that was measured to be 350 W and was included in the idle power. After initial idle phase, the tool plunged into the workpiece. This leads to an increase in power consumption over time. The plunge power averaged over time was 1745 W. Once the desired plunge depth was attained, the tool traversed along the seam resulting in an average FSW power of 3760 W. Fig. 11(b) shows the power levels associated with the post-processing (cutting) of the FSW joint. Similar to the FSW cycle, the cutting process is preceded and followed by idle phases that accounted for saw and blade movement, and standby functions. The idle power for cutting was 760 W and actual cutting power was 1225 W. The standby power for cutting equipment was measured to be 18 W. The GMAW product system required a pre-processing step of groove milling. It can be noticed from Fig. 12(a) that milling was also preceded and followed by an idle phase. The idle power level

Fig. 11. Transient power levels associated with (a) welding and (b) post-processing (cutting) for the friction stir welding (FSW) product system.

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 8

A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

Fig. 12. Transient power levels associated with (a) pre-processing (groove milling) and (b) welding for the gas metal arc welding (GMAW) product system.

for milling was 500 W before the milling tool contacted the workpiece and 564 W after the milling tool disengaged with the workpiece. The cutting power level was 545 W. Since GMAW was performed manually, there was no idle phase involved with it. It was characterized by an almost constant power level of 5303 W, except for a peak in the beginning of the process. The average stand-by power for the welding equipment was 110 W. It should be noted that the standby power level for FSW equipment is higher than for GMAW equipment. Therefore, for a production cycle, overall energy consumption results may be different depending on the time period spent in standby mode. However, this work considers the mere manufacturing processes as system boundary. Table 4 shows the process duration, average power and energy consumption for FSW and GMAW product systems. All the values refer to a welded assembly, i.e., for cutting, groove milling and gasmetal arc welding the values contain two passes that were required for each sample. The energy consumption for the product systems is determined from the consumption of each unit processes. The average energy consumption for the FSW and GMAW product systems was 175 kWs and 303 kWs, respectively. Therefore, FSW had a reduction in energy consumption of 42% compared to GMAW. Figs. 13 and 14 show the distribution of energy consumption for the joints created in this study by FSW and GMAW, respectively. The idle energy values in these figures are for complete FSW and GMAW cycles and are not limited to the idle phases before and after actual welding, i.e., it includes the energy consumed during the weld for other purposes besides welding. Therefore, FSW traverse energy and GMAW weld energy are representative of the energy utilized in actual creation of the joint. It can be noticed from Fig. 13 that 53% of the total energy (175 kWs) was consumed during the traverse phase for FSW,

which is approximately 93 kWs. Similarly, 60% of the total energy (303 kWs) was consumed during two welding passes for GMAW, which is approximately 179 kWs (Fig. 14). This suggests that the energy required for joint formation in GMAW can be almost twice what is needed for FSW. This is due to the solid-state nature of FSW and lower temperatures involved during the process. This results in overall lower power level for FSW than the GMAW. The process time, on the other hand, is 75% longer for FSW. The difference in process times is proportional to the number of welding passes and also a function of the processing parameters used. For this case, two GMAW passes and only one FSW pass was required. For thinner workpieces only one GMAW pass may be required and for thicker welds the number of GMAW passes increases significantly, whereas FSW up to 50-mm-thickness can be done in a single pass. Therefore, it is expected that there will be a tipping point in weld thickness above which FSW would require shorter process times as compared to GMAW. It can also be noticed from Fig. 13 that 16% of the total energy consumed (175 kWs) for FSW product system was required for post-processing (28 kWs). On the other hand, 40% of the total energy consumed (303 kWs) for GMAW was required for preprocessing (121 kWs). Therefore, the difference in energy consumption between the two methods is caused by both the welding process and the supporting processes.

Fig. 13. Energy consumption distribution for FSW joints.

Fig. 14. Energy consumption distribution for GMAW joints.

Environmental impact The key components of the life cycle inventories for the manufacturing processes are presented in Table 3. The occurring pre-consumer aluminum scrap can be considered a 100% substitute for primary aluminum [36]. This means, deviating from the values in Table 3, the processes are modeled with a reduced

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx Table 5 Results for life cycle impact assessment. Impact category

Category indicator

Results product system FSW

GMAW

Acidification Ecotoxicity Eutrophication Global warming Ozone depletion Photochemical ozone formation

moles of H+-Eq. kg 2,4-D-Eq. kg N kg CO2-Eq. kg CFC-11-Eq. kg NOx-Eq.

2.7 2.1 8  10 4 6.8 1.7  10 1.7  10

3.9 2.8 11  10 9.8 2.4  10 2.5  10

7 2

4

7 2

aluminum input for the life cycle impact assessment (substitution approach). For the FSW product system, an input of 530 g aluminum is needed. The waste of 130 g is recycled and reduces the required primary aluminum to 400 g. Similarly, the material input for the GMAW product system consists of 572 g of primary aluminum and 16 g of recycled aluminum waste. This modeling approach reduces the environmental impact of the primary aluminum production while accounting for the environmental effects of the aluminum recycling. The environmental effects of both product systems were evaluated according to the indicators: acidification, ecotoxicity, eutrophication, global warming, ozone depletion and photochemical ozone formation. These indicators describe different types of impact on the natural environment [37]. Acidification refers to the acidifying effects of chemicals on water and soil in terms of the generation of hydrogen ions. Ecotoxicity describes damages on the species within earth’s ecosystems. It is expressed in equivalents to dichlorophenoxyacetic acid, which is a herbicide. Eutrophication addresses the impacts on nutrient concentration on aquatic and terrestrial ecosystems. Global warming addresses the impact of the climate and is expressed in equivalents of carbon dioxide emissions. Ozone depletion refers to damages to the ozone layer. It is assessed in terms of the release of chlorofluorocarbon emissions with reference toward trichlorofluoromethane. The photochemical ozone formation describes the impact by pollutants that oxidize organic molecules. The results for the life cycle impact assessment are shown in Table 5. In all categories, the environmental impact of the FSW product system is lower than the GMAW product system. The reduction in environmental impact ranges from 23% (ecotoxicity) to 31% (all other indicators). The difference in ecotoxicity can be explained by the higher recycling rate in FSW process since the treatment of aluminum scrap has a relatively higher ecotoxicity. The contribution of each process to the overall environmental impact of the product systems is analyzed in more detail for the impact category of global warming. Table 6 shows the comparison of global warming potential between the FSW and GMAW product systems. The total CO2 equivalent emissions are 6.78 kg for the FSW and 9.82 kg for the GMAW product system. Hence, the FSW product system leads to a reduction of greenhouse gas emissions of 31% compared to GMAW. Table 6 Comparison of global warming potential between FSW and GMAW product systems. Product system

Aluminum Electricity Electrode wire Shielding gas Total

The highest share is caused by the aluminum input for the workpieces, which is 99.2% for FSW and 96.8% for GMAW, respectively. In the GMAW product system, the aluminum input for the electrode wire has the second largest share with 2.3%. The share of electricity consumption at the environmental effects is approximately 1% for both product systems. The difference in the environmental impact of aluminum production is caused by the variation in mass input and occurring waste (see Section ‘‘Material saving’’). For the FSW product system, the primary aluminum production causes 6.64 kg CO2-eq and the recycling causes 0.08 kg CO2-eq. For GMAW product system, primary aluminum leads to emissions of 9.5 kg CO2-eq and recycling to 0.01 kg CO2-eq. If no recycling was considered, i.e., the entire material would be provided by primary aluminum, it would cause 8.8 kg CO2-eq for FSW and 9.76 kg CO2-eq for GMAW. The environmental impact of the recycling process is considerably lower than the effects of primary aluminum production. Hence, the higher share of aluminum waste of FSW product system contributes significantly to the reduced environmental impact.

Conclusions Energy consumption was measured during friction stir welding (FSW) and gas metal arc welding (GMAW) of 6061T6 aluminum including associated pre and post-processes. The workpiece dimensions were chosen such that the maximum tensile force for the joints created by two processes would be similar (31 kN). The higher ultimate tensile strength of FSW joints allowed thinner workpieces (5-mm-thick) to be used, which resulted in a 10% reduction in material consumption as compared to GMAW (7.1-mm-thick). It was discovered that the overall energy consumed in joining, and pre and post processing for FSW was about 40% less than the corresponding energy for GMAW. Life cycle assessment (LCA) methodology was used to assess the environmental impact of the overall joint creation cycle for the two joining processes. It was discovered that FSW resulted in approximately 31% less greenhouse emissions as compared to GMAW. It is well accepted and further verified in this work that the energy required for actual joint formation is less for FSW as compared to GMAW. This is due to the solidstate nature of the FSW process, i.e., the process temperature remains below the solidus temperature of the alloy. It was also observed that the pre and post processes involved with FSW consumed less energy. It is anticipated that as workpieces get longer and welds get thicker the difference in total energy consumption and environmental impact of FSW as compared to GMAW will grow. This is because pre-process (cutting grooves) associated with GMAW scales with the length and thickness of the weld, whereas the post-process cutting of each end of the weld remains constant. A full-penetration GMAW in the 7.1-mm-thick aluminum required two passes, whereas FSW produced the 5mm-thick full penetration weld in a single pass. The number of GMAW passes will increase with weld depth/thickness, whereas FSW can weld aluminum up to 50-mm-thick in a single pass. Acknowledgements

FSW (kg CO2eq)

GMAW (kg CO2eq)

6.72 0.05

9.51 0.09 0.22 0.002 9.82

6.78

9

The authors gratefully acknowledge the partial support of this work by the National Science Foundation (grant CMMI-1332738), the Department of Mechanical Engineering (graduate student fellowship), the Wisconsin Structures and Materials Testing Laboratory, the College of Engineering Student Shop at University

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001

G Model

CIRPJ-290; No. of Pages 10 10

A. Shrivastava et al. / CIRP Journal of Manufacturing Science and Technology xxx (2015) xxx–xxx

of Wisconsin-Madison, colleagues in the Laser Assisted Multi-Scale Manufacturing Lab, and the Professorship of Conveyors at Technische Universita¨t Chemnitz. This work has been carried out during a research visit of one of the authors, which was funded by the German Academic Exchange Service (DAAD) under award number 91523257-57044987. The authors are indebted to Peter Severson for helping with the electrical power measurements and Mike Hughes for making the gas metal arc welds. References [1] Thomas, W.M. (1995). Patent No. EP0615480. [2] Mishra, R.S., Ma, Z.Y., 2005, Friction Stir Welding and Processing, Materials Science and Engineering, R 50:1–78. [3] Threadgill, P.L., Leonard, A.J., Shercliff, H.R., Withers, P.J., 2009, Friction Stir Welding of Aluminium Alloys, International Materials Review, 54/2: 49–93. [4] Kalpakjian, S., Schmid, S.R., 2001, Manufacturing Engineering and Technology, Prentice-Hall, New Jersey. [5] Lakshminarayanan, A.K., Balasubramanian, V., Elangovan, K., 2009, Effect of Welding Processes on Tensile Properties of AA6061 Aluminum Alloy Joints, International Journal of Advanced Manufacturing and Technology, 40: 286–296. [6] Heurtier, P., Jones, M.J., Desrayaud, C., Montheillet, F., Allehaux, D., 2006, Mechanical and Thermal Modelling of Frcition Stir Welding, Journal of Materials Processing Technology, 171:348–357. [7] Prasad, A.B., Prasanna, P., 2013, Experimental Comparison of the MIG and Frction Stir Processes for AA 6061(Al Mg Si Cu) Aluminim Alloy, International Journal of Mining Metallurgy & Mechanical Engineering, 1/2: 137–140. [8] Balasubramanian, V. (2010, 12 28). The Indian Institute of Welding. Retrieved from http://www.iiwindia.com/technical/technical_information.htm (25.07.14). [9] Dawood, H.I., Mohammed, K.S., Rajab, Y.M., 2014, Advantages of Green Solid State FSW over the Conventional GMAW Process, Advances in Material Science and Engineering, 2014:1–10. [10] Pfefferkorn, F.E., Bello, D., Haddad, G., Park, J.-Y., Powell, M., McCarthy, J., et al, 2010, Characterization of Exposures to Airborne Nanoscale Particles During Friction Stir Welding of Aluminum, The Annals of Occupational Hygiene, 54/5: 486–503. [11] Cole, H., Epstein, S., Peace, J., 2007, Particulate and Gaseous Emissions When Welding Aluminum Alloys, Journal of Occupational and Environmental Hygiene, 4/9: 678–687. [12] Matczak, W., Gromiec, J., 2002, Evaluation of Occupational Exposure to Toxic metals Released in the Process of Aluminum Welding, Applied Occupational and Environmental Hygiene, 17/4: 296–303. [13] Lienert, T.J., Stellwag, W.L., Warke, R.W., 2003, Friction Stir Welding Studies on Mild Steel, Welding Journal, 82/1: 1-S–9-S. [14] Defalco, J., Steel, R., 2009, Friction Stir Process Now Welds Steel Pipe, Welding Journal, 88/5: 44–48. [15] Kusuda, Y., 2013, Honda Develops Robotized FSW Technology to Weld Steel and Aluminum and Applied it to a Mass-production Vehicle, Industrial Robot: An International Journal, 30/3: 208–212. [16] Mononen, J., Siren, M., Hanninen, H., 2003, Cost Comparison of FSW and MIG Welded Aluminum Panels, Welding in the World, 47/11/12: 32–35. [17] Defalco, J., 2006, Friction Stir Welding vs Fusion Welding, Welding Journal, 85/ 3: 42–44.

[18] International Organization for Standardization (ISO). 2009, ISO 14040 Environmental Management – Life Cycle Assessment – Principles and Framework, . [19] Salonitis, K., Ball, P., 2013, Energy Efficient Manufacturing from Machine Tools to Manufacturing Systems, Procedia CIRP, 7:634–639. [20] Linke, B.S., Corman, G.J., Dornfeld, D.A., To¨nissen, S., 2013, Sustainability Indicators for Discrete Manufacturing Processes Applied to Grinding Technology, Journal of Manufacturing Systems, 32/4: 556–563. [21] Goeschel, A., Schieck, F., Schoenherr, J., 2012, Method for Energy and Resource Balancing Demonstrated as an Example of the Hot Sheet Metal Production Process, CIRP Annals – Manufacturing Technology, 61/1: 399–402. [22] Briskham, P., Blundell, N., Han, L., Hewitt, R., Young, K., Boomer, D., 2006, Comparison of Self-pierce Riveting, Resistance Spot Welding and Spot Friction Joining for Aluminium Automotive Sheet, . http://dx.doi.org/10.4271/200601-0774. [23] Dilthey, U., Lueder, F., Wieschemann, A., 1999, Technical and Economical Advantages by Synergies in Laser Arc Hybrid Welding, Welding in the World, 43:141–152. [24] Duflou, J.R., Kellens, K., Renaldi, M., Guo, Y., Dewulf, W., 2012, Critical Comparison of Methods to Determine the Energy Input for Discrete Manufacturing Processes, CIRP Annals – Manufacturing Technology, 61:63–66. [25] Krones, M., Mu¨ller, E., 2013, Towards a Methodical Framework for the Environmental Assessment of Factory Systems, in Pinheiro de Lima E, Gouvea da Costa SE, (Eds.) 22nd International Conference on Production Research (ICPR). Foz de Iguassu (Brazil), July 28-August 1 2013. Pontifı´cia Universidade Cato´lica do Parana´ (PUCPR), Curitiba. [26] Zhao, F., Bernstein, W.Z., Naik, G., Cheng, G.J., 2010, Environmental Assessment of Laser Assisted Manufacturing: Case Studies on Laser Shock Peening and Laser Assisted Turning, Journal of Cleaner Production, 18:1311–1319. [27] Kellens, K., Dewulf, W., Overcash, M., Hauschild, M.Z., Duflou, J.R., 2012, Methodology for Systematic Analysis and Improvement of Manufacturing Unit Process Life Cycle Inventory (UPLCI) CO2PE! Initiative (Cooperative Effort on Process Emissions in Manufacturing) – Part 2: Case Studies, International Journal of Life Cycle Assessment, 17:242–251. [28] Murray, V.R., Zhao, F., Sutherland, J., 2012, Life Cycle Analysis of Grinding: A Case Study of Non-cylindrical Computer Numerical Control Grinding via a Unit-process Life Cycle Inventory Approach, Proceedings of the Institution of Mechanical Engineers Part B: Journal of Engineering Manufacture, 226/10: 1604–1611. [29] GreenDelta GmbH. http://www.openlca.org. [30] ecoinvent – Swiss Centre for Life Cycle Inventories. http://www.ecoinvent.org/. [31] U. S. Environmental Protection Agency. 1995, Compilation of Air Pollutant Emission Factors AP-42 Section 12.19 Electric Arc Welding, . [32] Nakhla, H., Shen, J.Y., Bethea, M., 2012, Environmental Impacts of using Welding Gas, Journal of Technology Management and Applied Engineering, 28/3: 2–11. [33] U. S. Environmental Protection Agency. Tool for the Reduction and Assessment of Chemical and Other Environmental Impacts (TRACI). http://www.epa.gov/ nrmrl/std/traci/traci.html. [34] Fehrenbacher, A., Duffie, N.A., Ferrier, N.J., Pfefferkorn, F.E., Zinn, M.R., 2014, Effects of Tool-Workpiece Interface Temperature on Weld Quality and Quality Improvement through Temperature Control in Friction Stir Welding, The International Journal of Advanced Manufacturing Technology, 71/1–4: 165–179. [35] AWS D1.2/D1.2M. (2008). [36] European Aluminium Association. 2013, Aluminium Recycling in LCA, . Brussels, Belgium. [37] European Commission. 2010, International Reference Life Cycle Data System – ILCD Handbook: Framework and requirements for Life Cycle Impact Assessment Models and Indicators, Luxembourg, .

Please cite this article in press as: Shrivastava, A., et al., Comparison of energy consumption and environmental impact of friction stir welding and gas metal arc welding for aluminum. CIRP Journal of Manufacturing Science and Technology (2015), http://dx.doi.org/ 10.1016/j.cirpj.2014.10.001