A quality and cost approach for welding process selection

A Quality and Cost Approach for Welding Process Selection César Rezende Silva

Valtair Antonio Ferraresi

Américo Scotti

Faculdade de Engenharia Mecânica. Universidade Federal de Uberlândia.

C.P. 593. 38400-902. Uberlândia. MG

[e-mail]

The aim of this work was to propose, apply and evaluate a methodical approach to select welding processes in a productive environment based on market requirements of Quality and Costs. A case study was used. The welds were carried out in laboratory, simulating the joint conditions of a manufacturer and using several welding processes: SMAW, GTAW, pulsed GTAW, GMAW with CO₂ and Ar based shielding gases and pulsed GMAW. For Quality analysis geometrical aspects of the beads were considered and for Cost analysis, welding parameters and consumable prices. Quantitative indices were proposed and evaluated. After that, evaluation of both Quality and Costs was done, showing to be possible to select the most suitable welding process to a specific application, taking into account the market conditions of a company.

Keywords: Quality, costs, cost sensibility, process selection, welding

Introduction

The rise of competition has led many companies to pay more attention in their markets in order to attend them properly. The first step is to know well the market in which the company stands, determining market requirements (what and how much) in terms of price, quality, product diversity, delivery confidence, etc. Besides, it is necessary to choose the most suitable fabrication process to a specific situation, among them welding, considering technical and economical viability. The existence of a great number of welding processes in the market, with their variances (alternating, direct and pulsed currents, etc.), makes the best process choice for a specific situation difficult. Hence, an evaluation method that helps this task is very important to the final results of any company market strategy.

A comprehensive and precise analysis to select correctly a welding process in real situations is very hard and complex, because of many variables involved. An important point is that, in market strategy, Quality and Costs, as the other requirements, need to be analyzed as a whole. It is not enough simply to determine that the process "A" is the best in Quality and the process "B" is the best in Costs. There are minimum requirements of Quality and Costs that need to be determined and reached for each case, in order to be competitive. These requirements depend mainly on the mix between product and market of the company. In fact, the best process will be that one that presents the best overall performance.

One can say that the welding Quality is related to the bead and the heat affected zone (HAZ) characteristics, enclosing presence of defects (surface finishing, spattering, cracking, porosity, degree of penetration, excessive reinforcement, etc.), mechanical proprieties (strength, toughness, hardness, etc.) and chemical composition. Quality is a relative property. To describe Quality in a quantitative way is a hard task. A good or bad Quality is a function of the requirements for a particular application.

Welding Costs seem, at first glance, to be a more measurable property. However, they involve a great number of components, such as welding execution, process selection, personnel training, joint design, equipment definition/setting and even fabrication simulation. The determination of welding Costs requires to consider welding parameters and prices of consumables, labor, equipment, etc.. One must give close attention to the relevant components of Costs during the determination and control/reduction of them. Similarly to Quality, a target for low Costs depends on the particular application.

Even considering the specificity of each company and its welded products, Cost analysis, based on a given welding condition, is relatively achievable. There are, inclusively, softwares for this purpose. However, the introduction of the Quality concept in an overall analysis brings a new challenge/perspective to welding process selection approaches.

Therefore, in this work a new approach of Quality and Costs overall analysis is proposed, according to market requirements, to select the best welding process for a company that wishes either to change or to introduce new fabrication processes.

Determination of Welding Quality

As mentioned before, the quantitative assessment of Quality is a more complex task than the one for Costs. Therefore, in this work several factors were proposed to compose the analyses of Quality. These factors are expressed as quantitative indices, as following, whose terminology is based on welding representative literature (AWS, 1987 and AWS, 1988): (a) cracks; (b) porosities; (c) undercuts; (d) penetration index; (e) convexity index; and (f) spattering index.

Some of the factors are self-explained. Others need additional description. The penetration index (PI) was defined by relating the depth of the weld bead (P) to the sheet thickness (t) (Eq.1), for which a complete joint penetration is defined as PI = 100%, an incomplete penetration as PI < 100%, and an excessive penetration as PI > 100%. The convexity index (CI) was defined as a relationship between the bead reinforcement (r) and the bead width (w), in percentage (Eq.2). The spattering index (SI – Eq.3) is defined as the ratio between the spattering rate (S – Eq.4) and the deposition rate (D – Eq.5), given in percentage. For the determination of deposition efficiency (de), Eq.6 is used.

PI = (p / t) x 100 [%]. (1)

CI = (r / w) x 100 [%]. (2)

SI = ( S / D) x 100 [%]. (3)

S = (F_elect or F_wire) – D. (4)

D = 3,6 x (Mfcp – Micp) / tarc . (5)

de = D / (Felect or Fwire) x 100 [%]. (6)

F_elect = 3,6 x (Miel – Mfel) / tarc . (7)

F_wire = 60 x (p . f 2 . f_wire . g ) / 4, (8)

where p is the weld penetration [mm], t is the joint thickness [mm], r is the bead reinforcement [mm], w is the bead width [mm], S is the spattering rate [kg/h], D is the deposition rate [kg/h], F_elect is the covered electrode fusion rate [kg/h], Miel is the initial mass of the covered electrode, before welding [g], Mfel is the final mass of the covered electrode, after welding [g], t_arc is the arc duration time [s], F_wire is the wire fusion rate [kg/h], f is the wire diameter [mm], f_wire is the wire feed rate [m/min], g is the steel density (7.85 x 10-3 g/mm3), M_fcp is the final mass of the test plate, after welding [g], M_icp is the initial mass of the test plate, before welding [g] and de is the deposition efficiency [%].

Table 1 presents the Quality criteria adopted for the welding assessment. As there is no specific standards for such application and thickness of material (< 3 mm), these criteria were defined according to generic standards, general recommendations (ANSI/AWS, 1996; ISO, 1992; IIW/IIS, 1984) and the authors’ experience, based on the expectation of a dredging pipe fabricator. In this Quality analysis, three subjective levels of Quality were adopted, namely grade A for highest Quality; grade B for an acceptable Quality for the type of product and service, and grade C for non-acceptable welds.

Determination of Welding Costs

There are many objectives to have welding Costs calculated. According to Canetti (1992), they can be used for budget elaboration and/or for comparison and selection of welding processes. Machado (1995) states that Costs determination can be used for composing sale price, helping take decisions about a product fabrication opportunity, determining the necessary investment volume for an operation, predicting modifications owing to fabrication scale changes, establishing the principles to implement a cutting Cost program and providing assistance to a welding process selection. In the present case, Costs will be used as a balancing parameter during selection of the most suitable welding process.

The Costs can be based on estimate values (estimations of amount of weld to be deposited) or on actual values (amount in fact reached in experimental tests). In this work, the actual deposited amount was used. The reason for that is that the used joint, a butt weld joint with no groove and gap, makes difficult to estimate the amount of weld to be deposited. It is important to point out that, even in case of grooved joints, each process may deposit different height of reinforcements, misconducting calculations. Therefore, to apply the approach for process selection, weldments of test plates became necessary, simulating real cases.

The composition of Costs takes into account materials, electricity, labor and equipment. Indirect Costs will not be considered, since they are approximately the same in terms of comparison. Thus:

TWC = MC + LC + EC + EPC, (9)

where TWC is the Total Welding Costs, MC is the Material Cost, LC is the Labor Cost, EC is the Equipment Cost and EPC is the Electrical Power Cost.

All Costs are expressed in R$/m (1.00 R$ was about 0.90 US$ at that time), since it seems to be the most suitable index for welding process selection applied to the study case. Material Cost involves the electrode and/or wire and the gas Costs. The equipment Cost includes the investment, the depreciation and the maintenance Costs.

The proposed mathematics equations to the determination of each term of the Total Costs (TWC) are presented in Table 2, for all welding processes under evaluation, where de represents the deposition efficiency, that is, the rate between the weld mass deposited and the melted mass of the consumable; fop indicates the operating factor (or duty factor), that is, the rate between the running arc duration time and the total welding time; ee is the electrical efficiency of the equipment, that is, a factor relating input and output power and power factor and Pm is the monthly production of weld, given by the number of hours worked in a month (176 h) multiplied by the operating factor (fop) and by the deposition rate (D). For the deposition rate (D) and the deposition efficiency (de) determinations, Eqs.6 and 8 were used, respectively.

Experimental Procedure

To evaluate the proposed approach, a case study was taken. In this case, dredging pipes are manufactured by butt welding 2-mm-thick plain carbon steels. Therefore, welding test plates were prepared using sheets of plain carbon steel (ABNT 1010), with dimensions of 250 mm x 50 mm x 2 mm. For joint configuration, a typical joint applied by the manufacturer of dredging pipes was used: butt joint with no root opening, welding on the flat position (denominated by the American Welding Society – AWS as 1G).

The welding tests were carried out using an electronic multi-process welding source and an automatic system for welding torch translation. The welding setting was as follows:

¨ Shielding Metal Arc Welding process (SMAW): 2.0-mm-diameter AWS E-6013 electrode, direct current electrode positive (DCEP);

¨ Gas Tungsten Arc Welding process (GTAW): 4.0-mm-diameter AWS WTh-2 electrode, 6-mm arc length, tip electrode angle of 45o, pure Argon shielding gas at 12 l/min, direct current electrode negative (DCEN), torch angle perpendicular to the test plate and an 1.0-mm-diameter AWS ER70S-6 wire fed from the back of the torch;

¨ CO2 shielded Gas Metal Arc Welding process (GMAWC): 1.0-mm-diameter AWS ER70S-6 wire, contact-tip-to-work distance (CTWD) of 8 mm, pure CO2 shielding gas at 12 l/min, DCEP, torch angle perpendicular to the test plate;

¨ Ar based mixture shielded Gas Metal Arc Welding process (GMAWM): 1.0-mm-diameter AWS ER70S-6 wire, CTWD of 12 mm, 8%CO2-2%O2-Ar ternary mixture shielding gas at 12 l/min, DCEP, torch angle perpendicular to the test plate;

¨ Pulsed Gas Tungsten Arc Welding process (PGTAW): 4.0-mm-diameter AWS WTh-2 electrode, 6-mm arc length, tip electrode angle of 45o, pure Argon shielding gas at 12 l/min, DCEN, torch angle perpendicular to the test plate and an 1.0-mm-diameter AWS ER70S-6 wire fed from the back of the torch;

¨ Pulsed Gas Metal Arc Welding process (PGMAW): 1.0-mm-diameter AWS ER70S-6 wire, CTWD of 12 mm, 8%CO2-2%O2-Ar ternary mixture shielding gas at 12 l/min, DCEP, torch angle perpendicular to the test plate.

For the Quality analysis, visual inspection of the beads were applied along all their extensions, aiming to find defects such as cracks, porosities and undercuts. In addition, the geometric parameters (p, w and r) of two transverse sections of each bead were measured by a computerized image analysis system. For that, the specimens were cut off from the test plates, ground and chemically etched with an iodine-based reagent.

For the Cost term calculations, current, voltage, welding speed, wire feed speed, and gas flow were set and/or monitored. Welding times and initial and final mass of the test plates were also measured (by chronometer and a digital scale).

Results and Discussion

Table 3 shows the set/monitored welding parameters of each process. It is important to point out that these values regard acceptable conditions, yet not optimized (parameter optimization was not in the scope of this work). With the data from Table 3 and Eqs.(4), (5), (6), (7) and (8), the F_elect, F_wire, D, S and de values were calculated and are presented on Table 4.

Quality Analysis

Table 5 shows the outcome from the visual analysis and geometric measurements in two transverse sections, based on the proposed criteria presented in Table 1.

If a parameter selection optimization of each process is not considered (optimized parameters could lead to different results) and one concentrates only on the objectives of the work, the following observations can be extracted from Table 5:

¨ The SMAW process presented, in general, good performance. However, its big problem was the very high spatter index (SI), reaching 56%, a value much higher than the 20% considered acceptable.

¨ The GTAW and PGTAW did not present good penetration for the welding speed used, getting Quality grade C in this topic. They presented incomplete penetration in one of the transverse sections, less than the 1.5-mm-minimum value acceptable (75% of the thick sheet). It is worth to remind that, with some changes on welding parameters, it would be possible to achieve penetrations within the acceptability criteria.

¨ The PGMAW was the process that presented the worst results. The most critical factor was lack of penetration. For this process, it was noticed the necessity of essential adjusts on the welding parameters.

¨ The GMAWC and GMAWM presented results in conformance with the acceptability criteria, taking grade A on the factors named cracks, porosities and undercuts, and B on penetration, convexity and spatter indexes. Thereby, the processes that presented the best results in Quality were GMAWC and GMAWM.

Figure 1 shows the bead transverse sections produced by the GMAWC and GMAWM welding processes.

Cost Analysis

The prices for material, labor, equipment, maintenance and electrical power applied into this analysis are listed in Table 6, whose figures were practiced on the Uberlândia-MG market at that time. Table 7 presents the calculated Total Costs and their components for each process, which are illustrated by Fig. 2. A value of 2.5% a month was considered for the interest rate (Ir) used in the equipment cost calculation. Operating factor (fop = 30% for SMAW and fop = 65% for the others) and electrical efficiency (ee = 75%) were taken based on the current literature, such as Machado (1995), Canetti (1992), The Lincoln (1973) and AWS (1987).

As can be seen in Table 7 and Fig. 2, GMAWC presented the lowest Total Welding Costs among the processes under investigation (at the welding conditions of this work). This result reflects lower Gas Cost (the cheapest shielding gas), lower Investment and Depreciation Costs (mainly due to lower equipment price) and a lower Electrical Power Cost (low current level during operation). On the other hand, Wire Cost was higher than for GTAW and PGTAW, because these latter processes use lower wire feed speeds (and, consequently, less deposited material).

The GTAW achieved a very good position (second place), because of the medium equipment value and, mainly, the low Wire Cost (low deposition rate). It is worth to mention that in this process there is no need of a great amount of deposited material (due to the joint configuration), yet a deeper penetration is required.

The GMAWM presented Total Costs 42% higher than for GMAWC. The main cause of this difference is the Gas Cost (mixture price three times higher than for CO2), followed by the Wire Cost (higher deposition rate).

Excluding SMAW, the PGMAW was the process that presented the highest Total Welding Costs, because of the high Wire Cost (high deposition rate), Gas Cost (high gas price) and Investment and Depreciation Costs (high equipment value). The SMAW process presented the highest Total Costs (already expected), reaching a value close to four times higher than for the other processes which used shielding gas. The main reason for this high Cost is the Labor Cost, owing to the low welding speed and to the low operation factor.

Another approach of analysis is to consider the weighted fraction of each Cost component (Material, Labor, Equipment and Electrical Power) in relation to the Total Welding Costs. Table 8 shows these results. As presented, the Material and Labor Costs factors had a significant influence in every process with gaseous shielding, in which these two components were responsible for more than 80% of the Costs, except for the PGTAW. The Electrical Power Cost stayed in a very low level of significance for all the processes (< 5%).

Analysis of Cost Sensitivity

As much important as to determine the Costs of a process is to define the importance of each factor into the composition of the final Cost. A mean of doing this is through Analysis of Cost Sensitivity. This analysis was carried out firstly by selecting some factors that once varied would affect the costs, but with no influence on the welding parameter settings, such as fop (operating factor), Pw (wire price), Pg (shielding gas price), Sw (welder/operator salary) and Ve (equipment value). One can predict that travel speed (t_speed), for instance, would affect significantly the final cost. However, its action on the welding parameter setting is also remarkable, that is, tspeed variation leads to a new welding setting to keep the same bead Quality. Secondly, each of those factors was systematically varied from the initial value (for example, + 10%, + 25%, +50% and + 100% or up to a reasonable increment), simulating a scenario of factor variations, and the outcome of each factor variation on the Total Welding Costs is plotted, as can be seen in Fig. 3.

Figure 3

Then, a visual analysis of the plots is employed to assess the significance of the factor variations (sensitivity). The more significant the factor, the more attention is needed, in order to get a cost optimization, i.e., significant reduction of Total Welding Costs.

Process Selection, Based on Quality and Costs

The approach proposed in this work for welding process selection uses a balance between Quality and Cost. Applying this approach to find the most suitable welding process for the given product, the results have shown that, considering Quality only, both GMAWC and GMAWM processes presented the best performances. Therefore, the chosen process must be the one between these two processes that presented, in addition, the lowest Total Welding Costs.

In relation to Costs, the GMAWC was the one that presented the lowest Total Welding Costs, followed by GTAW. The GMAWM showed Total Costs 42% higher than GMAWC, fact that leads it to a less competitiveness. Hence, the GMAWC process was the one that presented the best Quality and Cost performances, considering the welding test conditions as representative and suitable. Therefore, this is the process selected for the application.

Conclusions

From the proposed approach and systematics presented for the welding process selection, based on the best global performance of Quality and Cost analyses (in this case, applied to weld thin sheets of carbon steel), it is possible to conclude that:

The selection of the best welding process is possible to a certain industrial activity, considering the best global performance of Quality and Costs, according to market requirements of the company. The procedure utilized in this work showed to be a suitable tool to this aim;

Assuming the data used in this work as representative and the welding conditions in each process as adequate, the best process for welding the manufacturer product in study would be the GMAWC process (CO₂ pure shielding gas) and the worst one would be the SMAW;

According to the Analysis of Cost Sensitivity, the factor welder/operator salary, among the analyzed factors, was the one that most impact causes in the Total Welding Costs, but this can vary according to the process and/or welding parameter setting.

Acknowledgements

The authors wish to thank DRAGAS HEFPEL LTDA, for the support and supply of test materials, as well as the Federal University of Uberlândia, for the technical support and use of laboratories, both located in Uberlândia, MG, Brazil.

Manuscript received: August 1999. Technical Editor: Alisson Rocha Machado.

• ANSI/AWS D1.1-96, 1996, "AWS Structural Welding Code – Steel", AWS, USA, 440 p.
• AWS, 1987, "Welding Handbook - Vol.1: Welding Technology", 8th Ed., AWS, USA, 638p.
• AWS, 1988, "Guide for the Visual Inspection of Welds", AWS, USA.
• Canetti, E.E., 1992, "Custos nos processos de soldagem", In: Wainer, E., Brandi, S.D. & de Mello, F.D.H., Soldagem: processos e metalurgia, Edgard Blucher, Brazil, chap.11, pp.449-461.
• IIW/IIS, 1984, "Geometrical Defects in Arc Welded Joints in Steel Materials - Classes of Requirements", Welding in the World, Vol.22, No.1/ 2, pp.34-52.
• ISO 5817, 1992,."Arc-welded Joints in Steel: guidance on quality levels for imperfections".
• Machado, I.G., 1995, "A Economia da Soldagem", Anais do XXI Encontro Nacional de Tecnologia da Soldagem, Vol.II, Caxias do Sul, Brazil, pp.999-1013.
• The Lincoln Electric Company, 1973, "The Procedure Handbook of Arc Welding", 12th. Ed., Lincoln Electric, USA.

Publication Dates

• Publication in this collection
  15 Dec 2000
• Date of issue
  2000

History

• Received
  Aug 1999