Welding Production, Economics and Lifecycle Management

Introduction

Welding is both a joining technology and a production system. A technically sound weld may still be commercially unsuccessful if it requires excessive labour, produces high rework, interrupts plant operation, or cannot be maintained economically. Industrial welding therefore requires simultaneous control of quality, productivity, cost, schedule, safety, asset reliability and environmental performance.

Production decisions extend beyond selecting a welding process. They include process planning, work sequencing, resource allocation, inspection, repair, maintenance, refurbishment and the eventual replacement or disposal of an asset. These decisions are connected through the asset lifecycle: design, manufacture, commissioning, operation, maintenance, modification and retirement.

This chapter presents methods for planning welding operations and evaluating their economic and lifecycle consequences. The emphasis is on decisions that a welding engineer, production engineer or maintenance manager may make in a fabrication shop, process plant, power station, transport workshop or heavy-equipment operation.

Learning Objectives

By the end of this chapter, the reader should be able to:

1. Develop a welding production plan from drawings, specifications and available resources.
2. Estimate labour, consumable, equipment, inspection, overhead and rework costs.
3. Calculate productivity measures and interpret their limitations.
4. Prepare a realistic welding schedule using precedence and resource constraints.
5. Evaluate make-or-buy alternatives using relevant cost and capacity information.
6. Distinguish corrective maintenance, preventive maintenance, repair and refurbishment.
7. Apply lifecycle cost and risk concepts to industrial welding assets.
8. Incorporate quality, safety, environmental, ethical and regulatory considerations into production decisions.

1. Welding Production as an Integrated System

A welding production system converts drawings, materials, specifications and labour into accepted welded products. Its principal inputs are:

• design data, weld symbols, tolerances and acceptance criteria;
• base materials, filler metals, shielding gases and consumables;
• welding power sources, manipulators, fixtures and tools;
• qualified procedures and competent personnel;
• inspection, testing and documentation requirements; and
• time, space, energy and financial resources.

The output is not merely a visible weld. It is a conforming product supported by traceability records, inspection evidence and, where required, a certificate of compliance.

A useful planning sequence is:

1. interpret the design and identify all welds;
2. classify welds by process, position, material and quality level;
3. select qualified welding procedures and personnel;
4. determine joint preparation, fit-up, preheat, interpass temperature and sequence;
5. estimate quantities, time and cost;
6. allocate equipment, labour and inspection resources;
7. plan quality control, safety controls and contingency time;
8. execute, inspect, document and learn from performance data.

The sequence is iterative. If the estimated production time exceeds the delivery date, the engineer may revise the process, increase parallel capacity, redesign the joint, outsource an operation or negotiate the schedule. Such changes must never bypass code requirements or reduce necessary inspection.

Important Production Terminology

Arc time is the time during which the welding arc is actually operating. Non-arc time includes fit-up, repositioning, electrode changes, cleaning, parameter checks and waiting. Travel speed is the rate at which the heat source moves along the joint. Deposition rate is the mass of weld metal deposited per unit time, commonly expressed in kg/h. Utilization is the proportion of available scheduled time spent productively on the defined operation. Rework is work performed again to correct a nonconformance. First-pass yield is the proportion of items accepted without rework.

These terms must be defined before comparing departments or suppliers. For example, a utilization percentage based only on arc time will be lower than one based on direct labour time, even though both may be correctly calculated.

2. Estimating Welding Time and Productivity

2.1 Weld-metal quantity

For a weld with cross-sectional area A and length L, the ideal weld-metal volume is:

where V is volume in mm³ if A is in mm² and L is in mm. If ρ is weld-metal density in g/mm³, the ideal weld-metal mass is:

The actual deposited mass is usually greater than the ideal mass because of reinforcement, starts and stops, spatter, trimming and dimensional allowances. If η_d is the deposition efficiency, expressed as a decimal, the mass of consumable required can be estimated by:

This formula assumes that the weld geometry and density are known and that η_d represents the ratio of useful deposited weld metal to consumed filler metal. It should not be confused with electrical efficiency or arc efficiency.

For a fillet weld with equal legs of length s, an approximate cross-sectional area is:

This approximation assumes an ideal right-triangular fillet profile and excludes convexity, concavity and reinforcement. Actual estimation should use the applicable design standard and measured profile where accuracy is important.

2.2 Deposition and arc time

If R_d is deposition rate in kg/h and m_i is required deposited weld-metal mass in kg, the ideal arc time is:

The equation applies when the deposition rate is reasonably constant and the mass is expressed on the same basis. It excludes fit-up, handling, inspection and other non-arc activities.

A practical operation time may be estimated using a utilization factor U:

where 0 < U ≤ 1 and t_o is the estimated direct operating time. The factor must be based on historical data for the same process, joint type, position, work environment and production method. A high deposition rate does not automatically produce a short total cycle time: a process may deposit metal rapidly but require extensive joint preparation, cleaning or post-weld treatment.

Worked Example 1: Time and Consumable Estimate

A fabrication contains 18 m of equal-leg fillet weld, each leg 8 mm. Assume weld-metal density 7.85 × 10⁻⁶ kg/mm³, deposition efficiency 0.90, deposition rate 4.5 kg/h and utilization 0.32.

The ideal area is:

The ideal volume is:

Therefore:

Estimated consumable mass is:

Arc time is:

Estimated operating time is:

The result does not mean that the job will finish in exactly 3.14 hours. It excludes unusual fit-up problems, waiting for inspection, crane availability, preheat, repairs and breaks. A schedule should add separately estimated inspection and contingency allowances rather than concealing them in an unjustified utilization factor.

2.3 Productivity measures

Common measures include:

where U_a is arc-time utilization and t_s is scheduled time.

where P_d is deposited weld-metal productivity in kg/h of direct labour time and t_d is direct labour time.

where FPY is first-pass yield, N_a is the number of units accepted without rework, and N_t is the total number inspected.

A broader equipment effectiveness measure may be expressed as:

where A_v is availability, P_r is performance rate and Q_r is quality rate, each expressed as a decimal. This multiplicative form is applicable when the three factors are defined consistently and independently enough for diagnostic use. It should not be treated as a universal measure of welding productivity. For job shops, product mix, setup time and engineering changes may make direct comparisons misleading.

Good performance analysis reports the denominator and boundary of every metric. “Labour productivity increased by 15%” is incomplete unless the report states whether the basis was arc hours, paid hours, completed weld length, accepted assemblies or mass of deposited metal.

3. Welding Costing

3.1 Cost structure

A welding cost estimate normally contains:

1. Direct labour: welding, fitting, handling and directly assigned supervision.
2. Consumables: filler metal, shielding gas, backing material, abrasive wheels and cleaning chemicals.
3. Equipment: machine time, fixtures, manipulators, cranes, power and depreciation or hire.
4. Quality costs: inspection, testing, calibration, documentation and certification.
5. Preparation and finishing: cutting, beveling, preheating, post-weld heat treatment, grinding and coating.
6. Rework and scrap: correction of defects and material that cannot be recovered.
7. Indirect overhead: facility, planning, maintenance, administration, insurance and information systems.
8. Risk allowance: a separately identified provision for uncertainty, not an arbitrary inflation of every rate.

A basic job cost is:

where the terms are labour, consumable, equipment, quality, preparation/finishing, rework/scrap and overhead costs. The categories should be adapted to the organization’s accounting system, but double counting must be avoided.

For labour:

where r_L is the fully loaded labour rate in currency per hour and t_L is labour time in hours. “Fully loaded” may include wages, benefits and applicable employment costs; overhead should not be included again if it is already embedded in r_L.

For consumables:

where q_i is the quantity of consumable i and p_i is its unit price. Gas cost may be estimated from flow rate and time:

where Q_g is gas flow in L/min, t_g is time in minutes and q_g is gas volume in litres. The calculation assumes stable flow and does not include losses from leaks or purge systems unless separately estimated.

Worked Example 2: Comparative Cost

A repair requires 6.0 direct labour hours at 42 currency units per hour, 7.5 kg of filler at 6.80 per kg, shielding gas costing 18, equipment at 24 per operating hour for 4.0 hours, inspection costing 95, and preparation costing 60. Expected rework is 5% of the subtotal before rework. The subtotal is:

Expected rework allowance is:

Thus the estimated cost before any general overhead or contingency is:

The result is an expected cost, not a guaranteed invoice. The estimate should identify whether tax, transport, engineering review and downtime are included.

3.2 Cost of quality and failure

Prevention and appraisal costs are often less visible than failure costs. A defective weld may create internal failure costs through repair, retesting and schedule disruption. External failure costs may include warranty claims, production loss, environmental damage, injury or loss of reputation. Quality economics therefore supports investment in qualified procedures, welder training, fit-up control and timely inspection.

A low quoted price can be misleading if it excludes lifecycle consequences. For safety-critical work, the relevant objective is not minimum initial cost but acceptable total cost at an acceptable level of risk.

4. Scheduling Welding Operations

A welding schedule should represent precedence, resources and uncertainty. Typical precedence relationships are material receipt → cutting → forming → fit-up → welding → inspection → repair, if required → coating → final release. Some operations can run in parallel, while others require the same crane, welder, furnace or inspection technician.

A useful scheduling method is:

1. list activities and their estimated durations;
2. identify predecessor activities;
3. identify scarce resources;
4. establish a network or bar chart;
5. calculate earliest and latest feasible times;
6. identify the critical path;
7. add inspection hold points and realistic contingency;
8. update the schedule from actual progress.

In a precedence network, the critical path is the sequence of activities with the greatest total duration. Delay to a critical activity directly threatens completion unless resources or methods are changed. Noncritical activities possess float, but float is not permission to consume all available time: shared resources may make apparently independent activities coupled.

Welding schedules should include:

• material and consumable availability;
• welder qualification and shift allocation;
• preheat and interpass monitoring;
• equipment setup and changeover;
• crane and manipulator bookings;
• nondestructive examination and report time;
• repair loops based on historical defect rates;
• post-weld heat treatment and controlled cooling;
• weather or environmental restrictions; and
• document review and client release.

Schedule compression can be achieved through parallel work, improved fixtures, modular fabrication, process automation or additional shifts. It should not be achieved by omitting inspections, reducing preheat below the qualified procedure, or assigning work to unqualified personnel.

5. Make-or-Buy Analysis

Make-or-buy analysis compares internal production with external procurement. The relevant comparison is incremental economic cost and risk, not simply the internal accounting charge.

For an internal alternative:

For an external alternative:

where variable cost, avoidable fixed cost, opportunity cost and risk cost are defined consistently. Sunk costs—costs already incurred and unchanged by the decision—should not influence the choice.

A supplier’s price must be evaluated with technical and commercial criteria: procedure qualifications, welder qualifications, traceability, inspection capability, capacity, delivery reliability, cybersecurity where digital records are exchanged, financial stability and control of subcontracting.

Short-term capacity example

An internal shop can produce a component for 1,850 currency units in variable cost. An additional 300 units of fixed cost would be incurred only if the work is performed internally. A supplier offers the component for 2,050, including delivery. Internal capacity is available and no other work is displaced.

Since 2,050 < 2,150, buying is economically preferable by 100 units, provided quality, delivery and contractual risks are equivalent. If the internal work would displace a contribution of 500 from another job, the opportunity cost changes the decision and buying becomes substantially more attractive.

For long-term decisions, fixed costs, learning effects, capital investment, supplier dependence and strategic capability must be modelled over the planning horizon. A shop may rationally retain a higher-cost internal capability for emergency repair, intellectual property protection or safety-critical control.

6. Maintenance, Repair and Refurbishment

6.1 Maintenance concepts

Corrective maintenance restores equipment after failure. It may be appropriate for inexpensive, noncritical items where failure consequences are low. Preventive maintenance is performed at predetermined intervals or usage levels. Condition-based maintenance uses evidence such as vibration, thermal images, oil analysis, electrical measurements or inspection results to determine when intervention is required. Predictive maintenance applies trend analysis or models to forecast degradation; it is not automatically superior unless the data are reliable and the decision improves outcomes.

For welding equipment, maintenance may include power-source inspection, cable and connector checks, torch-liner replacement, wire-feed calibration, gas-flow verification, cooling-system service, fume-extraction testing, fixture alignment and periodic electrical safety tests. Maintenance records should identify equipment, date, findings, corrective action, responsible person and next due date.

Reliability can be described using mean time between failures (MTBF) and mean time to repair (MTTR):

where A_i is inherent availability under the stated maintenance model. The equation assumes repeated operating and repair cycles and excludes logistics delays, planned shutdowns and external constraints. Operational availability may be lower because it includes waiting for parts, technicians and production access.

6.2 Repair versus replacement

A repair restores function, usually with limited intervention. Refurbishment is a broader process that may include dismantling, inspection, dimensional restoration, replacement of wear parts, upgrading controls, coating and recertification. Replacement provides a new or alternative asset.

The decision should consider:

• remaining structural life;
• failure mode and damage mechanism;
• access and outage duration;
• availability of qualified repair procedures;
• material compatibility and residual stresses;
• inspection and acceptance requirements;
• safety and environmental consequences;
• future duty, loads and operating temperature;
• spare parts and obsolescence; and
• total lifecycle cost.

Weld repair is not automatically acceptable merely because a defect can be filled. Repeated repairs may increase restraint, hydrogen risk, distortion, residual stress or local loss of toughness. Engineering assessment may require fracture mechanics, fitness-for-service analysis, hardness testing, metallography or pressure testing.

7. Lifecycle Cost and Decision-Making

Lifecycle costing evaluates costs incurred throughout ownership rather than only fabrication price. A simplified present-value model is:

where C_0 is initial cost, C_t is cost incurred in year t, i is the discount rate per year, SV_n is salvage value at year n, and n is the analysis period. Costs must be stated in consistent monetary terms. The model assumes a chosen discount rate and a defined analysis horizon; uncertainty in energy prices, failure rates and inflation should be tested through scenarios.

For a simple two-year example, an asset costs 10,000 initially, requires 1,500 maintenance at the end of year 1 and 2,000 at the end of year 2, and has salvage value 1,000 at the end of year 2. At a 10% discount rate:

This is a present-value estimate. It excludes downtime, risk and tax unless those are explicitly added. A cheaper asset may have a higher lifecycle cost if it consumes more energy, fails more often or requires expensive specialist repairs.

Risk-adjusted decisions

Where failure consequences differ, expected monetary value may be used as one input:

where p_j is the probability of consequence scenario j and C_j is its cost. This method assumes probabilities and consequences can be estimated meaningfully. It should not be used to justify unacceptable safety risks merely because the numerical average appears low. Regulatory limits, safety integrity requirements and ethical duties can impose constraints above and beyond cost optimization.

A decision matrix can combine cost, technical feasibility, reliability, safety, environmental impact, schedule and supplier risk. Weighting should be transparent, and stakeholders should be able to challenge assumptions.

8. Applied Case: Refurbishment of a Process-Pump Base

The following is an illustrative case based on a common industrial decision pattern.

Background

A chemical plant has a welded carbon-steel pump base showing cracks near a stiffener and corrosion loss around a drain opening. The pump supports continuous production, and an outage is planned for five days. The base is 14 years old, but the pump and motor are serviceable.

Problem

The maintenance team must choose among local weld repair, complete refurbishment, or replacement of the base. The base carries vibration and alignment loads, so a visually acceptable weld alone is insufficient.

Analysis

Inspection begins with cleaning, visual examination, dimensional measurement, magnetic-particle testing of accessible surfaces and ultrasonic examination of relevant thicknesses. The crack is traced beyond its visible end. Engineers review original drawings, material certificates, operating loads and the history of repeated repairs. The investigation identifies a stress concentration at an abrupt stiffener termination and inadequate drainage as contributors to cracking and corrosion.

A direct repair would have the lowest initial material cost, but it would leave the geometric stress concentration and provide little opportunity to correct drainage. Refurbishment would remove the cracked region, modify the stiffener termination, renew corroded plate, improve drainage and restore alignment surfaces. Replacement would reduce uncertainty but requires fabrication, dimensional verification, transport and more extensive disconnection.

Approach and solution

The selected option is refurbishment. The repair plan specifies removal of the crack to sound metal, qualified welding procedure selection, controlled preheat where required, low-hydrogen consumables, weld sequencing to limit distortion, in-process inspection, final nondestructive examination and dimensional alignment checks. The drainage modification is incorporated into the fabrication drawing. A hold point is established before closing the repair so that inspection personnel can verify preparation.

The plant also changes the inspection interval and records vibration and alignment data as condition indicators. The maintenance database links the asset, repair drawing, weld procedure, inspection reports and future observations.

Outcome

The work is completed within the outage. The initial cost is greater than a simple weld patch, but the modified geometry removes a known cause of recurrence. The decision avoids premature replacement of a serviceable pump and creates a traceable basis for future maintenance.

Learning points

1. Failure mechanism must be addressed, not merely the visible defect.
2. Cost comparison must include outage, recurrence and inspection consequences.
3. A repair plan requires engineering acceptance criteria and documentation.
4. Refurbishment can be a controlled lifecycle strategy between patch repair and replacement.
5. Condition data can improve the next maintenance decision.

9. Quality, Safety, Environmental and Ethical Governance

Production economics must operate within legal and professional boundaries. Applicable pressure-equipment, structural, occupational-safety, environmental and welding standards establish minimum requirements. The responsible organization must identify which edition and jurisdiction apply and retain objective evidence of compliance.

Safety controls include hot-work permits, fire watch, gas-cylinder management, ventilation and fume control, electrical isolation, confined-space procedures, fall protection and control of combustible materials. Economic pressure must never reward bypassing these controls.

Environmental decisions include filler and steel waste, shielding-gas use, electricity, grinding dust, coating emissions and disposal of contaminated components. Consumable efficiency and repair prevention can reduce environmental impact, but recycling or reuse must not compromise traceability or material performance.

Ethical practice requires accurate time and cost reporting, honest disclosure of defects, protection of inspection independence and refusal to certify work outside one’s competence. Artificial intelligence may help classify defects, forecast equipment failure or optimize schedules, but its outputs require validated data, human review, cybersecurity controls and documented accountability. An algorithm should not silently replace a qualified engineering judgment where safety is at stake.

10. Contemporary Developments and Best Practice

Current industrial practice increasingly combines robotic welding, digital work instructions, automated parameter recording, sensor-based monitoring and manufacturing-execution systems. These technologies can improve repeatability and traceability, but benefits depend on sound process qualification and reliable data. Automation is most attractive where joint geometry, production volume and access are stable; it may be uneconomic for highly variable repair work.

Best practice includes:

• use measured shop data rather than optimistic handbook rates;
• separate arc time, direct labour time and calendar duration;
• identify assumptions and exclusions in every estimate;
• include inspection, repair loops and resource constraints in schedules;
• evaluate suppliers on technical capability as well as price;
• maintain welders, power sources, fixtures and inspection equipment;
• preserve complete material, procedure and inspection traceability;
• assess the failure mechanism before approving repair;
• use lifecycle cost and risk rather than purchase price alone; and
• review completed jobs to update rates, defect probabilities and lessons learned.

Chapter Summary

Welding production is an integrated system in which technical quality, cost, schedule, safety and lifecycle performance are interdependent. Weld quantity can be estimated from geometry, density and deposition efficiency, while time estimates must distinguish arc time from non-arc activities. Productivity metrics are meaningful only when their definitions and boundaries are explicit.

A robust cost estimate separates labour, consumables, equipment, quality, preparation, rework and overhead. Scheduling requires precedence analysis, resource allocation, inspection hold points and realistic contingency. Make-or-buy decisions should use relevant incremental costs, opportunity costs and technical risk, while excluding sunk costs.

Maintenance may be corrective, preventive, condition-based or predictive. Repair and refurbishment decisions must address damage mechanisms, structural integrity, inspection and future duty. Lifecycle costing discounts future costs and may be supplemented by risk analysis, but safety and regulatory constraints cannot be traded away through simple monetary optimization. Digital and AI-enabled tools can improve production control when supported by validated data, human oversight and accountable governance.

Key Terms

Welding production: Coordinated conversion of materials, labour and procedures into accepted welded products.

Arc time: Time during which the welding arc is operating.

Utilization: Productive operating time divided by a defined available or scheduled time.

Deposition rate: Mass of weld metal deposited per unit time.

Deposition efficiency: Ratio of useful deposited weld metal to consumed filler metal.

First-pass yield: Percentage of inspected units accepted without rework.

Rework: Work performed to correct a nonconformance.

Direct labour: Labour directly attributable to producing or repairing the job.

Overhead: Indirect cost supporting production but not directly assigned to one weld.

Cost of quality: Prevention, appraisal and failure-related costs associated with conformity.

Critical path: Longest duration sequence of precedence-linked activities.

Float: Time by which an activity may be delayed without delaying the defined project completion.

Make-or-buy analysis: Comparison of internal production and external procurement alternatives.

Opportunity cost: Value of the best alternative use of a constrained resource.

Corrective maintenance: Maintenance performed after failure or identification of a fault.

Preventive maintenance: Planned maintenance performed at predetermined intervals or conditions.

Refurbishment: Restoration and improvement of an existing asset to extend useful service.

Lifecycle cost: Present-value cost of acquiring, operating, maintaining and retiring an asset.

Condition-based maintenance: Maintenance triggered by evidence of actual equipment condition.

Traceability: Ability to relate materials, procedures, personnel, inspection and records to the finished work.

Review Questions

Short Answer Questions

1. Distinguish arc time from non-arc time in a welding operation.
2. Why must utilization factors be defined with a clear time denominator?
3. Identify four categories normally included in a welding cost estimate.
4. What is the difference between repair and refurbishment?
5. Why are sunk costs excluded from a short-term make-or-buy decision?

Long Answer Questions

1. Explain how a welding engineer develops a production plan from design information through final release.
2. Discuss the strengths and limitations of deposition rate, utilization, first-pass yield and OEE as welding productivity measures.
3. Describe how precedence, scarce resources, inspection hold points and repair loops affect a welding schedule.
4. Compare corrective, preventive, condition-based and predictive maintenance for welding equipment and industrial assets.
5. Explain how lifecycle costing and risk analysis improve decisions between repair, refurbishment and replacement.

Analytical / Application Questions

1. A shop must choose between two welding processes with different deposition rates, utilization factors, consumable efficiencies and inspection requirements. Develop a calculation framework and explain which nonfinancial factors must also be compared.
2. An internally fabricated component has a variable cost below a supplier’s quotation, but it would displace a more profitable job and require additional fixed equipment. Analyse the make-or-buy decision using relevant cost and opportunity cost.
3. A repeatedly repaired welded support develops a new crack during service. Propose an investigation and lifecycle decision process that addresses failure mechanism, structural integrity, outage time, safety, cost and future monitoring.

Multiple Choice Questions with Answers

Question 1

Which quantity is calculated by dividing deposited weld-metal mass by deposition rate?

A. Ideal arc time

B. First-pass yield

C. Equipment availability

D. Salvage value

Correct Answer: A. Ideal arc time

Question 2

A weld has an ideal deposited mass of 4.5 kg and a deposition efficiency of 0.90. What consumable mass is estimated?

A. 4.05 kg

B. 4.50 kg

C. 5.00 kg

D. 5.40 kg

Correct Answer: C. 5.00 kg

Question 3

Which cost should normally be excluded from a short-term make-or-buy decision when it is unchanged by the decision?

A. Supplier transport cost

B. Incremental consumable cost

C. Opportunity cost

D. Sunk cost

Correct Answer: D. Sunk cost

Question 4

What is the primary significance of the critical path in a welding project schedule?

A. It identifies the lowest-cost weld process

B. It identifies the longest precedence-linked sequence controlling completion

C. It identifies all activities requiring nondestructive testing

D. It identifies the highest deposition-rate operation

Correct Answer: B. It identifies the longest precedence-linked sequence controlling completion

Question 5

Which maintenance approach uses measurements such as vibration or thermal condition to determine intervention timing?

A. Condition-based maintenance

B. Breakdown-only maintenance

C. Fixed-price maintenance

D. Corrective design

Correct Answer: A. Condition-based maintenance

Question 6

In the simplified lifecycle-cost equation, why are future costs discounted?

A. To remove all uncertainty from the estimate

B. To express future monetary amounts in present-value terms

C. To increase the apparent salvage value

D. To eliminate maintenance costs

Correct Answer: B. To express future monetary amounts in present-value terms

Question 7

A job has a subtotal cost of 800 currency units and an expected rework allowance of 4%. What is the allowance?

A. 16 units

B. 32 units

C. 40 units

D. 200 units

Correct Answer: B. 32 units

Question 8

Which statement best describes first-pass yield?

A. The fraction of scheduled time spent welding

B. The mass of filler deposited per hour

C. The percentage accepted without rework

D. The ratio of repair cost to purchase price

Correct Answer: C. The percentage accepted without rework

Question 9

Why may a simple weld patch be unsuitable for a recurring crack?

A. A weld patch always has zero strength

B. It may not remove the underlying failure mechanism or stress concentration

C. Inspection cannot be performed on repaired welds

D. Refurbishment is always cheaper

Correct Answer: B. It may not remove the underlying failure mechanism or stress concentration

Question 10

Which principle is ethically and technically appropriate when using AI for welding quality decisions?

A. Accept every algorithmic result without review

B. Use AI only when it replaces all qualified personnel

C. Require validated data, human oversight and accountable decisions

D. Ignore traceability if prediction accuracy is high

Correct Answer: C. Require validated data, human oversight and accountable decisions