Engineering Review: Air-cooled 6-Axis Collaborative Welder – Barcelona, Spain

In this article
- Field Report: Deployment of Air-Cooled 6-Axis Collaborative Welder in Barcelona
- Site Overview and Project Objectives
- Technical Specification: The 6-Axis Collaborative Welder
- Integrating Automated Welding into the Barcelona Workflow
- Mild Steel Welding: Process Optimization and Metallurgy
- Lessons Learned and Field Observations
- Conclusion and Deployment Recommendations
Field Report: Deployment of Air-Cooled 6-Axis Collaborative Welder in Barcelona
Site Overview and Project Objectives
The following report details the technical commissioning and performance evaluation of a 6-Axis Collaborative Welder at a medium-scale architectural steel fabrication facility in the Poblenou district of Barcelona, Spain. The facility specializes in structural frames and bespoke components using S235JR and S355J2 mild steel.
The primary objective was to transition a high-volume assembly line from manual GMAW (Gas Metal Arc Welding) to a semi-autonomous workflow. The client’s requirements were specific: reduce post-weld cleanup, ensure penetration consistency across multi-pass fillet welds, and mitigate the skilled labor shortage currently impacting the Spanish industrial sector. By implementing Automated Welding, the facility aimed to achieve a 25-30% increase in arc-on time.
Technical Specification: The 6-Axis Collaborative Welder
The heart of this installation is a high-precision 6-Axis Collaborative Welder. Unlike traditional industrial robots, this unit operates without safety fencing, utilizing force-torque sensors to detect human interference.
Kinematics and Joint Accessibility
The 6-axis configuration is critical for the complex geometries found in architectural mild steel. The first three axes handle the primary positioning, while axes 4, 5, and 6 allow for intricate torch orientations. In this Barcelona shop, we encountered several “closed-angle” joints where a 4-axis or 5-axis system would have resulted in torch-nozzle interference. The 6-axis freedom allows the system to maintain a consistent 15-degree push angle even when navigating internal corners of box sections.
Air-Cooled Torch Dynamics
We opted for an air-cooled torch system over water-cooled alternatives for this specific deployment. While water-cooled systems offer higher duty cycles, the air-cooled variant reduces system complexity and maintenance overhead. Given that mild steel welding at this facility typically involves 3mm to 8mm thicknesses, the current rarely exceeds 220A. Our field testing showed that the air-cooled neck could sustain a 60% duty cycle at 200A without significant contact tip degradation, which aligns perfectly with the collaborative “operator-load/robot-weld” cycle.
Integrating Automated Welding into the Barcelona Workflow
The transition to automated welding in a traditionally manual shop requires more than just hardware; it requires a shift in the “prep-to-weld” philosophy.
Shift from Manual to Collaborative Operations
In Barcelona, the welders were transitioned to “Cobot Operators.” The 6-Axis Collaborative Welder was integrated into a dual-zone workstation. While the robot performs a circumferential weld on a flange, the operator jigs the next component in Zone B.
The automated welding software was localized into Spanish/Catalan to ensure the shop floor team could interface with the teach pendant effectively. We utilized “lead-through programming,” where the operator physically moves the robot arm to the start and end points. The 6-axis sensors then interpolate the path, calculating the linear or circular interpolation required for the bead.
Programming and Path Optimization
A recurring challenge in automated welding is path repeatability when parts are not perfectly jigged. We addressed this by implementing a “Touch Sense” routine. The 6-Axis Collaborative Welder uses the welding wire itself as a probe to find the part’s edge before initiating the arc. This compensates for the ±1.5mm variance found in the facility’s plasma-cut mild steel blanks.
Mild Steel Welding: Process Optimization and Metallurgy
The core of the production remains mild steel welding. Despite its reputation for being “forgiving,” automating the process requires strict adherence to parameter windows.
Material Grade and Surface Prep (S235JR)
The S235JR mild steel used in the Barcelona facility often arrives with a layer of mill scale and light oxidation due to the humid Mediterranean coastal air. We observed that automated welding is significantly less tolerant of mill scale than manual welding. Manual operators unconsciously slow down to “boil out” impurities; the robot does not.
Lesson Learned: We mandated a mechanical grinding step for the weld zone. This eliminated porosity issues and ensured the 6-Axis Collaborative Welder maintained a stable arc voltage.
Parameter Stabilization for S355J2
For the thicker S355J2 sections (8mm), we utilized a pulsed spray transfer mode.
– Wire: 1.0mm ER70S-6.
– Gas: 82% Argon / 18% CO2 at 16 L/min.
– WFS (Wire Feed Speed): 8.5 m/min.
– Travel Speed: 35 cm/min.
The 6-Axis Collaborative Welder provided a level of bead ripple consistency that manual operators could not match over an 8-hour shift. This consistency is vital for the Barcelona facility’s compliance with EN ISO 5817 Level B quality standards.
Lessons Learned and Field Observations
Deployment in a real-world Spanish workshop revealed several technical nuances that “lab testing” often misses.
Cable Management and Singularity Avoidance
The most frequent error during the first week was “Robot Singularity”—a mathematical state where the robot’s axes align, causing it to freeze. This typically happened during long mild steel welding runs on cylindrical tanks.
– Solution: We re-oriented the base of the 6-Axis Collaborative Welder to a 45-degree angle relative to the workpiece. This shifted the “dead zones” away from the primary weld paths.
– Cable Management: The air-cooled umbilical is stiffer than expected. We had to install a top-mounted spring balancer to prevent the cable from dragging on the workpiece, which was causing torque-sensor trips (false collisions).
Thermal Accumulation and Duty Cycles
In the Barcelona summer, ambient temperatures in the workshop reached 34°C. The air-cooled torch’s cooling rate was slower than documented in the technical manual (which assumes 20°C).
– Adjustment: We increased the “Inter-pass Delay” by 10 seconds. While this slightly lowered the automated welding throughput, it preserved the contact tip life and prevented the wire from “bird-nesting” at the drive rolls due to heat soak.
Grounding and Electrical Noise
We encountered intermittent arc-start failures. Investigation revealed that the facility’s earthing was shared with a nearby CNC plasma cutter. The high-frequency noise from the plasma cutter was interfering with the 6-Axis Collaborative Welder‘s control signals.
– Action: Installed a dedicated ground stake for the welding cell and used shielded communication cables. The failures dropped to zero.
Conclusion and Deployment Recommendations
The deployment of the 6-Axis Collaborative Welder in Barcelona confirms that automated welding is no longer reserved for automotive giants. For small-to-medium enterprises focusing on mild steel welding, the cobot provides a scalable entry point.
Key takeaways for future deployments:
1. Surface Prep is Non-Negotiable: Automated systems cannot “see” rust or scale; clean steel is the foundation of a successful program.
2. 6-Axis Versatility: The additional degree of freedom provided by the 6th axis is mandatory for architectural steel where joint access is non-standard.
3. Environmental Context: Always factor in local ambient temperatures when sizing an air-cooled system. In Mediterranean climates, a 20% buffer on the rated duty cycle is recommended.
The facility has now moved to 24/5 operation on the cobot cell, with the 6-Axis Collaborative Welder handling all structural fillet welds, allowing the human welders to focus on complex tacking and final assembly. This synergy has successfully modernized the Barcelona workshop’s output.







