Dual XP2010-10W Robots for Teach-Free Welding of Large Stainless Steel Box Bases

Large stainless steel structures can be challenging to automate, especially when the workpiece contains long weld seams, multiple welding planes, internal corners, and strict dimensional requirements.

For this application, XARP developed a dual-robot teach-free GMAW/MAG welding solution using two XP2010-10W 6-axis welding robots combined with a vision-guided teach-free welding system.

The system is designed for a large stainless steel box base measuring 2,000 mm in width, 660 mm in height, and 500 mm in depth, with a net weight of 55 kg.

Instead of relying entirely on traditional manual teach-pendant programming, the system uses vision, CAD data, and workpiece geometry to help generate and verify welding trajectories before the two robots perform the welding operation.

Project Overview

1. Why Large Stainless Steel Box Bases Are Difficult to Weld

Automating a large stainless steel box structure is not simply a matter of installing a welding robot and programming several welding points. The combination of large dimensions, long weld seams, multi-plane joints, stainless steel heat sensitivity, and complex corner transitions creates several engineering challenges.

Long Welding Seams

With a workpiece width of 2,000 mm, welding paths can extend across a large working area. With conventional robotic welding, each new workpiece or modified structure may require engineers to manually teach numerous points using a teach pendant. For customized or small-batch production, this programming workload can become a significant part of the production preparation process.

Multi-Plane Welding

The box base contains welding areas in different directions and positions. The robot must maintain an appropriate torch orientation while moving between horizontal sections, vertical sections, internal joints, and corner transitions.

Stainless Steel Heat Deformation

Stainless steel requires careful control of welding heat input. When welding is performed sequentially on one side of a large structure, heat can accumulate unevenly. Different areas cool and contract at different rates, creating residual stress and increasing the risk of structural deformation.

Internal Corners

Internal 90-degree corners create additional challenges for robotic GMAW. Changes in robot posture, travel direction, torch angle, and welding speed can affect the weld pool and bead formation. A suitable trajectory and welding sequence are required to maintain consistent welding conditions through these transitions.

2. Why the Customer Chose a Teach-Free Welding Solution

Traditional robotic welding normally requires an operator or programmer to teach the robot by moving it through the required positions and recording welding points. For standardized products produced in high volumes, this approach can be practical. However, large customized structures create a different challenge. A new workpiece may require:

• New welding paths
• New reference points
• Different robot positions
• New torch orientations
• Additional collision checks
• Repeated test welding and path adjustments

The customer therefore needed more than a conventional welding robot. The key requirement was a vision-guided teach-free welding system that could reduce manual programming work and adapt to the actual geometry of the workpiece. This is where the combination of vision, forward modeling, reverse modeling, and dual-robot welding becomes important.

3. Vision-Guided Teach-Free Robotic Welding

The teach-free welding system can work with two different modeling approaches: forward modeling and reverse modeling. These two workflows allow the welding process to start either from existing CAD data or from the actual physical workpiece.

3.1 Forward Modeling: From CAD to Welding

When the customer already has a CAD model, the digital model can be used as the starting point.

This approach reduces the dependence on manually recording every welding point with a traditional teach pendant.

3.2 Reverse Modeling: From Physical Workpiece to Welding

Not every customized workpiece starts with a complete and ready-to-use CAD model. For customized structures or situations where the physical workpiece is the most reliable reference, the system can use a reverse modeling workflow.

This approach provides an alternative to traditional manual teaching when the actual workpiece needs to be used as the primary geometric reference.

4. Dual XP2010-10W Robot Collaboration

The second major part of the solution is the use of two XARP XP2010-10W 6-axis welding robots. For a large 2,000 mm stainless steel box base, a dual-robot configuration provides greater flexibility for accessing different welding areas and arranging the welding sequence.

• Large Working Area: The dual-robot configuration helps cover different areas of a large structural workpiece without relying on a single robot to reach every welding position.
• Parallel and Coordinated Welding: Depending on the welding sequence, the two robots can perform welding operations on different areas of the structure in a coordinated manner, reducing unnecessary waiting and improving cell utilization.
• Heat Distribution: Where the process permits, welding can be arranged in a symmetrical or balanced sequence so that heat input is distributed more evenly across the structure, helping control residual stress and reduce dimensional deviation.

5. GMAW/MAG Welding Process Control

The dual XP2010-10W system performs Gas Metal Arc Welding (GMAW/MAG) for the stainless steel box base. Three basic principles are particularly important during the welding process:

• Stable Arc Starting: The robot and welding system must establish a stable arc before entering the main welding path.
• Consistent Travel: Maintaining a controlled travel speed and torch movement helps achieve consistent bead formation along long welding seams.
• Controlled Arc Ending: The welding sequence needs to control the end of each weld to reduce defects associated with abrupt arc termination.

6. Welding at 90-Degree Corners

Internal corners are among the more demanding areas of the box base. When the robot moves through a corner, the welding trajectory changes direction while the torch orientation and travel speed also need to remain appropriate for the joint. The robot program can adjust movement speed and welding parameters during transitions to maintain a stable weld pool, achieving consistent penetration and bead formation while reducing the risk of undercut, excessive spatter, or burn-through.

7. Controlling Welding Deformation

Large stainless steel structures can be sensitive to welding deformation due to uneven thermal input. The welding sequence can incorporate:

• Symmetrical welding
• Balanced heat input
• Segmented welding sequences
• Controlled travel speed
• Appropriate inter-pass cooling
• Alternating welding locations where required

A segmented back-step welding approach can also be used where appropriate to distribute heat instead of concentrating it continuously in one area.

8. Key Benefits of the Solution

• Reduced Manual Teaching: Vision-guided path generation reduces the need to manually record every welding point with a teach pendant.
• Faster Setup for Customized Workpieces: CAD-based and reverse-modeling workflows provide more flexible ways to prepare welding paths for different workpieces.
• Improved Welding Consistency: Robot-controlled travel speed, torch movement, and welding paths help maintain consistent welding conditions across repeated production.
• Better Access to Large Structures: Two XP2010-10W robots provide a flexible configuration for handling different welding areas of large box structures.
• Improved Heat Distribution: A coordinated welding sequence can help distribute heat more evenly and reduce the risk of excessive thermal accumulation.
• Suitable for Customized Fabrication: The combination of vision and teach-free trajectory generation is particularly useful when workpiece geometry changes frequently or production involves customized structures.

9. Frequently Asked Questions (FAQ)

Q1: What is teach-free robotic welding?

A1: Teach-free robotic welding uses technologies such as vision, CAD data, and automated trajectory generation to reduce the need for manually recording individual robot welding points with a teach pendant.

Q2: How does vision-guided teach-free welding work?

A2: The vision system detects or confirms the actual workpiece geometry and position. The welding system then uses this information together with CAD data or a generated digital model to determine and verify the required welding trajectory.

Q3: What is forward modeling in robotic welding?

A3: Forward modeling starts with a digital CAD model. The system imports the CAD data, uses vision to confirm the physical workpiece position and geometry, confirms the weld seams, and generates the corresponding robot welding paths.

Q4: What is reverse modeling in robotic welding?

A4: Reverse modeling starts with the physical workpiece. The vision system captures the workpiece geometry and generates or establishes a digital representation, which is then used to confirm weld seams and generate robot welding trajectories.

Q5: Why use two welding robots for a large stainless steel structure?

A5: A dual-robot configuration can provide better access to different areas of a large workpiece, support coordinated or parallel welding operations, and provide greater flexibility for arranging a balanced welding sequence.

Q6: Can the XP2010-10W be used for stainless steel GMAW welding?

A6: Yes. The XP2010-10W is a 6-axis industrial welding robot designed for robotic welding applications, including GMAW/MAG processes for stainless steel and other materials.

Q7: What types of workpieces are suitable for this welding solution?

A7: The approach is particularly suitable for large or customized fabricated structures with long weld seams, multiple welding planes, frequent product changes, or a high manual programming workload. Examples include stainless steel frames, box structures, equipment bases, cabinets, and other fabricated assemblies.

Application Video