ROBOTICS IN STRUCTURAL STEEL FABRICATION

The structural steel fabrication industry is poised for a key year. U.S. stimulus spending on infrastructure will reach its highest level in 2010, with over $20 billion earmarked for various infrastructure projects around the country. Firms will continue pursuing welding productivity improvements and will remain vigilant in looking toward technology and advanced processes that can provide a competitive advantage.

While CNC was introduced as a first generation of automation for structural applications, (typically 2D cutting applications on flat stock) a number of challenges prohibited robots from becoming the next generation, even though the tolerances required in the steel fabrication industry are easily achievable with robots.

The historical challenge has been that robots are highly precise, and yet structural steel fabrications are not always precise or repetitive. Indeed, product in structural steel fabrication is not a small batch business ? each weldment is unique. So even though robots may be more flexible, economical, efficient, agile, and mobile in adapting to various operations, reprogramming the robot for each application has been cost-prohibitive.

Standard off-line programming tools have been introduced to reduce some of the manufacturing inefficiency associated with the programming time that is unique to each part. This allows the programmer the ability to simulate a robotic arc welding and cutting application in 2D or 3D space. The user imports a work-piece CAD file and is able to create weld paths with proper torch angles and process parameters.

All programs and settings from the virtual work-cell are then transferred to the robot to increase productivity. More than 500 copies of this type of software have been sold in recent years, many with repeat buyers, and many fabricators have used this process to manage ?what-if? scenario simulations to improve cost and throughput predictability.

In the interim, translation software has entered the market that can automatically program the robot to weld every intersection that exists in a 3D model. The user is required to review each of the potential weld joints, eliminate the unnecessary joints, repair simulated collisions, and then enter information for each of the weld joints, such as the actual size and length of the welds, whether they are skip-welds or continuous, proper sequencing, etc.

Then, after performing a lengthy touch sensing sequence to manage part variability, the robot is finally able to weld the part. This concept requires a fairly expensive and dedicated system. Few fabricators comprehend that this format will bring structural steel fabrication to a totally new, higher level of production, and so the search continues.

The entry of Building Information Modeling (BIM) technology (4D CAD) is becoming more widely accepted, and appearing more often in magazines related to construction and structural steel fabrication. CIS/2 is the file exchange format that facilitates BIM for structural steel and was adopted about ten years ago as a standard for data interchange among different software for structural steel design, analysis, detailing, and fabrication.This file exchange format is the key element for turning the switch on to a mass conversion from CNC machinery to welding and cutting robots.

Design data in this 4D format includes all information needed in each phase of data processing: project and product definition, part geometry, and includes weld positions and descriptive attributes. This high level of information allows welding and cutting robots to calculate the approach angles and the correct tool paths ?on-the-fly?, along with the selection of the necessary parameters such the welding process, gas mixture, wire diameter, voltage, wire feed speed and travel speed to make the required cut or weld.

BIM software allows the ability to automatically reproduce all of the required layout information to a part, resulting in a lower level of waste, and an elimination of shop drawings. This is a benefit because there is no opportunity to lose information in the translation of drawings or specifications ? it?s a straight line of information flow from the designers to the robotic welding / cutting workcell, and the software is intelligent enough to provide advance notice to events where there may be conflicts with weld-sequencing or potential collisions.

Fabricators will see this as a competitive advantage and as a way to realize a higher operating factor and an overall improvement in the quality of the fabrication. It is with these expanded capabilities that off-line simulation can be accomplished more quickly, easily, and efficiently.

Many details need to be worked out, even as this capability expands the ability to program the robot for welding and cutting applications. Tolerances and accuracy of layout and operations are still required ? welding procedures still operate best when the tolerance of the parts remains less than .020 in. Without this level of tolerance, intelligent robotic options that have been introduced in the last few years are primed to solve this remaining challenge. Integrated robot vision, therefore, becomes an increasingly important component of many automation opportunities in the structural steel industry to manage variation. Ease of programming and cost effectiveness has improved in recent years with the tight integration of vision to robotics.

Robots can use a vision ?snap-shot? sensor to ?see? the location and orientation of parts in a fraction of a second versus 5 to 20 seconds for some touch sensing sequences. The camera can examine and verify part fit-up, find features pre-weld, and measure the joint position. ?Snap-shot? vision systems can be commonly applied for multi-pass welding sequence management and can also be used for error-proofing which relates to the ability of a system to either prevent an error in a process or detect it before further operations can be performed. It can be performed on every weld in a process or to monitor critical welds of a process.

As a final expansion of technology, robots are increasingly integrating digital technology to network welding equipment to bring data from the factory floor to the business productivity arena. Production monitoring enables any networked power source to be set up so that users can monitor weld data, store and share files, monitor production task, set weld limits and tolerances, track consumable inventory, perform serial number traceability, and perform diagnostic troubleshooting remotely.

To summarize, the ability to successfully implement state-of-the-art automation technology is essential for the modern fabrication shop. Four main tasks for automated robotic cutting and welding of structural steel can now be accomplished with BIM (4D) methodology, CIS/2, and robotic welding systems. The tightly integrated systems can:
(1) recognize features such as copes, holes, and welds and automatically generate tool paths ?on-the-fly? for scribing, cutting and welding,

(2) recognize sub-weldments associated with a member assembly and generate the tool path needed to mark their position and piece mark on the main member material,
(3) quickly simulate in 3D and evaluate alternative methods for automation, and
(4) use simulation technologies to provide easy and accurate preparation and verification of complex programs for robots and CNC machinery.

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

Geoff Lipnevicius is the operations manager for the automation division at The Lincoln Electric Company, 22800 Saint Clair Avenue, Cleveland, OH 44117-8542, 216-383-8027, Fax: 216-383-8823, www.lincolnelectric.com, geoff_lipnevicius@lincolnelectric.com.