Arc Welding Automation Projects

Insider secrets to help your customer determine the best system for their requirements

Understanding the benefits of automating a welding process is easy:  faster, simpler, better, prettier, stronger, safer and more. Manufacturers know that already, which is why everyone who can automate tries to. What is not as easy to understand is the difference between a part that is good for automation from one that is not so good. Manufacturers can struggle with the decision to automate, often stuck in the analysis paralysis of not knowing how to evaluate and which factors are key. Manufacturers that rush in haphazardly can end up with an automated system that does not meet their requirements, a situation that unfortunately occurs far too often with dusty robots in storage facilities as proof.

Distributors already working with automation know that the manufacturer end-user wants and needs help in determining the best system for their requirements. For distributors just entering this market, the decision by your customer to automate can be made much clearer and much more confidently by adopting the following Nine Automation Evaluation Key Points. These are not hard and fast rules, merely points to consider. It has been said that exceptions prove the rule. During 25 years of experience in welding automation, there has been only one project which completely satisfied all nine points. It was a high volume, partial penetration and partial length fillet weld on a small but thick and accurate mild steel part. It was robotically welded using gas metal arc (GMA). All the other projects had some sort of challenge which must be addressed with tooling, programming, adaptive control, weld process changes or redesign of part.

Use the following list of nine key points to identify and address issues during the planning stages to insure a successful weld automation project.

  1. Production Volume
  2. Part Size
  3. Part Accuracy
  4. Weld Acceptance Criteria
  5. Weld Process
  6. Joint Configuration
  7. Material Thickness
  8. Base Material
  9. Technical Capability of Facility

Production Volume

First, the production rate of a part needs to be evaluated to ensure the rate is sufficient to justify the expense of automation. Generally parts with low production rates are not cost effective for automation. Evaluating the production rate also helps determine the scope and level of automation. For example, intermediate production volumes are often good fits for flexible automated systems, such as robotic cells, that can be programmed to weld a broader range of parts to maximize the system’s utilization. Parts with higher quantities are typically ideal for dedicated systems that are designed to be highly efficient at welding one type of part, or a family of parts with slight variation.

Part Size

The size and weight of the part is evaluated next to determine the work envelope and load capacity of the system. Larger parts require robots with larger work envelopes, longer slides, possibly a robot transport unit with longer linear track, and larger positioning equipment with greater load capacities. A large part size commonly affects part accuracy, one of the most important issues in automation and the next key point.

Also, large or complex parts require more programming. Each start, stop, or change in a weld path’s direction is an additional line of software to be written and debugged. This means more programming time which affects cost. Multiple layer welds mean that each layer must be programmed only after the preceding layers are programmed and welded. Often large or complex parts violate the KISS principle, or “Keep It Simple Stupid.”  A simple project has a better chance of succeeding.

Part Accuracy

A part’s accuracy has the most influence over repeatability, and is therefore one of the most important areas of concern when evaluating an automation project. Part accuracy is even more so crucial in automated welding, where any deviation greater than one wire diameter can be the difference between a good weld and a reject. Thermal expansion is also a source of dimensional error, particularly with long parts and high preheat temperatures.

Without some form of adaptive control, the weld torch’s programmed path will be the same every time, so it requires a part with accurate clamping points and geometry for the positioning and fixturing to be accurate. Joint gaps must also be consistent to tolerate the welding process.

It is still possible to automate parts with less than ideal accuracy using adaptive control, or a way for the system to detect inaccuracies and make adjustments accordingly. Common forms of adaptive control include vision, touch sensing or laser sensing. Adaptive control requires some searchable feature on the part. Searching routines add to overall cycle times, lowering the production rate. Adaptive control adds to the cost of the automated system through additional hardware, software, programming time and complexity.

Weld Acceptance Criteria

Weld acceptance criteria is a key variable that requires close consideration. Parts with partial penetration joints are preferred, whereas full penetration welds are more complex and require machined joints. Partial length welds are also preferred. Welding to an absolute edge rather than stopping just short of it increases the degree of difficulty and programming complexity. The aesthetic appearance criteria frequently dictates the welding process selection. When appearance is a primary criteria, gas tungsten arc or plasma transferred arc is recommended over gas metal arc welding. Additional criteria such as leak proof welds or sanitary welds also present additional problems that should be considered.

Welding Process

The welding process is evaluated to fit the best process for a project. Each process is unique in required equipment and operation costs, the degree of difficulty to automate, deposition rates and limitations. Furthermore, as technology advances, innovations can push some processes to the forefront that were previously not considered ideal for certain applications.

Gas metal arc (GMA) welding offers low cost, simplicity, high deposition rates, and is more tolerant of a misaligned part. GMA, however, is limited with which alloys can be welded with it due to available wire chemistries. Also ,weld quality and appearance may not be as good as other processes like gas tungsten arc (GTA).

In addition to better weld quality and appearance, GTA typically has lower defect rates, which is critical for expensive parts. GTA deposition rates are much lower than GMA; however, rates can be improved with techniques like adding hot wire, weld parameter optimization, pulsed power, or twin wire.

Furthermore, GTA requires the wire feed to rotate following the direction of the weld, which may require an additional rotational axis for the torch. Tungsten electrode, an added consumable, is also required. Tungsten electrodes are  vulnerable to contamination, something difficult to detect, and must be maintained by grinding periodically.

Robotic plasma transferred arc welding.

Plasma arc welding (PTA) offers unique advantages like a more consistent current density due to its rigid columnar arc. The plasma arc resists deflection, making joint alignment more repeatable. PTA is unique in that filler metal can be added in a powder form rather than wire. This allows for more variety and blends of alloys unavailable in wire form, and powder is generally less expensive than wire. PTA equipment, however, is usually the most expensive arc welding process. Deposition rates are typically less than GMA.

Submerged arc welding offers some of the highest deposition and lowest defect rates. Despite these benefits, submerged arc welding has historically been the least considered process for automation due to the constant management of flux and slag, and being limited to flat weld positions. However, advances in coordinated motion control and process sequencing have revived submerged arc welding for applications previously thought unviable.

Joint Configurations

Different joint configurations alter the degree of difficulty to automate and must therefore be examined. Fillet and lap joints tolerate misalignment better and have surfaces for adaptive control measures like joint sensing and seam tracking. Fillet and lap joints are also partial penetration joints, and the walls of the joint tend to guide the wire into the joint. Grooved butt joints have similar advantages, but a V butt joint needs a consistent root opening. U grooves without any root opening are preferred for full penetration projects.

Troublesome joints include outside corner joints and square butt joints. Repeatable joint conditions are difficult to maintain on outside corners and tolerances typically are tight. Square butt joints have no identifiable features required by adaptive control measures to locate and track.

Material Thickness

Similar to part size, material thickness can affect a part’s accuracy. Thin parts commonly become less accurate as their flexibility increases and can be easily burned through. Thick parts are frequently inaccurate due to their production techniques, such as flame cutting, or liberal production tolerances.

Base Material

The base material’s weldability is also evaluated before automating. Mild steel has good weldability due to its resistance to cracking and oxidation. High strength / low alloy steels often require a preheat process which adds complexity and equipment cost to the automated system. Aluminum’s weldability hinges on the part’s cleanliness. Stainless steel, high nickel alloys, titanium and other reactive metals require special measures to mitigate oxidation.

Technical Capability of Facility

One of the most neglected areas of evaluation is ensuring that the technical capabilities of a facility match the level of automation selected. The staff must be receptive to automation. Engineers, operators and maintenance personnel must be capable, trainable and willing to support the system once installed. The human-to-machine interface, or controls, needs to be designed to fit the operator’s technical level. Simple is better, but some applications require complex controls and additional programming. Simplifying complex processes or, for instance, controlling complex motions of a robot without actually writing the pages of robot code on the fly, requires wizard-style software and graphical user interface.

Whether using internal resources, or utilizing the expertise of an automation specialist, a successful automated welding project requires leadership with a strong understanding of automation, welding and metallurgy. Additional considerations exist. However, these nine key points can help you guide your customers to identify and address the challenges of a project while it is in the planning stages and ensure a successful project in the end.

Gases and Welding Distributors Association
Dan Allford Meet the Author
Dan Allford is president of ARC Specialties located in Houston, Texas, and at