Since its founding in 1919, the American Welding Society (AWS) has played a vital role in supporting the needs of the welding industry. The nonprofit group has provided technical resources, certifications, standards development and educational programs all while advancing the professional growth and development of individuals interested in the trade.
To lift up the trade and those interested in it, AWS also organizes conferences, seminars and training programs to facilitate knowledge sharing and professional development within the welding community. A good example of those efforts was seen in the “Women in Welding Virtual Conference 2023: Automation in Manufacturing,” a webinar hosted by AWS that featured several women welders and welding engineers.
Each presenter focused on the automation technologies for which they’ve become experts, culminating with Karen Gilgenbach, who, in a short amount of time, offered a nearly full compendium of considerations to keep in mind when implementing a robotic welding cell.
Gilgenbach is zone vice president at Matheson, a major industrial and specialty gas supplier, and is the chair of the AWS D16 Committee on Robotic and Automatic Welding. She has a bachelor’s degree of science in engineering mechanics from Michigan State, a master’s degree of science in welding engineering from Ohio State, two business degrees from Indiana University and three AWS certifications, including certified welding inspector, certified welding supervisor and certified robotic arc welding technician.
Relying on her extensive background in welding and welding automation, Gilgenbach offered valuable advice to webinar attendees considering the adoption of robotic welding technologies. She focused on how to identify automation opportunities through part selection, welding process, joint design, programming, repeatability, upstream processes, fixturing and more. Most importantly perhaps, she kicked things off with the safety considerations that must be made to truly ensure implementation success.
Safe and sound
At the beginning of her presentation, Gilgenbach stressed the importance of safety when implementing any new welding process or equipment. To ensure a safe environment, she said that automation adopters must first and foremost understand the risk assessments and safety protocols that should be provided or recommended prior to the delivery of the robotic cell or cobot.
“Typical preventative measures prohibit an operator from coming into contact with a full-speed robot, or in other words, a robot other than a cobot,” she explained. “I have seen robots that don’t have the proper safety protocols where somebody could get hurt, and unfortunately, if you don’t get those safety mechanisms at the time you get the robot, they can be very tricky to integrate later.”
One of the main attractions to cobots is their ability to work in proximity to humans due to enhanced safety features. Gilgenbach clarified that the safety protocols being discussed were primarily for standard, industrial robots and not necessarily cobots.
Those include light curtains, steel walls and other machine guarding, which add to the cost and footprint of the welding cell and should, therefore, be considered during the early planning phases of any automation implementation. There are also common welding risks that should be considered that aren’t exclusive to industrial robots, including the arc and fumes that may require fume extraction systems and other welding PPE.
Taking safety protocols seriously means keeping employees safe, but it also means staying in compliance with regulatory legislators and governmental bodies, such as OSHA. During her presentation, Gilgenbach hinted at the idea that the regulatory world isn’t necessarily easy to navigate. Case in point: OSHA, the Robotic Industries Association (RIA), the AWS D16 Committee and ISO all have different safety standards. But don’t fret:
“Instead of only considering the codes written specifically for robotic arc welding, a robotic safety expert on the D16 Committee provided a lot of insight on the importance of OSHA standards that apply to machinery and machine guarding, environmental controls and electrical,” Gilgenbach said. “Even though there are no OSHA standards that are specific to robotic arc welding on OSHA’s web page, that doesn’t mean that other OSHA standards aren’t applicable to robotics and your robotic installation. An example is a standard that covers machinery and machine guarding, which may include steel walls and laser light curtains.”
Gilgenbach added that the OSHA web page does serve as a good resource; in addition to applicable OSHA standards, it also references a variety of documents, including ANSI/RIA 15.06. She also recommended the ISO 10218 document, which includes two sections for robots in industrial environments, and the ISO 15066 document on cobots. And of course, as chairperson of the D16 committee she encourages robot uses to look at AWS D16.1M/D16.1 and AWS D16.3M/D16.3.
Despite any elusive regulatory guidelines, installing proper safety protocols should be fairly straightforward and something that the manufacturer of the robot or robotic cell can help with. If the robotic manufacturer or integrator can’t offer safety advice, it might be worthwhile to look for a different, more credible provider.
The reasons for considering a robotic welding cell or a cobot are broad, but it’s often seen as a solution for labor, quality and throughput issues. To determine whether a part might be a good fit for robotic welding, Gilgenbach spent the bulk of her presentation on considerations around part size, part presentation and fixturing, welding process and welding cell design.
Part size: To prepare for robotic operations, consider whether a part is small enough to fit on a pre-engineered cell. Big parts can be welded with robots, of course, but the bigger the part, the more complex the robot, the more the cost goes up, and therefore, the harder it can be to justify the investment, depending upon the application. For the first robot for a shop, she recommended taking a look at common subcomponents that may exist.
“Don’t rule out big, complex parts, though,” Gilgenbach said. “Perhaps it’s a part that’s regularly made, which is another consideration. Or, if it’s a part with a lot of arc-on time, which may help with the cost justification.”
Part presentation and fixturing: There are nearly unending options as to how a part can be presented for welding, but fortunately, Gilgenbach offered several rules of thumb and advice. For starters, if all of the welds can be accessed from one side, the overall application cost may be less (if positioners are not required) and, therefore, might be easier to justify. To further justify a robotic installation, she encouraged attendees to think outside of the box.
As a few examples: If the time to load a fixture and robotically weld the part is the same, an indexing table can double (or more) a company’s output. Consider different positioner options for robots, like a headstock and tailstock, J hooks or sky hooks as well as tooling that’s designed so that when the part is flipped over the robot can access all of the welds. Or, consider using two robotic arms instead of one to decrease the cycle time and to reach all areas of very big parts.
Tooling and fixturing can be an expensive component of an overall automation implementation, so Gilgenbach cautioned attendees to not skip over that part of the plan when budgeting. “Tooling can be as complex as pneumatic fixturing where all the clamps close automatically,” she said. “You can also have sensors in the fixture that can alert the operator if one of the components wasn’t loaded properly. There is also modular tooling that can be used for high-mix, low-volume manufacturing instead of investing in multiple specialized fixtures.”
The current use of fixtures can also be an indicator of the part’s repeatability and suitability for automation. In regard to pre-existing fixtures for parts under consideration for automated operations, Gilgenbach said, “If it takes a hammer to put the part in the current fixture, it’s probably not a good robotic application.”
Welding process: Before considering a part for robotic welding, Gilgenbach said it may need a certain level of repeatability and weld accessibility. “If only the best welder in the shop can make the part, it might not be a good candidate. It is important to understand exactly what the challenges are with the part when a welder welds it.”
Tolerances should be within one half of the size of the wire diameter. As an example, for 0.035-in. wire, tolerances should be within 0.017 in. Another good rule of thumb is to keep the size of any gaps within one-half of the thickness of the material.
“In regard to joint design considerations, certain joints are going to be more forgiving for the robot,” she said. “Slot welds are very common for robotic applications and can be really nice. Joints that have an outside corner should typically be avoided due to burnthrough.”
Another area to consider is whether there are multiple steps in the welding process. If so, Gilgenbach said that will add complexity, which is okay, but something to think through. “Multiple steps may lend themselves to progressive fixturing, such as having multiple fixtures on one side of an indexing table,” she said. “People can get really creative.”
Welding cell design: Considering the vast options that integrators and robot manufacturers offer in how an automation cell is designed and customized, Gilgenbach mentioned a few rules of thumb for that, too. She reminded webinar attendees that cobots have much smaller footprints because of the lack of guarding required. And, she said to be sure to consider whether the robot can continue to weld while the next part is being loaded – that way, both the robot and the human operator are busy at the same time.
Plan for programming
When gearing up for an automation implementation, planning for programming is a must. Who’s going to program it? When are they going to program it? And where are they going to program it, as in at the robot or offline?
“If you’re going to program a weld at the robot, it’s going to take some time away from the robot actually welding,” Gilgenbach said, “which can be a factor in terms of cost justification and overall productivity.”
Offline programming allows the robot to continue working versus sitting idle while programming or troubleshooting the next job. It can also be incredibly helpful when determining the design of the welding cell and the type of fixtures to use.
“Offline programming software is a great way to make sure that the robot can reach all the areas that it needs to reach,” she explained. “In doing so, you can determine whether your tooling will work. It’s important to think through whether the robot can reach all the joints you want welded and in what position they will be welded – vertical down, horizontal, flat, etc.”
She also mentioned that in high-mix, low-volume environments “every job that you have to program might take a couple parts to get the welds just the way you want them, so it may take some time to program no matter what kind of robot you’re using. That is part of the reason the convention has been to pick parts where you have a higher volume for automation, with the conventional exception being very large parts with a lot of arc-on time.”
At the end of the day, though, Gilgenbach said the efficiencies in robotics can still sometimes justify short runs. It will all come down to proper planning.
Finally, don’t forget to plan for the order in which the welds will be programmed. “When you’re looking at putting a part on a robot, I always suggest starting the same way the welder has been welding the part,” she said. “The way he or she does it might not look intuitive, but the welder has probably trouble-shot through a lot of issues over time and might already be accommodating for distortion and spatter, for example.”