By Joe Hoffman, senior welding engineer, FANUC Robotics America, Inc.

Vision systems can be used in automated MIG welding for Al.

W elding aluminum with robotic MIG welding equipment presents many challenges. The aluminum MIG process is not as forgiving as steel and requires special control to achieve successful results, and a good understanding of the welding process and how to control it from the robot are critical.

From a robotic perspective, the welding process can be broken down into defined steps.

Starting the arc
Starting the arc can be one of the most difficult steps of the aluminum MIG process. Mechanical properties of aluminum work against the welding process to make an arc more difficult to start. Some of the key factors that affect arc starts are base-metal oxidization, thermal conductivity, and soft or ductile filler wire.

Base-metal oxidization is a natural enemy to the welding process, and measures should be taken to minimize this contamination.

Oxides act as an insulator and require greater arc energy to burn through. Because arc starting routines in robotic applications are predefined at set energy levels, there may not be enough energy to burn through a part that has excessive oxidization, and that can result in a failure at the start. That is why it is important to control base-metal oxidization and to implement measures into the arc starting routine to overcome this natural occurrence.

Touch retract arc starting is one method that is used to overcome the natural oxidization process and assist the starting of the arc.

Touch retract starting is a controlled process in which the robot puts the weld power supply and the wire drive through defined sequences to start the routine, drawing the arc to ignition. The process of drawing the arc eliminates the harsh, dead short, explosive start routines that are conventionally used. This method of starting the arc provides the reliability required for robotic applications without impacting cycle time. Meanwhile, this starting method dramatically improves contact tip life and the mean time between failures.

Weld formation
Weld formation is the next step after the start of the arc.

The objective is to transition from the starting sequence to weld formation, and common techniques include run-in , ramping or direct entry editing.

Choosing the appropriate technique is determined by the part being welded. Material thickness, the requirement for multiple welds on a part, weld sequencing, and fixture design all play a role in this decision.
Run-in typically is treated as a universal technique in which the robot uses a defined weld condition to start all welds for a given process. This is a good technique to use when the base metal temperature is consistent and does not fluctuate during processing.

If the base metal heats up due to weld processing, the run-in conditions used in the beginning phase of the process may not be appropriate to complete the weld. Run-in can be disabled, and an alternative starting condition may be used for other welds.
Ramping is a common technique used to weld thicker material. The theory behind ramping is to change gradually from the starting parameter to the welding parameter over a defined time. During this duration, the weld output shifts from parameter “A” to parameter “B” to provide a smooth transition into the welding mode. Ramping is not a universal technique and can be specific for each weld.
Direct entry is a common technique used on thinner material for which the base metal temperature changes as welds are applied, making it necessary to have specific control at each weld. This technique is different from ramping in that the change between parameter “A” to parameter “B” is immediate. Often on thinner material, the time between the start and weld is so short that there is no advantage to using ramping.

Each of these techniques operate on the same principles.

Touch Retract initiates an arc and a defined set of weld values are used to stabilize the arc, then weld values are changed and the weld is made. When welding aluminum, it is common to use higher weld values to start, stabilize and penetrate, then switch to a cooler parameter to make the weld. Starting slightly hotter helps arc initiation and assists in overcoming the thermal efficiency of aluminum.

Weld Deposition
Weld Deposition is the reward of successful starting and weld stabilization. The robot continues to play an important role and can not be overlooked. The stability of the welding process is directly related to the ability of the robot to control the welding process. Programming techniques such as weaving may need to be applied to overcome variations in the part. Weld process changes may need to be made on-the-fly without interruption of the arc. Advanced process techniques such as “Heat Wave” may need to be used to overcome large gaps, weld variations in metal thickness, or provide the cosmetic “TIG” appearance.

The limitations of the robot should never have an affect on the welding process. Understanding the common aluminum welding modes and how to apply them to robotic applications will assist in achieving success.
Pulse welding is a common deposition mode used in conventional robotic aluminum welding applications. The deposition of this mode is stable, the penetration is consistent and the cosmetic appearance is good. Because of the good stability of the arc, this mode often is used on fillet welds to maintain good travel speeds.
Variable pulse welding is a special deposition mode only supported by a handful of power supply manufactures. The deposition of this mode is stable, the penetration is slightly greater than conventional pulse welding, and due to the nature of the deposition, it tends to tolerate a wider degree of variation over conventional pulse welding. The cosmetic appearance is exceptional and, when properly tuned, resembles that of the TIG stacked-dime analogy.
Power mode is special for aluminum, and provides a clean, fast, spatter-free deposition. It is ideally suited for applications with good material fit-up and has little or no limitations to material thickness. On thicker material the combination of this deposition mode with a circular weave delivers outstanding results. On thinner material this mode can be cranked up and let rip.
Heat wave is a proprietary weld process control unique to Fanuc Robotics. The robot controls the welding deposition by changing the process parameters based on wire location. This advancedprocess control has been instrumental in the evolution of robotic aluminum welding. The sought after TIG appearance can be achieved easily, gaps can be bridged without problems and precise control of penetration simplifies the welding of dissimilar metal thicknesses.

Ending the weld
Arc ending on aluminum requires some special techniques to close the weld crater.

The weld crater is the void that remains at the end of all welds. The amount of current used to make the weld influences crater size.

Failure to fill this void leaves a stress point in the weld that will promote the formation of a defect called a crater crack. A crater crack typically will propagate through the rest of the weld to cause weld failure.

There are several welding techniques that can be used to fill and close aluminum MIG craters. These techniques operate on the same principle, reducing the weld current while adding time to allow the weld puddle to close and the crater to fill.

Personal preference as well as joint design play a role in determining the method that is used, and the end result is most important, the crater gets filled.

The methods to close the weld and fill the crater are analogous to the methods used to start the weld. They are:
Ramping to a cooler parameter. This technique provides a gradual transition from a “hot” welding parameter to a “cooler” parameter. The ramping of the weld schedule alone typically will not fill the crater; some additional time – which is called dwell – must be added to hold the weld process at the cooler settings until the crater is filled.
Process Switching between two modes of deposition. This technique is used to weld with a “variable pulse” process. Variable pulse welding modes do not fill craters as consistently as conventional “pulse” welding modes. Process switching is a good technique to use to switch from a variable pulse process to a straight pulse process that will fill and close the crater. The straight pulse process is predictable and can be programmed to achieve consistent crater results.
Changing weld parameters with an included dwell designed to close and fill craters. This technique is similar to ramping, but does not have the gradual change between the weld parameter and the crater parameter. Changing the weld parameter and including a dwell is called for when welding thin materials. When using this technique, the transition from the “hot” welding mode to the “cooler” mode to close the crater is instantaneous. There is no down ramping. However, as with ramping, a dwell must be added to allow the cooler parameter to close and fill the crater.

Burn Back
Burn back is the final step in making a weld. During this phase, the filler wire is separated from the weld puddle and the arc is extinguished. A system with properly controlled burn back will terminate the wire crisply, leaving no ball on the end of the wire. Systems with poor burn back control end the wire erratically, often leaving a large melted ball of wire on the end of the wire or the wire is consumed into the contact tip. A clean ending properly prepares the wire for the next start.

The manufacturer of the weld power supply dictates where the burn back control resides. Control can be located within the weld power source or remotely from the robot. It is important to understand where burn back control resides to avoid communication conflicts.

If point of control resides at the weld power source, the robot needs to be configured accordingly. The burn back feature on the robot should be disabled, and additional communications to synchronize the motion of the robot with the shut down routines of the power source may be necessary. Failure to synchronize the motion of the robot with the power supply shut down routines often will result in poor ending conditions and failed arc starts.

Burn back and the influence it has on weld wire is very important and often is misunderstood. Improper settings could result in a failed arc start for the next weld. The poor ending condition of the previous weld creates a failure condition for the next weld and adds confusion to the troubleshooting process.

When trouble shooting a failed arc start condition always look at the ending conditions of the previous weld. Understanding where as well as how to adjust the burn back to get the desired ending results can minimize process problems.

Burn Back Rules The weld wire should be separated from the weld puddle. The weld wire should extend past the contact tip half the distance of the taught tool center point upon weld termination. For example:If the taught tool center point (TCP) is 12 mm, then the wire stick-out after burn back should be 6 mm or more. The weld wire should not have a large ball formation on the end of the wire at the start or end of a weld. The wire should look as if it was cut. Understanding the welding process and the capabilities of your equipment are the keys to success. If you have inadequate equipment, an upgrade to the latest technology may be necessary. If you do not thoroughly understand the welding process, or how to program the robot to give you the desired results, pursue training. With proper equipment and thorough understanding of the welding process, robotic welding of aluminum can be successful. As the process gains acceptance, unique and more difficult challenges are presented. Understanding the challenges and the ability to develop the necessary tools to succeed should be the goal of your robotic supplier.