When most engineers think of resistance welding, they picture six-axis robots welding car bodies on a moving assembly line with sparks flying everywhere.
But, says Kevin J. Ely, Ph.D., manager of microjoining and plastics at the Edison Welding Institute (Columbus, OH), if you're welding parts that are less than 0.5 millimeter thick, the last thing you want to see is sparks. "In small-scale resistance welding, you never want to see expulsion," says Ely. "If you're welding car bodies, you can afford to lose a bit of metal. In small-scale resistance welding, if you have expulsion, you might lose 20 percent of your material."
Resistance welding is used to assemble many small products, such as medical devices, automotive ignition modules, sensors, heating elements, microwave conductors, smoke detectors, solar cells, batteries, relays, lightbulbs and air bag detonators. The process can join similar or dissimilar metals, including aluminum, brass, copper, gold, Inconel, molybdenum, nickel, Nichrome, nitinol, platinum, silver, stainless steel, titanium and tungsten.
In resistance welding, the parts to be joined are pushed together while an electrical current is passed through them for a brief time. The heat for welding comes from resistance to the current both within and between the parts and electrodes. When the current stops, the electrodes continue to hold the parts together, and the molten metal rapidly solidifies, forming a weld. The challenge with using this process to join small, thin parts lies in precisely controlling time, current and force to get a good weld without overheating the assembly.
"Most of the blame for poor quality welds is put on the weld pulse, so engineers try to solve the problem by increasing the current or buying a fancier power supply," says Ely. "However, force, squeeze time and hold time are just as important."
Overcoming Resistance
Resistance is one variable that engineers can manipulate to get better welds.
Bulk resistance-the resistance to current within the parts and electrodes-increases as their temperature increases. As a result, bulk resistance can be a factor with high-volume operations or with weld times over 50 milliseconds. "The only way to change bulk resistance is to change the materials," says Ely.
Contact resistance-the resistance to current between the parts and electrodes-decreases as the parts are forced together, and it drops to zero within a few milliseconds after current is applied. Contact resistance lasts longer with hard, resistive metals, like stainless steel, than with soft, conductive metals, like copper. Contact resistance is also influenced by the condition of the parts' surfaces, such as roughness, platings, coatings, oxidation, dirt and oil.
"Platings and coatings are rarely chosen with consideration for welding," Ely points out. "To produce a satisfactory weld, the process designer should understand the metallurgy of the coating as well as that of the parent materials."
Naturally occurring oxides on stainless steel, copper and brass are usually thin enough to weld through. Oxides on titanium and anodized aluminum should be removed before welding.
"Clean parts weld better than contaminated parts," says Doug Stokes, technical support manager for MacGregor Welding Systems Ltd. (San Diego). "So you need to think about how the parts are handled and stored."
Platings can be problematic because they often melt at lower temperatures than the base metals. For example, nickel-zinc plating is used to prevent corrosion on steel. However, this plating can be difficult to weld, because zinc tends to vaporize, leaving pores in the weld zone, says Ely. As a result, engineers should select a plating that has the same thermal conductivity as the base metal. Gold plating on a copper part will not affect the welding process, but nickel plating will cause weak welds because the nickel cannot dissipate heat as quickly as copper.
Chemical reactions are also a problem with platings. Tin-based platings are often applied to copper, brass and steel to prevent corrosion. If the electrode will contact the tin-plated side of the part, the electrode should be pure tungsten. Tin reacts with molybdenum and reduces electrode life.
Another problem with platings is inconsistent thickness. If the thickness of the plating on a 0.005-inch thick part varies by as little as 5 microns, that can be enough to cause inconsistent welds.
Parts with uneven surfaces can cause random changes in contact resistance and thus, variation in weld quality. As a rule of thumb, peak-to-peak variations in surface finish should not exceed 10 percent of the part's diameter or thickness. For example, when welding a wire that is 12.5 microns in diameter, the peak-to-peak variation in surface finish on the mating part should not exceed 1.2 microns.
One way around the problem of inconsistent surface finishes is to increase the force applied to the parts or lengthen the time that force is applied. Engineers may be tempted to reduce squeeze time to decrease overall cycle time, but they should not reduce it too much. "Squeeze time allows contact resistance to stabilize and reduces weld-to-weld variation," says Ely.
Another way to solve the problem is to preheat the parts by applying a small pulse of current just before the primary pulse. Alternatively, the weld current can be gradually increased over time. This also helps prevent current spikes.
The most accurate way to measure contact and bulk resistance is with a constant-current power supply and a voltmeter, says Ely. This four-probe technique is preferable to the two-probe ohmmeter, which can be affected by contact resistance between the probe and the parts.
Striking a Balance
The goal in resistance welding is to generate the required amount of heat at the joint interface without excessively heating the surrounding material. The weld nugget should be centered at the interface with equal amounts of fused material on both sides. However, achieving this goal can be challenging when the parts to be welded have different thicknesses and conductivity.
Engineers can balance heating by varying the shape and composition of the electrodes, says Ely. For example, when welding a high-conductivity alloy to a low-conductivity alloy of equal thickness, an electrode with a small face is used on the high-conductivity alloy, while an electrode with a large face is used on the low-conductivity alloy. The smaller contact area will increase the current density in the high-conductivity alloy. Less heat will be conducted away from the joint by the base metal and the electrode. More heat will be generated in the workpiece, and the fusion area will shift from the low-conductivity alloy to the high-conductivity alloy.
Alternatively, the high-conductivity alloy could be paired with a high-resistance electrode, which will draw less heat away from the joint. Or, the thickness of the high-conductivity alloy could be increased.
"When choosing an electrode, use the rule of opposites," says Stokes. "If the metal you're trying to weld is fairly resistive, like steel, you would choose an electrode made from a copper alloy, which is very conductive. If the metal you're trying to weld is fairly conductive, like brass, you would choose an electrode made from molybdenum or tungsten, which is very resistive."
A similar strategy can be used to weld parts with similar conductivity, but differing thicknesses, says Ely. Because the thick part will have more resistance than the thin one, it will heat more. As a result, the weld nugget will penetrate more of the thick part than the thin part. This imbalance can be corrected by decreasing the current density in the thick part or decreasing the heat loss from the thin part. For example, applying a large-diameter electrode on the thick sheet will concentrate the current density in the thin part.
If one part is more than five times the size of the other, a projection can be formed in the part with the most mass or the part that is the most conductive, says Ely. The projection constricts current flow and becomes a point of high resistance. As a result, heating occurs preferentially at the point of contact. As the material heats, it becomes soft, and the projection collapses under the force applied by the electrodes.
"Using a projection in resistance welding significantly increases the robustness of the process. It also extends the life of the electrodes," says Ely. "The main disadvantage of projection welding is the need to stamp or machine a projection in the part, which adds an extra manufacturing step. Also, it's very important that the weld head allows the electrode to follow the collapse of the projection. If it can't, splashing and arcing could occur at the interface."
Stokes warns engineers that parts with radically different thicknesses cannot be welded. "You can't weld a piece of foil that is 0.001 inch thick to a piece of steel that is 0.01 inch thick," he explains. "By the time you get the large part heated, you've vaporized the small part."
Large differences in thermal conductivity between parts can be overcome by altering their shape. For example, welding a nickel wire to a brass terminal could be problematic, because the brass absorbs heat more quickly than the nickel. Cutting a hole in the terminal will reduce its mass and thus, reduce its ability to drain heat from the welding area.
