Improving Resistance Welding Process Control in Medical Applications

Medical Device & Diagnostic Industry Magazine
MDDI Article Index

An MD&DI November 1997 Column

PROCESS CONTROL

New systems enhance manufacturers' control over key variables during the welding process.

During the past two decades, resistance welding has become a widespread method for joining metals in medical manufacturing applications such as biosensors, catheters, pacemakers, surgical instruments, and electronics. As a result, the resistance welding equipment industry has developed increasingly sophisticated systems for controlling heating profiles, welding force, and the duration and intensity of weld current­all critical factors for achieving consistent results.

Welding of the filament, foil, and post is a key process in the making of x-ray and other medical lamps. Photo courtesy of Unitek Miyachi

New resistance welding systems using constant feedback loops to control energy, time, and force open the door for improving overall process control programs. Built-in process control and higher levels of automation not only improve product quality but also increase profitability by shortening the amount of time required to set up, test, and certify new assemblies and processes. The repeatability of built-in process control mechanisms also allows manufacturers to easily replicate existing joining processes at a number of geographically dispersed production facilities.

THE FUNDAMENTALS OF RESISTANCE WELDING

Resistance welding is a thermoelectric process in which heat is generated by passing an electric current through the parts to be joined. The process uses electrodes to push the parts together with sufficient force to remove the surface oxides that would otherwise inhibit the current flow and to establish the initial contact resistance. Once physical contact is established, a carefully controlled combination of resistance-generated heat and physical pressure is applied through the electrodes to create the actual joint. The amount of heat generated is a function of the amount of current applied, the length of time it is applied, and the resistance profile between the parts being joined.

A typical resistance welding heat profile shows that temperatures rise highest at the point between the parts to be joined. Temperatures also rise between the electrodes and the parts. The objective of resistance welding process control is to consistently maximize a short-duration temperature rise between the parts while minimizing the secondary temperature effects on surrounding materials.

Three basic types of bonds can be created using resistance welding techniques:

Brazed/soldered bonds. Resistance heating of the parts to be joined generates sufficient heat to melt a third metal, such as silver solder alloy or tin/lead solder, which then bonds to both parts.

Forge welds. A very short weld-time current is used to forge the parts together without melting them. This is useful for joining two dissimilar materials such as molybdenum and tungsten.

Fusion welds. A longer pulse is used to melt both parts to a liquid at their interface point. The pieces are held together as they cool, and the recombination of the materials forms a nugget that contains all of the possible alloys to the two materials. Fusion welding is ideal for joining two similar metallic materials.

WELD VARIABLES

The key variables that must be controlled to achieve consistently acceptable welds are:

Three of the four variables that affect resistance welding­energy, time, and force­can be enhanced by using the appropriate systems and finding the best force/energy profiles for the materials being joined. The fourth, materials, should be analyzed with respect to the first three by using design of experiments (DOE) techniques to enhance process control.

Improvements to resistance welding equipment usually focus on controlling the first and second variables, energy and time. Modern resistance welding controls, with dynamic feedback, have substantially improved control over these two variables, permitting attention to be focused on improving the third variable, weld force. The newest generation of resistance welding equipment features feedback loops built into the force control mechanisms, thus removing operator-introduced variations.

The fourth variable, materials, is entirely controlled by the manufacturer. Medical manufacturers should work closely with their resistance welding equipment supplier to tailor the system's features, electrodes, and process settings to the materials being joined. They must also understand and control the metallurgic and resistance characteristics of the workpiece materials.

CONTROLLING ENERGY AND TIME

Effective resistance welding requires applying constant weld energy to the parts for a precisely controlled interval. There are three microjoining resistance welding systems used for medical applications: direct energy (alternating current), stored energy (capacitor discharge), and high-frequency direct-current (HFDC) inverter systems.

Direct Energy Systems. These systems have been around for a long time and are relatively simple. They use available alternating-current power to deliver weld current. Direct energy systems function much like a water faucet: The speed and duration with which the valve is opened and closed creates the desired weld pulse profile. For instance, slowly opening the valve creates an up-slope of weld current to preheat hard, curved, or aluminum parts; slowly closing the valve creates a down-slope of current to anneal the parts after the weld. Sequentially rising spikes of current can be used to temper the parts or remove coatings prior to the final weld pulse.

The total heating capability of a direct energy system depends on its transformer size and is directly proportional to the amount of alternating current that is drawn from the power line. Weld current pulse time control is synchronized with the underlying alternating current line voltage cycles (e.g., 60 cycles in the United States and 50 cycles in Europe and Asia). This means that pulse profiles for the same application may differ between the United States and other countries­something to take into account when conducting geographically dispersed manufacturing operations. In addition, direct energy systems can be sensitive to line voltage changes or inconsistent power sources.

Stored Energy Systems. These systems use a capacitance discharge to deliver current. Resembling a bucket of stored water more than a spigot connected to a constantly available source, the systems extract energy from a power line and store it in the weld-energy capacitor banks. Thus the heat generated by a stored energy system is directly related to the size of its capacitor banks and welding transformer, rather than to the level of available alternating current.

Because a stored energy system does not depend on the immediate flow of alternating current, it is quite flexible. For example, higher levels of pulse energy can be attained by "dumping the bucket" all at once than would have been available using line current directly. In addition, most stored energy systems are designed to provide rapid sequential pulses of current. This dual-pulse capability is particularly useful for coated or plated materials; the first pulse displaces the surface oxides and the second pulse welds the underlying materials. Dual pulses can preheat or postheat the workpieces and control overall temperature rise/fall profiles to prevent material expulsion and cracking. Moreover, stored energy systems can reverse the polarity of the sequential pulses, which is useful for welding dissimilar or polarity-sensitive parts.

The greatest limitation of stored energy systems is the relatively long time it takes to charge their capacitor banks. This can make them unsuitable for high-volume production situations requiring a constant series of rapid welding actions.

High-Frequency Direct-Current Inverter Systems. These systems use sophisticated power-switching technologies to provide weld energy that is controlled with constant current, voltage, or power feedback loops. The switching technology minimizes internal power losses in the control circuits, provides immunity from line voltage fluctuations, and delivers high weld energy through relatively small, lightweight transformers. The feedback loop in new-generation HFDC inverter systems is updated at extremely rapid rates, as often as 0.00004 seconds per update cycle. This high-frequency switching technique permits precise control of weld current during weld cycles as short as 2 to 4 milliseconds.

HFDC inverter systems combine the best aspects of direct energy and stored energy systems. They allow continuous rapid welding cycles because the weld energy is immediately available (there are no capacitors to recharge), but they do not depend on line current timing cycles or quality.

CONTROLLING FORCE

Precisely managing the amount of force applied during welding is a key but often overlooked aspect of resistance welding process control. The amount of force applied to the parts through the electrodes during each phase of the welding process (e.g., weld force buildup, heating, rapid melting, and cooling phases) affects the overall integrity of the bond. For example, too little weld force can produce insufficient contact between the parts to break through their oxide layers, while too much force during the weld can reduce interface resistance between the parts and actually degrade the heating profile.

Statistical Process Control Techniques. Manufacturing process engineers should routinely develop a force/energy profile map for each application by correlating various force levels and energy levels with weld quality measures such as pull strength, determined by destructive testing. These data create a real-world set of parameters that can be used with statistical process control (SPC) techniques to monitor and control the ongoing production process.

For example, Figure 1 shows a typical DOE profile map, comparing four different electrode force levels (8, 10, 12, and 14 lb) by correlating the relationship between various levels of weld energy and resultant pull strengths for each force level. For each curve, the weld-heat setting in wattseconds (Ws) has been systematically varied (on the x-axis) while the weld force has been kept constant. The empirical data on pull strength has then been plotted (on the y-axis) as a key quality outcome. The acceptable range of pull strength for this particular application has been specified as 12­12.5 lb. Each curve presents a significantly different profile regarding the amount of pull strength that can be achieved and the range of energy levels over which that pull strength can be maintained.

Figure 1. Weld schedule diagram­pull strength versus weld energy.

The primary objective of a DOE analysis is to define a stable process that can be controlled within a reasonably wide parameter window. In this case, this means that the curve must not only reach the required pull strength level but also maintain that pull strength level across a range of energy, time, and force levels. DOE process optimization can be compared to balancing a plate on the whole hand versus one finger. A barely controlled process has only a few points where the key parameters intersect to produce acceptable results, whereas a well-controlled process has a broad range of combinations that yield satisfactory outcomes.

Thus, in Figure1, the 8-lb and the 14-lb curves are unacceptable because they do not produce the required 12-lb pull strength at any weld current energy level. The 12-lb curve yields acceptable pull strength between energy levels of approximately 35 and 62 Ws but experiences a 2-lb pull strength variation range that may be unacceptable from an SPC standpoint. The optimal force profile is provided by the 10-lb curve, which has a relatively flat peak that produces a pull strength variation of less than 1 lb across a wide range of energy levels (from 45 to 70 Ws).

If a resistance welding system cannot control the weld force at a consistent level, it will create bonds that vary randomly across the spectrum of acceptable and unacceptable parameters. Therefore, the newest innovations in resistance welding systems are more sophisticated built-in force-control mechanisms.

Linear-Magnetic Actuators. Today's most advanced systems include linear-magnetic actuators that can be electronically controlled to manage force levels throughout the welding cycle. For instance, in microjoining operations, the rapid expansion or deformation of the extremely small workpieces during the heating phase can cause significant variations in weld quality during the weld, producing sparking, splashing, and inconsistent weld results. In such cases, applying a controlled amount of additional follow-up force during the rapid melting phase can make a significant difference in maintaining control over the final bond. As shown in Figure 2, the position of the electrode against the top workpiece being joined may need to vary as much as 0.2 mm during the course of the weld cycle.

Figure 2. Electrode position on top part surface during melting.

At the beginning of the weld force buildup phase, the electrodes are brought into contact with the weld materials and the prescribed weld force is established. When the resistance weld current is turned on and the weld materials start to soften, their reaction force against the electrodes decreases, calling for a follow-up movement by the electrode to maintain the prescribed force. During follow-up, the electrode must accelerate rapidly to maintain force on the weld material and then stop precisely when the reaction force builds up again. This uniform force must also be maintained during the cooling phase to ensure the final integrity of the fusion nugget.

Table I. Comparison of force control methodologies.

Table I compares new-generation linear-actuated control systems to traditional coil-spring and air-drive systems. Because many medical applications now involve the microjoining of ever-smaller components, maintaining consistent follow-up force has become even more important in reducing splash and preventing deformed bonds. Since the rapid melting phase can take as little as 0.5 millisecond, the electrode must be able to move very rapidly to maintain constant force. Traditional coil-spring systems lack the precision for such minute force adjustments, and air-actuated systems generally cannot provide enough speed to track over such short melting times. This inability to match the force with the rapidly changing temperature rise times can cause weld splash and cracking, leading to weld inconsistencies and parts damage. These problems are avoided by electromagnetic linear actuator weld-head control systems (Figure 3).

Figure 3. Rise time, overshoot, and ringing versus time in linear-magnetic actuators versus coil-spring and air-drive systems.

Magnetic weld force control systems definitely improve process consistency but at a price. A magnetic weld force control system can cost two to three times more than a standard coil-spring weld head. In addition, it may be difficult to synchronize the weld force profile with the weld energy profile without using more complex and expensive welding controls, such as an HFDC inverter system. Finally, magnetic weld heads are limited to about a 50- to 100-lb weld force because of the high current necessary to produce and control the weld force profile.

CONCLUSION

Driven by the globally competitive market environment and a myriad of regulatory and quality oversight regulations, medical manufacturers must continuously improve their control over every aspect of production. However, older resistance welding systems often relied on operator skill levels and on-the-floor experimentation to optimize results.

Today's new generation of resistance welding equipment takes a major part of the process control equation and incorporates it directly into the systems. The bottom line is that manufacturers can now use much more adaptable welding systems that extend the users' capital investment by bringing new applications on-line faster and tightly controlling the key quality factors throughout the welding cycle. Although such systems cannot replace the manufacturers' own development of process control techniques, they can enhance those techniques by greatly reducing the potential for error.

David Steinmeier is the business development manager for Unitek Miyachi Corp. (Monrovia, CA). He has 17 years' experience in microjoining with resistance welding.

Copyright ©1997 Medical Device & Diagnostic Industry