Friday, September 23, 2011

A Primer to Spin Welding

By Jerry Wibben, Regional Sales Manager, Dukane Corporation

Spin welding is a simple process that has been around (pun intended) longer than thermoplastics. Spin welding of metals has been known and practiced for at least a hundred years. It is no surprise, then, that it is one of the oldest methods of joining thermoplastic parts. It is a fast way to join parts that have a circular joint, and is very reliable at delivering a hermetic seal in those products that require one.

The fundamental idea is to spin one part against another under clamp force, the surface friction creating heat that melts the interface, and then to stop rotation and allow the parts to fuse together. It is a process that is deceptively difficult while in truth, as with many materials, it really is quite simple.

The Phases of Spin Welding

The phases of a spin weld are the approach, weld, hold, and retract. There are several ways to begin the rotation. If the part is to be rotated by frictional contact with the spin tool (as opposed to engaging a detail such as drive dogs or hose barbs), the tool can be spinning before contact. Alternatively, the tool can approach not spinning, pick the part up either with friction or vacuum, retract slightly, start spinning, and then extend again; or the tool can apply pressure prior to spinning. Of course, the part also can be directly loaded into the spin tool and held in place by friction or vacuum.

During the spinning portion of the cycle, the first thing that happens is some small amount of heat build-up that softens but does not melt the surface. When this happens, material will be stripped from the surface and rolled up into little balls. This is why spin welding produces particulate. As spinning continues, heat continues to build. Once sufficient heat has built up, true melting of material will occur. At this point, bond line thickness is established; in other words, additional spinning will not add to bond line thickness or strength of the part. Additional spinning, however, will cause the parts to travel toward each other with excess melted material thrown out of the joint in the form of flash and particulate. Concealing this flash and particulate is an important part of designing the joint.

Once true melting is established, rotation can be stopped. It is important to maintain clamp force on the parts and to stop the rotation as abruptly as possible. If spin down occurs in a gradual fashion, the material can start to solidify before rotation stops and the joint interface can be severely weakened by shear forces applied during this cooling period.

Virtually all thermoplastic material can be spin welded if they have a high enough ratio of the coefficient of friction to thermal conductivity. To put is more simply, heat needs to be created faster than it can dissipate. Usually, only materials with very high lubricity are excluded from consideration. Care should be taken to ensure that materials on both sides of the joint are not only chemically compatible, but have similar melt temperatures and similar melt flow indices. If materials are reinforced (i.e. with glass), there will be no benefit of reinforcement in the joint itself; the spin welding process does not promote any fibers crossing the joint line, rather it encourages fibers to lay parallel to it.

The joint itself needs to be a circle, but the remaining part geometry is fair game, as long as it does not interfere with rotation or create a severely out-of-balance condition for the rotating part. This is particularly a concern for small parts, as the rotation speeds may need to be high.

Rotation speed is based on the surface speed at the joint, so smaller parts turn at a higher rate than larger parts. A rule of thumb is that the target rotation speed in rotations per minute should be about 8,000 divided by the joint diameter in inches, or 200,000 divided by the joint diameter in millimeters. This is a ballpark speed, with successful applications running at speed deviating from this rule by 50 percent or more both high and low.

Three Main Classes of Spin Welders

Inertial spin welders typically use an air motor to spin up a flywheel, which can be clutched in, but is more commonly attached to the tool in a fixed manner and engages the part frictionally. The part acts as a brake for the flywheel; so the kinetic energy stored (mass of the flywheel multiplied by the rotational speed) is the energy imparted to the joint. Because of the high speeds air motors can generate, these machines work well for very small parts, and they are simple and reliable. They can, however, be hard to adjust and hard to convert from one job to another. These machines also cannot control radial orientation of the finished part.

Conventional electric direct drive or geared spin welders use AC induction motors to drive the spin tool. They generally use digital motor controllers so rotational speed can be adjusted. Many also have electronic dynamic braking. In situations where dynamic braking is not sufficient, a physical brake may be installed. These machines can be built to any speed or torque requirement, but high torque motors tend to have heavy armatures so they do not stop quickly if turned at high speeds. Induction motors also delivery their highest torque only at high speed; so gearing is often used n high-torque applications (large parts, low rotation speeds) to keep the motor shaft speed up in the high torque part of the power curve. While certain of these machines do have sophisticated controls, many are quite simple. These systems also are not capable of controlling the radial orientation of the finished part.

Servo driven spin welders, whether geared or direct drive, typically can control the finished part radial orientation to within one degree or better – a specification that usually depends more on the tooling than the machine. Servomotors have relatively flat torque curves all the way from a near-standstill to maximum rated speed, so they typically are more versatile than the other types of machines. A servomotor also is typically more compact for a given torque rating than an induction motor, and is much better at staying on speed under load. Machines using servomotors usually have more sophisticated controls and a variety of welding methods. Servomotors also can deliver feedback to the control, so torque ad energy can be monitored during the process. Gearing, if used, is almost exclusively intended to multiply torque rather than simply to change the output speed. These systems are typically very precise and repeatable productions tools.

Spin welding is a simple process that, when coupled with modern servo motor technology, delivers precision and repeatability beyond that dreamed of when it was first tried during the open-cockpit days of the plastics industry. The process is fast and stable, and will continue to be utilized for as long as making products out of thermoplastics maintains its popularity.