In the friction welding process, one component is commonly held in a self-centering clamp.  In this example, the other component is held in a centering chuck mounted on a rotatable spindle, and driven by an electric motor (either AC or DC).

To make the weld, the chuck is rotated, and the components are thrust together to produce frictional heat at the faying surfaces.  When the combination of speed and thrust has achieved suitable conditions in terms of heat and temperature at the flying surfaces, rotation is stopped and the thrust is maintained or increased to create the weld.

The most commonly used friction welding machine joins two components together, although three or more components can be simultaneously joined on a specially designed machine.  In a basic form, the machine spindle can be driven directly from an AC or DC motor source and then allowed to stop under its natural deceleration characteristics, and the retarding torque developed from the components being formed.   In practice, a clutch is normally added between the motor and spindle so that the motor can run continuously, and the main spindle clutched in and out only when required for friction welding.  This reduces the energy that would be consumed by the motor if it was started from zero speed for each operation and drew full starting current for each weld cycle (typically 60 to 300 times per hour, depending on component size). Some modern machines include a fast acting brake to the spindle.  The function of the brake is to provide the operator with a means for holding and maintaining a close overall length balance on the parts being welded.

The spindle can be rotated under thrust for as short or as long a period of time as demanded by each individual set of components, depending upon their pre-weld length tolerance.  This allows for the discreet differences between components, i.e. saw cut ends, forging surfaces, etc., to be consumed during the welding process before the real burn-off phase is carried out.

Alternatively, the operator can elect to rotate the spindle under thrust for a pre-set time, thereby reproducing a finite equal amount of energy for each set of components to be joined.  It is usually necessary under these conditions to pre-machine the surfaces to be joined to obtain a good overall length tolerance after welding.

 Generally, the former method is selected when friction welding components for high duty applications, because it is necessary to ensure that at least a minimum loss of length occurs between the components to be joined.  Measurement of length loss is a more direct measurement of the welding cycle, compared with measurement of time

There are two phases to the process - heating and forging.  The whole basis of the process is to heat materials to their forging temperature, and then forge them together to produce a forged solid-phase bond.  During the heating phase, the relative speed of component surfaces and pressure on the surfaces produce heat.   One set speed and one set pressure are adequate for the majority of materials.   Some materials however, e.g., those normally requiring a pre-heat before being joined by other welding processes, can be dealt with by having a two-stage pressure application during the heating phase.  A first low pressure is applied, which generates heat into the components with no length loss.  This is followed by the normal frictioning pressure, and will produce a desirable condition in any medium carbon steel, such that when forging is complete, the pre-heat applied earlier will produce a slower quench rate of the heat affected zone.  This will produce a more ductile joint, and overcome the likelihood of cooling cracks.

A typical friction welding machine consists of the following:

    1. Friction welding machine head
    2. Machine base
    3. Component clasping arrangement and backstop
    4. Hydraulic power supply
    5. Electrical/electronic control
    6. Automatic machine lubrication
    7. Machine monitoring device (optional)

The machine designer has the choice to make as to whether the friction welding machine head moves to a fixed component clamping arrangement, or the reverse.   Most machines in operation today use the former approach, which is discussed in the following descriptions:

l. Friction Welding Machine Head

The head consists of a steel casting carrying a machine spindle supported in tapered roller bearings.  At the work end of the spindle, a work holding device in the form of a draw bar operated chuck or collet arrangement is used to grip one component to be welded.  Extending upwards from the spindle, located between the bearing, is a silent chain drive to a motor which is generally an AC or DC electric motor.   The motor shaft-extension carries a brake and/or clutch.

The Head is moved forward on machine ways by a hydraulic cylinder which supplies the heating and forging pressures.

2. Machine Base

The machine base is of heavy fabricated construction. On its top surface are machine ways which carry the friction welding machine head.  It also carries the self-centering clamps and backstop.  On small machines (30 ton and smaller), the bed is often designed to absorb the loads associated with the process.   On larger machines (60 ton and larger), a system such as the tie-bar and base design often provides the required machine rigidity.

3. Component Clamping Arrangement and Backstop

In front of the friction welding machine head is a self-centering clamping arrangement. Its function is to clamp the stationary component to be welded.   The clamping device is usually bolted to the machine bed. Behind the clamps is a component backstop, which is also bolted to the machine bed.

In some cases, the clamping device is replaced with special work holding assemblies, depending on the shape of the component to be welded.

4. Hydraulic Power Supply

The friction welding machine uses hydraulic cylinders to move the head, open and close the chuck, open and close clamps, apply brake, etc.  The hydraulic control valves and pumps are usually designed to operate the machine at pressures lower than 2000 psig.  A double vane pump is used to move the head forward rapidly, and to supply high pressures for heating and forging phases.

5. Electrical Control

Many modern friction welders are equipped with a linear encoder located between the head and base.  Some utilize an MDI disc to control and monitor the process and machine.  A standard controller might have a memory capacity which can store up to 30 different component areas.

6. Automatic Machine Lubrication

Most modern friction welders have automatic lubrication of key machine parts, such as bearings, machine ways, etc., so that the frictional resistance of machine moving parts is reduced to a minimum.  On automatic machines, loss of lubrication is indicated as a machine fault (see section on machine monitoring device).

7. Machine Monitoring Device

Most modern controllers will monitor the following:

    1.  Initial component lengths
    2.  Rate of length loss during the heating phase
    3.  Forge collapse length
    4.  Final component length

Process faults are often indicated by a flashing light warning system.   The faulty component cannot be removed on some systems until the fault release switch is key activated.  This key can be removed and kept by the supervisor to enable further control.

When changing over from one diameter component to a different diameter on a friction welding machine, the typical items that must be changed are as follows:

    1. The position of one set of 3 chuck jaws on the chuck face
    2. One set of 2 clamp inserts
    3. One spindle backstop. (If also a change in length is required)
    4. One clamp backstop
    5. Two pressure settings - heating and forge
    6. One loss of length setting.

The variable parameters of the friction welding process are as follows:

    1. Speed (only when DC drive is used)
    2. Pressure
    3. Loss of length (or time)

Let's now take a look at these parameters one at a time:

1. Speed

The function of the rotational speed is to produce a relative speed at the periphery of the components in excess of 250 SFM (for steels).  This empirical figure is the same for solids and tubular components.  Speeds below 250 SFM produce very high torques in the material (for a set pressure), and have a tendency to tear the metal fibers.  There is no real limit as to the highest speed. Welds have been made up to 2000 SFM.  However, in production welding machines, the SFM is usually arranged to be within 300-650 SFM.  As an example, a machine spindle speed of 600 RPM will comfortably weld steel products of 2" dia. to 4" dia. (in fact larger than 4" dia.).

The formula for calculating SFM is as follows:

    Surface feet per minute (SFM)  =  Spindle RPM  x   Component dia. in feet

While high rotational speeds can be used, they do not increase the speed of welding.  Very high speeds ultimately lead the machine designer to use specialized bearing arrangements, which can be a source of maintenance problems.   From a weld quality standpoint, speed is generally the least important parameter.

2. Pressure

There is a wide range of pressures that might be applied (for steels) to obtain a sound weld.  However, it is recommended that the heating pressure is 4 tons per. sq. in. of the component cross sectional area, and the forge pressure is 10 tons per sq. inch.  These pressures can be varied if required to produce specific changes in physical properties, e.g. tensile, torsion, ductility, fatigue life, impact, etc. (one relative to the other).

In order to calculate the machine settings for pressure, one needs to know the area of the hydraulic cylinder and the area of the component to be welded.   It is then possible to calculate the PSI gauge setting for the machine.   Tables for these settings for various diameters are supplied with production machines.

The formula for calculating the PSI guage setting is as follows:

    PSI Gauge = Tons/sq. in on component x component area (sq. ins.)
                                           2000 x cylinder area (sq. ins.)

As previously mentioned, it is possible if required to modify the pressure input during the heating phase.  This can be done to produce "back heat" or pre-heat into the components, to achieve a relatively slower quench rate.   If this is required, a pressure of 1.5 tons per sq. in. of component area is applied before the heating pressure.

3. Loss of Length (or Time)

A fundamental requirement of a solid-phase weld is that at least a minimum amount of the original component surfaces are removed within a certain period of time in order to bring together clean parent materials at the forging phase.  The simplest means to achieve this situation is to apply the heating phase for a pre-set time.

For a set spindle speed, low pressures produce heat at low temperature with little or no loss of length.  High pressures produce high temperature with low total heat, because the loss of length is rapid and the heated metal is pushed into the flash.

Friction Welding to a Length Tolerance

Because the continuous drive process can be controlled on loss of component length, and because it has a basically infinite energy source, it is possible to load the machine with components having normal engineering cut-off length tolerances from a saw or a forge shop (draft angles), and join them to length tolerances of typically +/- .0.020".

The chucked component is backstopped in the head spindle, and has a physical dimensional relationship to the head.  The clamped component is backstopped at the rear of the clamps on a bracket bolted to the machine base, and has a physical dimensional relationship to be base.  Movement of the head relative to the base therefore represents movements of the components.

A machine can be set to weld minimum length toleranced components.   The minimum length loss for optimum quality welds is set on the linear encoder located between head and base, on machines so equipped.

Components having lengths greater than minimum tolerances are subjected to slightly more length loss before the forging phase due to the continuous drive system.   Components that are shorter than minimum lengths are recognized by such machines and rejected.

Flash Removal

In some component geometries, the flash collar produced by the process can be sheared off in the machine while still hot.

An optional feature available on many modern continuous drive systems is flash removal.  The flash removal would typically be accomplished by rotating the spindle, welding, and then removing the flash by plunge cutting in about 3-5 seconds.   In this operation, the friction welder is used much like a lathe.

Care should be taken when considering immediate flash removal in a machine.  In some cases, such as hardenable steels, it is advantageous to leave the flash on for a while, as it provides a heat sink which reduces the weld cooling rate and provides a rare ductile joint.

Continuous Drive Friction Welding

The continuous drive process is the type most commonly used for friction welding throughout the world.  Its parameters can be applied very simply as pressure and loss of length against a fixed spindle speed (or, it can be modified to provide a greater range of desirable variables).