A laser is an energy conversion device. The industrial laser converts electrical energy to light energy, but it does not do this very efficiently. A carbon dioxide laser has a wall plug efficiency (that is, the energy output, measured in watts of average power, divided by the electrical energy drawn from the wall) of less than 10%, and Nd:YAG lasers have wall plug efficiencies of 1 or 2 %. The remainder of the electrical energy input is carried away by the flow of cooling water. In spite of these inefficiencies, the beam can be delivered very efficiently to the workpiece. Newer generations of lasers such as diode lasers have a considerably larger energy efficiency than the carbon dioxide and Nd:YAG lasers currently widely used.
Because of the low energy efficiency, it is important that the energy be used expediently when applied to the weldment. There are two energy efficiencies that are important. The first is the energy absorption efficiency, also called the energy transfer efficiency, or arc efficiency in the case of an arc welding process. This designates the fraction of the incident beam energy that is absorbed in the workpiece. The second energy efficiency is designated the melting efficiency and is characteristic of what happens to the energy once it is absorbed in the workpiece.
The energy absorption efficiency is the fraction of the laser energy directed at the workpiece that is absorbed into the workpiece. Two reasons why incident energy may not be totally absorbed by the material are reflection from the workpiece and transmission through the workpiece. Some of the energy may be reflected from the surface and not absorbed. The absorption of laser radiation into metals depends on the nature of the metal, the temperature, the wavelength of the laser, the roughness or surface condition of the metal, and the angle of incidence of the radiation onto the material. The wavelength of the Nd:YAG and diode lasers result in greater absorption into most metallic materials than that of the carbon dioxide laser. As temperature increases, however, so does the absorption of the laser radiation by the material. Thus, even in a material that is largely initially reflective, as long as the material absorbs part of the energy of the beam and starts to heat, a larger fraction of the beam is absorbed and the heating process accelerates. If a keyhole is formed in the material, the keyhole acts as a trap and most of the beam is absorbed.
If the keyhole extends completely through the material, some energy may pass through the bottom of the keyhole. In the optimized welding process, this energy loss is minimal. A more serious energy loss is from absorption of laser light in a plasma which may occur above the material surface, or light scattering from a plume above the surface. The plasma consists of ionized gas which absorbs energy from the laser beam by a process known as inverse bremstrahlung absorption (Offenberger et al., 1972). The absorption increases with the square of the wavelength and hence is considerably more severe for the carbon dioxide laser than for the Nd:YAG laser. When welding with carbon dioxide lasers with power greater than a few kilowatts, a plasma suppression jet directs a flow of helium gas at a slight angle to the surface, immediately above the keyhole region. The jet is usually positioned to blow the plasma, and other gas coming from the keyhole, onto the cold surface immediately ahead of the weld area. This has been shown to limit the loss of easily vaporized alloying elements from the weld metal (Blake and Mazumder, 1982) and also may serve to preheat the metal ahead of the beam.
The net absorption of laser energy (or energy from an arc welding source) into the weldment is most accurately measured by a Seebeck calorimeter. This is an insulated box with an insulated lid that is manually shut immediately after the weld is completed. The box contains thermocouples that measure the total flux of heat through the walls of the box during its return to room temperature; by integrating the flux of heat, the amount of heat absorbed by the metal can be evaluated. Measurements of absorption efficiency have been presented by Banas (1986) for a variety of different materials.
The melting efficiency is the fraction of the energy absorbed by the material that is used actually to melt the metal to form the weld. Measuring this quantity requires the use of instrumentation such as the Seebeck calorimeter to determine the amount of energy absorbed by the material. The amount of energy to perform the weld is determined by examining cross-sections of the weld to determine the area, knowing the speed of the weld, and using data for the heat capacity for the material between room temperature and the melting point and the heat of fusion. Usually several cross-sections are taken, and an average calculated, to account for possible fluctuations in the material or the process.
The melting efficiency is a consequence of the heat flow patterns in the material. Many conduction welding processes can be thought of as originating from a point source moving across the surface of the material being welded. In this case, heat can be conducted in three dimensions away from the point source, in the direction of motion of the heat source, perpendicular to the direction of motion, and into the depth of the material. On the other hand, a deep penetration weld can be thought of as originating from a line source of heat, extending through the thickness of the material. In the latter case, heat conduction away from the heat source is only two-dimensional as there is already a distribution of heat through the thickness of the material. Thermal losses are less severe in the latter case since there are fewer dimensions in which the heat can be conducted away. Some thermal losses are inevitable; it is impossible for one area to be heated to above the melting point without some heating of the surrounding area. It has been shown that the maximum possible melting efficiency for welding is 37 % for three-dimensional heat conduction and 48% for two-dimensional heat conduction (Swift-Hook and Gick, 1973).