- Text Size:
Information Sheets
Beam Delivery
Beam delivery for CO2 laser systems
Introduction
The importance of the optical elements in a laser beam delivery system lies in the ability of the system to transfer the laser energy to the workpiece in the desired way. Aspects of this energy transfer which may be of importance include:
- Minimization of power losses.
- Minimization of losses of beam 'quality'.
- Control of polarization.
- Energy density/depth-of-focus tradeoff.
If it were possible to develop complete optical models for a system, (including optical conditions vs. process efficiencies) then, with some effort, it would be possible to define the ‘best economy’ system for any purpose. In general, the laser industry is not in a position to do this, although some broad guidelines do exist.
Laser power transfer
In a simple laser beam-delivery system there is a sequence of mirrors and lenses. It is possible to design and build fully-reflective optical systems under certain circumstances. In order to calculate the expected power transfer it is required to know the transmission (or reflection) from each optical element as a fraction of unity.
If a mirror reflects 98.5% of incident radiation, then, for our calculation, reflectivity R = 0.985.
The figure below is an example system, with an in-line phaseretarder unit consisting of one reflective phase-retarder component plus three gold-coated mirrors. Next is a two-lens beam expander, followed by two more mirrors and a final lens. The overall laser power transmission of the beam line, T, is found from multiplying together the values for each element:
T = R1 x R2 x R3 x R4 x T1 x T2 x R5 x R6 x T3
If all mirrors have R = 0.985 and all lenses have T = 0.995, then we obtain T = 0.8997. In other words, 10% power loss can be expected in this case, even when the optics are new and uncontaminated with process debris.
Power losses can be reduced by choosing enhanced-reflectivity mirrors. Enhanced-coated Si or Cu mirrors have R = 0.996 or better, but can be more expensive than gold-coated mirrors. Calculating the economics of mirror choice on this basis can be extremely difficult, and laser users will often make choices that are based on sound personal experience at best, ‘availability’ sometimes, and inexplicable prejudice at worst. Other power losses may occur through beam-clipping at apertures within the system.
Calculation and modelling of beam clipping is easy for the highest quality (TEM00) of laser beam passing centrally through a circular aperture - a simple formula exists for this calculation. If the beam is not centred, then exact calculations become very difficult, but computer simulation of this condition has been found to work very well.
Exact calculations, or computer simulations, for multi-mode beams are, in general, not possible. This is because the actual intensity distribution of the laser beam is not always available even if the beam quality factor is known.
The table below gives the fraction of power lost at an aperture of diameter A for a TEM00 beam of 1/e² diameter D. The first column states the power loss for a centred beam; the second and third columns give power loss for de-centred beams for de-centration 0.1A and 0.2A respectively.
Figures in Table 1 are from a computer simulation and are not exact analytical results. Inspection of the table shows the importance of selecting optics (and designing optic mounts) with sufficient clear aperture, and of maintaining beam alignment/centration through the system.
Transfer of ‘beam quality’
Laser beam quality is related to ‘lateral mode structure’. Various symbols have been used to describe beam quality, such as ‘M-squared’ value, the ‘Quality factor’, Q, and the ‘Beam Propagation Factor’, K. These are numerically related as follows:
Q = M² = 1/K
A TEM00 beam has M² =1. Laser beams of higher order mode structure have M² values greater than 1. If the optical components in a beam delivery system introduce non-spherical errors to the laser beam wavefront, then the beam quality reduces, with consequent reductions in focused energy density, or depth-of-focus, or both. Most industrial CO2 lasers have excellent beam quality, which can, in principle, be maintained through the beam delivery system and up to the focus. In order to do this, we have to avoid, as far as is practicable, optical aberrations and thermal lensing. Optical aberrations can be calculated by an optical designer, and lens and mirror systems which either minimize or eliminate aberrations are often possible. Where residual optical aberrations do exist, it is extremely difficult to ‘combine’ their effects with the laser beam properties in such a way as to arrive at an exact, or fully analytical, result. Thermal-lensing effects can be modelled. Relatively small amounts of thermal lensing can be ignored in most laser systems, but as thermal-lensing increases there comes a point where transmissive optics cause intolerable axial shifts of focus, and where mirrors (used at 45° incidence) give rise to unacceptable astigmatism.
Current models for thermal-lensing indicate that, in general, reflective optics are better. However, transmissive optics are often much more convenient.
Polarization
The efficiency of some laser processes is dependent on the state of beam polarization at the workpiece. The cutting of metals is highly polarization dependent. Process quality and efficiency is greatly reduced if the electric field of the radiation is at 90° to the direction of cut, for example. Most industrial lasers emit linearly polarized radiation, such that the electric field of the radiation is in a constant direction.
A compromise solution has been found whereby the direction of the electric field was made to spin once for every cycle of the radiation (once for every wavelength), or about 3x10^13 times per second. This is called circular polarization, and has the effect of allowing processing with equal efficiency in all directions in an X-Y plane.
Linear polarization is converted into circular polarization by use of a component called a phase-retarder, which imparts a 90° phase-shift. This component (or sometimes two components each with 45° phase shift) can be in error. A typical specification for phase-shift accuracy is 90°±6°.
Other mirrors in the system can contribute to phase-shift error. A few years ago standard Si mirrors had a phase-shift error of about 5°, but modern developments in optical coatings have reduced this to about 2° or better. Gold-coated mirrors have a natural phase shift of about 1°, but cannot offer the high reflectivity of enhanced-coated Si mirrors.
Enhanced-gold coatings with low phase shift and high reflectivity are available, but can be more expensive. True circular polarization can be achieved by the design and construction of a variable phase retarder instrument, but no market demand has been observed. Perhaps the ultimate solution for flatbed metal cutting will be systems capable of producing directed linear polarization at the workpiece. Energy-density and depth-of-focus
In a simple laser cutting system the influences on choice of lens are straightforward to describe.
The energy density which will be achieved at the workpiece depends on the laser power and the focused spot ‘area’; the spot area being proportional to the square of the spot diameter. Whichever (aberration-free) lens is chosen, the product of the energy density and depth-of-focus (DOF) remains constant. So, if the lens focal length is doubled we have twice the spot diameter, one quarter the energy density, and four times the DOF, assuming a constant beam at the lens. In seeking the optimum cutting conditions, the laser user is trying to identify the combination of optical conditions and gas flow conditions which give the required cut quality at the best cutting speed. Although the cut kerf width may be roughly linearly dependent on lens focal length, the energy density and DOF will each depend upon the square of focal length. At this time it is not possible to model process efficiency for a variety of materials, thicknesses, assist gas conditions and optical parameters. Accordingly, only empirical trials are able to guide the user’s selection of lens.
Some laser systems involve more complex optics. Examples include plastics processing where axicon based optics may be used to drill neat, round holes of ‘large’ size, and galvo-scanner systems for cutting textiles and paper in pre-determined patterns at high speed. Galvo-scanner systems are a special case, where (for CO2 lasers) the cost of building truly aberration-free focusing optics is prohibitive.
In these cases the solution most often employed is the design of an ‘optimum-compromise’ single element lens, where there is a tradeoff between the desired focusing characteristics, these being:
- Smallest spot size
- Flatness of scanned field
- 'F-theta' scan characteristic
Some types of cutting system usefully employ special optics which modifies the conditions at the focus. Certain CO2 laser systems used for high-speed partial cutting of artwork materials have benefited from axicon-based focusing lenses which yield a more flat-topped focused energy distribution.