Cutting on edge

Conventional laser cutting is routinely used for plate cutting, however, the thickness is limited to 25 mm using 5- or 6kW CO₂ lasers. Plate cut for subsequent welding processes often requires some form of joint preparation, with the most common being either single or double bevel. Generating these edge preparations is both costly and time consuming.

Although lasers have revolutionized the thermal cutting market recently they have yet to make their mark in the area of bevel cutting. A major factor limiting their capability is that for thicker C-Mn steels the laser cutting process uses low-pressure oxygen assist gas and relies on gravity to assist in molten material removal. There have been some high-profile examples of laser bevelling, such as the work done by General Motors Diesel Division, where 12mm-thick ballistic-grade steel was cut at 45° giving a total cut depth of 17 mm. On thicker steels the oxygen-assisted laser cutting process suffers from uncontrolled gouging and burning away of the lower edge. It also has limitations due to sensitivity to steel composition and surface condition.

Most fabricators think of oxy-fuel cutting as the traditional process for the manufacture of weld preps. Commercial systems comprising three separate torches are available that can make a complex double V with nose in a single pass. Using oxy-fuel for bevel cutting requires a skilled operator, controlling of the torch tip stand off, the cutting speed, and the gas flows in order to get an optimized result. The high heat input where the oxy-fuel flame preheats the material results in a HAZ that can extend to several millimeters away from the cut edge and which can also result in top edge melting, a feature that can be accentuated in bevel cutting. The oxy-fuel process requires significant time to preheat the material prior to cut initiation (20 seconds on materials >30 mm).

Robotic plasma cutting is another widely used process for bevel cutting; either with a twin head system or as a two-pass operation. The plasma cuts can be done with either nitrogen or oxygen, however, nitriding of the cut face can lead to weld porosity, and so oxygen is typically the preferred gas. Plasma cutting is considerably faster than oxy-fuel cutting but maintaining bevel edge quality can sometimes be problematic.

The Lasox process was developed by BOC and a system was installed at Bender Shipbuilding (Mobile, AL) in 2002 to extend the cutting thickness range of its Tanaka LMX III laser cutting system.

In the Lasox process the laser beam passes coaxially through the gas nozzle. Using an appropriate optical arrangement a condition can be achieved whereby the footprint of the laser beam on the plate is larger than the gas jet. The laser power required by the process is dictated by the ability to increase the surface temperature to >900°C over the whole of the gas jet interaction zone. This can typically be achieved with only 1kW of incident laser power. Additional laser power fails to give any benefits to the process in terms of productivity; in fact it can be detrimental leading to top edge melt. Careful balancing of the laser energy input with the process requirements ensures that the oxidative reaction will be initiated across the whole width of the oxygen jet. Failing to meet this condition results in intermittent and uncontrolled reaction with very poor edge quality.

Lasox cutting speeds are comparable to those achieved with oxy-fuel. The width of the kerf is governed by the impinging jet diameter with values of 2-3 mm being typically achieved in the 20-50 mm thickness range. Taper is generally low at <2° and the cut edge quality is similar to oxy-fuel with a smooth cut face from top to bottom exhibiting little variation in roughness through the thickness. The process does not exhibit the striation features that are characteristic with conventional laser cutting.

The key benefit that Lasox offers over plasma is the squareness of the kerf, while the main benefits over oxy-fuel are the lack of top edge melting and elimination of the need to preheat the plate before cut initiation.

In order to demonstrate the bevelling capability of the process a Lasox system was installed on a 5-axis TRUMPF laser-cutting machine at the National Laser Centre in Pretoria South Africa.

A trial was made to assess the bevelling capability of the Lasox process in 25mm and 30mm plate. A wide range of bevel angles and some specific bevel geometries were attempted to evaluate the process window of operation of the process.

The Lasox process was initially optimized for normal cutting, and the key parameters in terms of beam position, nozzle stand off, gas pressure, and laser power were established as well as the range of cutting speeds over which good cut quality was achieved as defined by the following:

• Smooth cut face with no signs of deep fluting or gouging
• Parallel drag lines perpendicular to cut face (curved drag lines tend to indicate non-optimized speed)
• Complete oxidation at the lower edge of the plate (no resolidified metal)
• Crisp square top edge (no top edge rounding)
• Readily removable oxide dross on bottom edge of cut (easily removed with chipping hammer)
• Negligible cut taper <2° (or bowing of the cut face)

The results showed that no specialized parameters were needed and that regular cutting parameters gave good results for the equivalent effective bevel cut depth.

The material used throughout the experimental work was a Grade 43A, which is generally considered to be the most "open" C-Mn steel, highly variable in terms of quality and consistency. The material was used in an "as received" condition—no surface preparation was carried out.

The bevel angle was achieved by tilting the laser beam to the plate, and all of the samples were made with edge starts without preheating the plate.

Conventional bevel cuts were made in 26mm and 30mm plate at angles ranging from 15° to 45° as well as more complex joint preparations, including double V preps and single V preps. All of the trial cuts were made using <2 kW of laser power, 8.5 bar O₂, and at cutting speeds of 160-200 mm/min. The maximum cut depth achieved was 42 mm based on a 45° bevel in 30mm plate.

The complex bevels such as the double V and those with small noses required multiple cuts to be made a relatively straightforward programming task for the 5-axis CNC. The controlled heating in the Lasox process and the high stability of the oxygen jet mean that, when coupled with the highly accurate laser 5-axis positioning system, accurate and repeatable results were obtained.

One of the key benefits of Lasox when compared with oxy-fuel bevelling is the reduced HAZ, particularly on the top surface of the plate where there is a considerable HAZ associated with the zone of flame impingement. The etched samples cut by Lasox clearly show a very parallel HAZ, which widens slightly at the bottom of the bevel as a result of dross attachment. In all cases this dross comprised oxide slag and not resolidified material and hence could be readily removed with a chipping hammer—common practice in oxy-fuel bevel cutting.

Not only are the dimensional tolerances important in order to achieve the required fit up but also the edge quality is of significant importance in the weld prep. The cut faces of selected samples were measured and showed that for the 26mm plate a stable edge quality in the 4-µm range was obtained, which correlated well with the edge quality achieved with oxy-fuel cutting/bevelling.

The trials clearly demonstrated that the Lasox process offers the capability to bevel cut C-Mn plate and may offer a viable alternative to oxy-fuel and plasma. A simple cost analysis does show that producing bevels using the Lasox process is more expensive than conventional processes, but it can be done with existing laser systems with minor hardware modifications. A laser cutting system with only 2kW that can profile cut and bevel C-Mn steel up to 50 mm is now a viable proposition.