[Skip to content]

TWI
Search our Site
.

Cutting the cost of cutting

TWI Bulletin, March/April 1995

 

Martin Bourton
Martin Bourton

On graduating in Electronic Systems Engineering from the University of East Anglia, in 1987, Martin Bourton joined the Microcomputer Section of the Arc, Laser and Sheet Processes Department at TWI. His research work has largely been devoted to the development of expert system and multimedia computer programs.




Lynda Mackridge
Lynda Mackridge

Lynda Mackridge joined TWI in 1988. Following a one-year training programme, she became part of the TWI Microcomputer Section (Arc, Laser and Sheet Processes Department), where she completed an HNC in Mechanical and Production Engineering. Her work has largely been related to advanced computer software packages.





Paul Anderson
Paul Anderson

Paul Anderson is a Senior Research Welding Engineer in the Arc Welding Section of the Arc, Laser and Sheet Processes Department at TWI. After obtaining BEng (Hons) in Materials Technology at Coventry Polytechnic, he spent a year with British Steel, at the Welsh Laboratories, Port Talbot. He joined TWI in 1990 and has been actively involved with the development of the gas-shielded arc welding processes. His recent activities include the development of shielding gases for the TIG welding of duplex stainless steels and the assembly of a prototype top face penetration control system.

Reducing the pounds and pence spent on oxyfuel and plasma arc cutting is crucial to the economic success of many fabricators. A new piece of software makes detailed expertise on cutting selection available at the keyboard as Martin Bourton, Lynda Mackridge and Paul Anderson explain.


The cutting process can have a significant impact on the cost of a fabrication, due to the 'visible' cost of the labour, consumables and equipment, and the 'invisible' costs arising from the subsequent effort to achieve the required surface quality of the component. A wide range of cutting processes is available to the fabricator, which differ widely in terms of cutting speed and the quality of the cut surface. In order to select the cutting process to achieve the optimum combination of cost and quality, consideration must be given to the performance of every process option - a large and complicated task. Historically, this has been avoided by using the experience of the cutting engineers and moving directly to the practical trials. However, there is a limit to the expertise that can be developed in this way and the practical trials may involve significant expense.

TWI, in conjunction with Air Products, have developed CUTTING SELECTOR, a PC-based expert system for oxyfuel and plasma arc cutting, which makes detailed cutting expertise immediately available to fabricators, and allows the latest developments in cutting technology to be included in updated versions of the program.

Cutting processes

The cutting processes which are most widely applied in industry are oxyfuel and plasma arc cutting, these account for approximately 90% of the equipment sold. [1]

In oxyfuel cutting, a preheat flame, formed by mixing oxygen and a fuel gas, is first directed on to a spot on the metal, which is heated to above the ignition temperature (typically 900°C for C-Mn steel). A jet of pure oxygen is then injected through the flame which exothermically reacts to melt the metal and burn through the spot. The molten metal and slag are removed by the velocity of the oxygen stream. The critical factors which determine the cutting performance for a given application include the selection of the fuel gas and the design of the gas nozzle. Options for the fuel gas include acetylene, propane, propylene, MAPP (methylacetylene-propadiene) and natural gas. The range of commercially available nozzle types includes parallel, convergent, curtain and divergent designs.

The plasma cutting process operates by using heat from an electrically generated arc plasma to melt the metal. The plasma gas may be either essentially inert, such as nitrogen or argon-hydrogen, or oxidising, such as air or oxygen. The critical factors which determine the cutting performance for a given application include the plasma gas, the torch design and the electrical power of the equipment.

Each combination of process variants ( ie the fuel gas and nozzle type during oxyfuel cutting or the plasma type, torch design and electrical power during plasma cutting) offers an individual combination of direct costs due to labour, consumables and equipment and of the characteristics of the cut surface (including the kerf width, kerf angle, level of dross, level of oxidation and degree of edge hardening).

Traditionally, the cutting processes have been selected on the basis of experience and practical trials. However, although useful, there is a limit to the expertise that can be developed in this way. Information can be slow to spread and inevitably industry bears the expense of conducting separate trials in order to discover the optimum conditions for the same application. Furthermore, the rapid development of process variants can result in accepted applications lagging far behind the optimum configuration.

Expert systems

Expert systems are computer programs that encapsulate the specialist knowledge of a human expert in a form that can accessed by users. In the same manner as talking to an expert, an expert system will process uncertain data or opinions to provide a recommended solution to a given problem. [2 , 3] Compared to simple databases or conventional procedure-based programs, written in a language such as Pascal, expert systems can be applied more readily to the provision of expertise specific to an entered application, whilst at the same time reducing the programming effort both to initially construct the sophisticated user interface required for this type of application, and to make subsequent modification or updates to the program.

Cutting selector

The selection of a suitable cutting process on the basis of technical and economic factors is an area which is ideally suited to an expert systems approach. An object-orientated expert system shell, Egeria, was chosen as the programming medium. [4] The selection process was accomplished using object-orientated code, and the expertise particular to the entered application represented using conditions and rules. Furthermore, the use of Egeria permitted the program to be linked to the shielding gas selector expert system, GAS SELECTOR.

Selection of the cutting process

The selection of a suitable process is dependent upon a number of factors. During a consultation with the CUTTING SELECTOR system, the user is requested to enter the following details about a joint:

  • Application details: the material type to be cut, component thickness, length of cut and starting position.
  • Process information: equipment availability, purity of the available cutting oxygen, frequency of operation, degree of mechanisation, duty cycle and labour cost (including overheads).
  • Acceptable properties of the cut surface: kerf width, kerf angle, level of dross and level of oxidation.
Fig.1. The influence of oxygen purity on travel speed for oxyfuel cutting.
Fig.1. The influence of oxygen purity on travel speed for oxyfuel cutting.

The program has a number of features to assist the user in answering these questions including a series of help screens defining the technical terms used in the program (for example, oxygen purity and degree of mechanisation), graphics screens showing the kerf width, kerf angle and the influence of oxygen purity on travel speed, Fig.1, and input ranges to ensure that the user does not enter invalid values.

The details of the proposed application are compared with the capabilities of the available cutting processes. This allows a list of suitable process options to be compiled. Currently, the expert system contains characteristic data for 15 variants of the oxyfuel and plasma cutting processes.

Economic analysis

In order to recommend the most cost-effective cutting process, the program carries out a detailed comparative economic analysis of the cost per metre of cut for each of the suitable processes. Using the application data and process information data, the program first calculates the time required to complete the cut, based on mid-range performance data from a number of published applications, Fig.2. The cost of consumables, direct labour and indirect charges/overheads is then calculated.

Fig.2. Selected data for cutting speed/thickness relationship.
Fig.2. Selected data for cutting speed/thickness relationship.

The cost of consumables is the sum of the cost of the gas, the electricity consumed (during plasma cutting) and an allowance for consumables (for example, cutting nozzles for oxyfuel cutting and electrodes for plasma cutting). The cost of gas consumed per cut is calculated separately for plasma gas or fuel gas and cutting oxygen, based on the time required to pierce and to make the cut, gas flow rate and gas cost.

The direct labour cost per cut is calculated using the time required to complete the cut and the labour cost per hour.

The indirect labour cost is calculated using the direct labour cost and the duty cycle. An allowance is added to cover cost of capital depreciation associated with both the cutting equipment and the level of mechanisation.

The sum of the consumables, direct labour and indirect labour/overheads is then divided by the length of the cut to calculate the cost per metre of cut.

Display of results

The list of suitable processes is displayed in order of total cost per metre, ranging from cheapest to most expensive, Fig.3 (the user has the option of viewing a bar graph representation of the costing analysis, Fig.4). The expert system demonstrates that in many applications labour cost is the largest component of the cost per metre of cut. This means that relative to a process selected by convention or prior practice, the process recommended by the expert system can often significantly reduce the cost per metre of cut.

Fig.3. CUTTING SELECTOR results screen.
Fig.3. CUTTING SELECTOR results screen.
Fig.4. CUTTING SELECTOR bar graph representation of cost analysis.
Fig.4. CUTTING SELECTOR bar graph representation of cost analysis.

Technical assistance

In addition to identifying the optimum cutting process, CUTTING SELECTOR can be used to provide technical assistance particular to the entered application. As well as the recommendation of the cutting speed and gas flow rate, the program can provide the following assistance:

  • Assistance in correcting the most common deficiencies in the cut surface, Fig.5.
  • General advice on oxyfuel cutting.
  • The approximate welding duty cycle.
Fig.5. CUTTING SELECTOR technical assistance screen.
Fig.5. CUTTING SELECTOR technical assistance screen.

The technical advice is determined by simple conditions and rules. The conditions identify where a feature of the application makes technical advice necessary, and the rules determine the advice to be given. For example, the duty cycle is mainly dependent upon two conditions: the length of cut and the degree of mechanisation. Rules allow the program to calculate an appropriate duty cycle for a given combination of these parameters.

Summary

CUTTING SELECTOR makes detailed expertise concerning oxyfuel and plasma arc cutting immediately available to fabricators. It can be used to reduce the cost of cutting by initial process selection, and via technical assistance particular to the entered application.


References

Author Title
1 Lucas W and Rennie S: 'Cutting processes - the right choice'. Welding and Metal Fabrication April 1993 122-127. Return to text
2 Brightmore A D: 'Microcomputer software helps welding engineers'. Joining & Materials 1988 1 (5) 221. Return to text
3 Lucas W and Brightmore A D: 'Expert systems for welding engineers'. Metal Construction 1987 19 (5) 254. Return to text
4 Harrison C: 'A detailed look at Egeria'. Expert Systems User 1989 5 (8) 19. Return to text

Originally published in TWI Bulletin, March 1995