For more than 5,000 years metal has been casting raw ore into usable tools. At first the castings were limited to very simple and small structures, but nowadays we see castings weighing anything from an ounce to thousands of pounds. New casting processes developed over the last hundred years allows modern casters to produce very thin and complex castings that can be made with predetermined quality characteristics at levels well below the visual. These are not single castings produced in a laboratory; these are mass productions with minute tolerances for dimensions, grain structures, mechanical properties and strength.
19th Century
Originally, casting was limited to gravity pouring of liquid metal into a mould. In the nineteenth century, a more mechanical casting approach was developed. The first die casting-related patent was granted in 1849 for a small hand operated machine for the purpose of mechanized printing type production (1), the forerunner of today's high pressure die casting machines.
In the high-pressure die casting, process liquid metal is not poured by gravity into a sand mould, but forced by high pressure into a steel die. The advantage of this process is that it is the shortest way from liquid metal to a final product. Casting walls can be very thin and most dimensions do not have to be machined after the casting process. Hundreds of thousands of castings can be made from a single die, keeping the costs low and the quality high.
Where there is light there is shadow. The advantages are bought by a costly investment in the die casting machine and related equipment and in the die in particular, high quality steel in which the casting shape, a melt feeding system (runner) and cooling lines are machined into. Finding the right die design fulfilling its purpose required a lot of experience of the responsible engineers, as well as extensive attempts using trial and error.
21st Century
There has been tremendous progress in modernization and improvements in high pressure die casting during the last decades, including automatic sprayers and casting extraction systems, more efficient melting furnaces and pouring devices, better controlled die-casting machines and computerized monitoring systems. All these improvements made die casting processes more efficient, better controllable and helped producing higher quality castings.
At the same time, however, today's die casting industry is facing challenges early casters did not see. Experienced metalcasters are moving into retirement - yet at the same time, many shops are having difficulty in recruiting young engineers to train as replacements. The demand for higher quality standards for the castings, international competition, customer demands for lower prices and the difficult economic situation across the world, all affect the bottom line and company profit margins. Add to this the increasing need to shorten product development times requires that the caster produces good castings the first time, or the cost structure to make money is in jeopardy.
Process Simulation
Toward the end of the 1980's, process simulation was introduced to the world of die casting. Process simulation allows the metalcaster (or designer or engineer) to create a simulation of the part to be cast, predict possible outcomes based on input variables, and optimize those outcomes by using the processing power of the computer, rather than the traditional experience-based trial-and-error methodology. With the dawn of the desktop computer and workstations on every desk, process simulation became available to every engineer. Supported by Computer Aided Design (CAD) systems volume models could be created fast and be used as direct input for simulations.
Simulations done in1993 could take up to a month for the casting shape input and an additional month for the calculation and result evaluation. Today, almost 20 years later, complete casting files are loaded into the software by mouse clicks and be calculated in minutes. Now the result evaluation is the most work-intensive part of the simulation and takes the longest time. Engineers have to go through different results files and compare them to each other to get a good understanding about the used design and process parameter. Despite the fact that the soft- and hardware are much faster than years ago, the human factor is not and engineering takes still hours or days until a final recommendation can be made.
Autonomous Optimization
Leading software provider did recognize the result evaluation as an engineering bottleneck and changed their proprietary software from a single simulation approach to autonomous optimization.
The software takes a defined set of designs and parameter for the first simulation iteration. Then it changes these designs and parameter based on given tolerances and the results found. The process follows the rule of evolution: each layout variation is kept, eliminated, modified or combined with an already calculated or new design (2).
Fill Parameter Optimization
The filling process is determined by plunger movements pushing the liquid melt into the die cavity. Melt is poured through a hole into a tube, called a shot sleeve, which is closed on one side by the plunger and on the other by the die. After filling the sleeve to a specific level, the plunger slowly moves forward pushing the melt towards the die. When the melt reaches the casting cavity, the plunger is accelerated to high speed to fill the cavity in 1/10th of a second. The filling speed is critical as filling slower can allow the melt to get too cold while filling faster does not give the trapped air time to escape.Melt that is too cold or contains too much air will reduce the casting quality. Finding the right compromise is an important task.
This 'change and comparison' allows the software to optimize the casting process within the given tolerances, but without having an engineer spending time to go through results and make improvements manually. The engineers have time to concentrate on more important project tasks than simulation set-ups, result evaluation and comparison.
Simulating all possible variations and comparing them would be nice, but based on the high number of possible variations it is, in most cases, impractical. For example, a simulation having only six parameters and five variations would end in 7,776 simulation runs. Assuming an average time of three minutes per simulation, the entire project would take over sixteen days to complete, which is time that most metal casting engineers do not have anymore. Today software uses genetic optimization algorithm. As in the biological world, the evolutionary process of autonomous optimization occurs over several calculation generations. Based on the defined objectives, such as high temperature and/or low air volumes, a generic algorithm creates new variations of the filling parameter. The process is repeated until design modifications do not lead to additional improvements.
Using variations of these fill parameter is an easy task for the optimization software. The engineer simply creates a template including the parameter to change, the step variation and tolerances, and the software takes over. It selects a starting design and changes the parameter according the variations and tolerances for single simulations.
Interpretations of filling results are done automatically based on the melt or die temperature, fill time or air volume values at end of fill. A once tedious and time-consuming, but critical, engineering task has now been relegated to an automatic and more accurate process.
Component Optimization
Getting a new project or the opportunities to change the casting or die design, such as adding fins, more wall stock, overflow locations, gate locations, ingate variations, cooling or heating lines, die materials, etc., calls for action in the engineering group. Selecting the best design out of all the given possibilities is often connected with failure, even when those involved are experienced. Autonomous optimization, on the other hand, opens the possibility to simulate and find the right elements for the casting projects. Simply by designing these single objects and switching them from simulation to simulation is an easy task for the software.
For example, a particular set of components is calculated and the runner model is selected to be replaced by one with more or less branches; without engineering intervention, the software automatically loads the new design and calculates it. Based on the defined objective, the computer will find the best component configuration in a short time. What's left for the engineers to do is to evaluate the best simulation results, then release the optimal design to build the tool.
Optimization with Parametric Objects
The most exciting alternative is to manipulate objects by simply changing numerical values. Instead of drafting and designing multiple runner systems to be selected for the optimization, runner systems can be manipulated using numerical parameters. Ingate areas can be increased or reduced simply by adjusting the parameter for the ingate thickness or width. Balancing a runner system for one or multiple cavity dies is easily done by changing the numerical values for runner length, angle and/or direction. With the given templates provided a runner system can be defined, built, and prepared for optimization in minutes, including variations of multiple runner branches and ingates. Overflow and vent dimensions can be adjusted in size and volume to reduce air entrapment and porosity.
Balancing the thermal profile of a die to reduce die flashing, cold runs and to increase die life, placing cooling lines is crucial, but becomes a simple task when optimization tools are used. Line start and end locations are defined numerical and can be easily changed and dependencies can be readily defined. By changing the line temperature values, cooling and heating can be simulated as necessary.
Summary
Over thousands of years, foundry technology did not change much until the dawn of the computer age. Electronics allow building fully automatic production processes with handling robots and monitoring systems, increasing quality and productivity by reducing hard working labor forces. With autonomous optimization tools, research and development is taken out of the direct production area. Trial and error is not done on the plant floor anymore with production machines wasting time, casting volumes and costs.
In a very short time, optimization has taken over the process of simulation, eliminating manual corrections and the high cost of wasted engineering time. Engineers today define a process with tolerance windows and the objectives the process has to achieve, while optimization will come up with the best results to do so.
1.) From Wikipedia, the free encyclopedia
2.) Casting Designs Become Self-Optimizing
Metal Casting Sep/Oct 2009 Christof Heisser, Magma Foundry Technologies, Inc, Schaumburg, IL
http://www.kktooldie.com/cae-simulation/
19th Century
Originally, casting was limited to gravity pouring of liquid metal into a mould. In the nineteenth century, a more mechanical casting approach was developed. The first die casting-related patent was granted in 1849 for a small hand operated machine for the purpose of mechanized printing type production (1), the forerunner of today's high pressure die casting machines.
In the high-pressure die casting, process liquid metal is not poured by gravity into a sand mould, but forced by high pressure into a steel die. The advantage of this process is that it is the shortest way from liquid metal to a final product. Casting walls can be very thin and most dimensions do not have to be machined after the casting process. Hundreds of thousands of castings can be made from a single die, keeping the costs low and the quality high.
Where there is light there is shadow. The advantages are bought by a costly investment in the die casting machine and related equipment and in the die in particular, high quality steel in which the casting shape, a melt feeding system (runner) and cooling lines are machined into. Finding the right die design fulfilling its purpose required a lot of experience of the responsible engineers, as well as extensive attempts using trial and error.
21st Century
There has been tremendous progress in modernization and improvements in high pressure die casting during the last decades, including automatic sprayers and casting extraction systems, more efficient melting furnaces and pouring devices, better controlled die-casting machines and computerized monitoring systems. All these improvements made die casting processes more efficient, better controllable and helped producing higher quality castings.
At the same time, however, today's die casting industry is facing challenges early casters did not see. Experienced metalcasters are moving into retirement - yet at the same time, many shops are having difficulty in recruiting young engineers to train as replacements. The demand for higher quality standards for the castings, international competition, customer demands for lower prices and the difficult economic situation across the world, all affect the bottom line and company profit margins. Add to this the increasing need to shorten product development times requires that the caster produces good castings the first time, or the cost structure to make money is in jeopardy.
Process Simulation
Toward the end of the 1980's, process simulation was introduced to the world of die casting. Process simulation allows the metalcaster (or designer or engineer) to create a simulation of the part to be cast, predict possible outcomes based on input variables, and optimize those outcomes by using the processing power of the computer, rather than the traditional experience-based trial-and-error methodology. With the dawn of the desktop computer and workstations on every desk, process simulation became available to every engineer. Supported by Computer Aided Design (CAD) systems volume models could be created fast and be used as direct input for simulations.
Simulations done in1993 could take up to a month for the casting shape input and an additional month for the calculation and result evaluation. Today, almost 20 years later, complete casting files are loaded into the software by mouse clicks and be calculated in minutes. Now the result evaluation is the most work-intensive part of the simulation and takes the longest time. Engineers have to go through different results files and compare them to each other to get a good understanding about the used design and process parameter. Despite the fact that the soft- and hardware are much faster than years ago, the human factor is not and engineering takes still hours or days until a final recommendation can be made.
Autonomous Optimization
Leading software provider did recognize the result evaluation as an engineering bottleneck and changed their proprietary software from a single simulation approach to autonomous optimization.
The software takes a defined set of designs and parameter for the first simulation iteration. Then it changes these designs and parameter based on given tolerances and the results found. The process follows the rule of evolution: each layout variation is kept, eliminated, modified or combined with an already calculated or new design (2).
Fill Parameter Optimization
The filling process is determined by plunger movements pushing the liquid melt into the die cavity. Melt is poured through a hole into a tube, called a shot sleeve, which is closed on one side by the plunger and on the other by the die. After filling the sleeve to a specific level, the plunger slowly moves forward pushing the melt towards the die. When the melt reaches the casting cavity, the plunger is accelerated to high speed to fill the cavity in 1/10th of a second. The filling speed is critical as filling slower can allow the melt to get too cold while filling faster does not give the trapped air time to escape.Melt that is too cold or contains too much air will reduce the casting quality. Finding the right compromise is an important task.
This 'change and comparison' allows the software to optimize the casting process within the given tolerances, but without having an engineer spending time to go through results and make improvements manually. The engineers have time to concentrate on more important project tasks than simulation set-ups, result evaluation and comparison.
Simulating all possible variations and comparing them would be nice, but based on the high number of possible variations it is, in most cases, impractical. For example, a simulation having only six parameters and five variations would end in 7,776 simulation runs. Assuming an average time of three minutes per simulation, the entire project would take over sixteen days to complete, which is time that most metal casting engineers do not have anymore. Today software uses genetic optimization algorithm. As in the biological world, the evolutionary process of autonomous optimization occurs over several calculation generations. Based on the defined objectives, such as high temperature and/or low air volumes, a generic algorithm creates new variations of the filling parameter. The process is repeated until design modifications do not lead to additional improvements.
Using variations of these fill parameter is an easy task for the optimization software. The engineer simply creates a template including the parameter to change, the step variation and tolerances, and the software takes over. It selects a starting design and changes the parameter according the variations and tolerances for single simulations.
Interpretations of filling results are done automatically based on the melt or die temperature, fill time or air volume values at end of fill. A once tedious and time-consuming, but critical, engineering task has now been relegated to an automatic and more accurate process.
Component Optimization
Getting a new project or the opportunities to change the casting or die design, such as adding fins, more wall stock, overflow locations, gate locations, ingate variations, cooling or heating lines, die materials, etc., calls for action in the engineering group. Selecting the best design out of all the given possibilities is often connected with failure, even when those involved are experienced. Autonomous optimization, on the other hand, opens the possibility to simulate and find the right elements for the casting projects. Simply by designing these single objects and switching them from simulation to simulation is an easy task for the software.
For example, a particular set of components is calculated and the runner model is selected to be replaced by one with more or less branches; without engineering intervention, the software automatically loads the new design and calculates it. Based on the defined objective, the computer will find the best component configuration in a short time. What's left for the engineers to do is to evaluate the best simulation results, then release the optimal design to build the tool.
Optimization with Parametric Objects
The most exciting alternative is to manipulate objects by simply changing numerical values. Instead of drafting and designing multiple runner systems to be selected for the optimization, runner systems can be manipulated using numerical parameters. Ingate areas can be increased or reduced simply by adjusting the parameter for the ingate thickness or width. Balancing a runner system for one or multiple cavity dies is easily done by changing the numerical values for runner length, angle and/or direction. With the given templates provided a runner system can be defined, built, and prepared for optimization in minutes, including variations of multiple runner branches and ingates. Overflow and vent dimensions can be adjusted in size and volume to reduce air entrapment and porosity.
Balancing the thermal profile of a die to reduce die flashing, cold runs and to increase die life, placing cooling lines is crucial, but becomes a simple task when optimization tools are used. Line start and end locations are defined numerical and can be easily changed and dependencies can be readily defined. By changing the line temperature values, cooling and heating can be simulated as necessary.
Summary
Over thousands of years, foundry technology did not change much until the dawn of the computer age. Electronics allow building fully automatic production processes with handling robots and monitoring systems, increasing quality and productivity by reducing hard working labor forces. With autonomous optimization tools, research and development is taken out of the direct production area. Trial and error is not done on the plant floor anymore with production machines wasting time, casting volumes and costs.
In a very short time, optimization has taken over the process of simulation, eliminating manual corrections and the high cost of wasted engineering time. Engineers today define a process with tolerance windows and the objectives the process has to achieve, while optimization will come up with the best results to do so.
1.) From Wikipedia, the free encyclopedia
2.) Casting Designs Become Self-Optimizing
Metal Casting Sep/Oct 2009 Christof Heisser, Magma Foundry Technologies, Inc, Schaumburg, IL
http://www.kktooldie.com/cae-simulation/
Comments
Post a Comment