Frequently Asked Questions
What is the objective in die design?
For flat dies, one usually wants to produce a film, sheet or coating of uniform thickness across its entire width. To do this, it is necessary for the die to produce a uniform volumetric flow rate across its exit. Although this is almost always the primary objective in flat die design, often there are secondary ones. For materials which can degrade (such as PVC), a secondary objective could be to eliminate any re-circulation or "dead zones" in the die. For other situations it may be to ensure that a maximum shear rate or shear stress is not exceed anywhere in the die in order to avoid "melt fracture", "shark skins" or other similar instability. For foam applications, the pressure profile is also important to ensure that "pre-foaming" does not occur before exiting the die.
For profile dies, usually the primary objective is to have a uniform average velocity at the exit of the die along the entire profile. Such dies are said to be "balanced". As with flat dies, there can secondary objectives as well.
What is involved in designing dies?
The performance of a particular die depends on the rheology of the material passing through it, its density and flow rate as well as the temperature of the die. In the past, arriving at an optimal design for a given situation required designing a die, machining it, performing trials to determine its performance, redesigning the die, and repeating the procedure until a satisfactory design was found. This can be an expensive and time-consuming procedure because of the many times this trial and error process has to be repeated.
The purpose of Dieflow is to eliminate the cost and time involved in die design by replacing the "cut and try" procure discussed above with highly accurate and dependable computer trials. Dieflow uses fully 3-D finite element flow algorithms to solve for flow everywhere in the die. The flow in the die is examined and the die design changed until one satisfying the design criteria of the die is obtained.
How is Dieflow different from other 3-D CFD (computational fluid dynamics) codes?
Most commercial CFD codes are "general purpose" codes, which means the same code is marketed to a wide range of industries to solve a wide range of flow problems. As a result they do not solve any one problem very efficiently. They also do not include efficient "mesh generators" so, as a result, one often has to spend days or even weeks just generating a single mesh. Because of this, other codes are not practical for design or optimization problems where the geometry (i.e. the mesh) must be changed over and over again. Dieflow is designed for that.
Dieflow contains only the necessary components needed to solve continuous extrusion problems in dies. As a result, it is much more efficient and faster than the other 3-D "general purpose" codes. We also include automatic mesh generators with the program depending on the style of die. These programs read a table of numbers (i.e. various dimensions of the die) and create a 3-D finite element mesh in seconds containing hundreds or thousands of nodes and elements. Again, since design of a die involves optimizing or changing geometry many times, fast mesh generation is a must. No other code comes with this capability.
How are dies for foams different from those for non-foaming materials?
In non-foaming dies, the pressure profile through the die is not important. But in foam dies it is. As a result, the internal geometry of a foam die is quite different. Foam dies are much more open before reaching the exit (lips) of the die where most of the pressure drop occurs.
Designing both non-foaming and foaming dies require an accurate rheology model (i.e. viscosity vs. shear rate) for the polymer or gel at the temperature at which the die will operate. Although, determining the rheology of foam gels can be much more involved than that for non-foaming polymers, the die design procedure is otherwise much the same.
What about the effect of die swell which occurs downstream of the die?
In the design of flat dies, it is usually assumed that a uniform flow across the exit of the die will produce an extrudate with a uniform thickness downstream. Die swell downstream of the die will influence the thickness of the flat extrudate but its effect on extrudate shape is usually limited to the extreme edges. Since the edges of a flat extrudate are usually trimmed and changing the speed of the line can control thickness, the effects of die swell are usually ignored in the design of flat dies. As a result, simulations for flat dies usually end at the exit of the die.
However, this is not necessarily the case with profile dies. Because of the effects of die swell, the final shape of an extrudate is often substantially different from that of the exit opening of the die. As a result, the design of profile dies producing complex shapes often involves more than just "balancing" the die but also involves compensating for die swell as well. In this case, one must calculate the shape of the extrudate downstream of the die, the flow within the extrudate AND the flow in the die.
A version of Dieflow with this capability is also available. It has the ability not only to determine the shape of the extrudate produced downstream of a die with a given shape but also can determine what the shape of the exit opening of the die should be in order to produce a profile with the specified shape downstream.
How practical is using 3-D CFD to design or optimize extremely complex profile dies?
The goal in profile die design is usually to produce a "balance" die, i.e. one that has a uniform average velocity at the exit of the die along the entire profile. Arriving at a satisfactory design requires designing a die, machining it, performing trials to determine its performance, redesigning (i.e. re machining) the die, and repeating the procedure until a satisfactory design is found.
The re-machining step is usually accomplished by one of three methods:
Because of the number of "cut and try" cycles that can be involved in the process, designing profile dies can take many days if not weeks. Dieflow uses these same methods (as well as combinations of them) to design and optimize profile dies. However, the "machining" of the die and the "trials" are replaced by fast and highly accurate 3-D computational simulations. Using these tools requires the repeated steps of making changes in the internal geometry of the die, recreating a 3-D finite element mesh of the internal flow volume incorporating these changes, solving the 3-D flow equations for the flow throughout the die and comparing the flow results against the design objective(s) of the die. The problem in using 3-D CFD to design complex dies is NOT in the die flow simulation but in the speed of creating (and re-creating) the 3-D finite element mesh on which the flow equations are solved. Without fast mesh generation capabilities, creating 3-D meshes on which flow equations are solved simply takes too long (also many days if not weeks) and cost too much. There is no program available anywhere which can be used to quickly generate a 3-D finite element mesh for every style of profile die. Each style of die requires its own mesh generator. Fortunately, there are relatively few die styles used in practice for the vast majority of flat dies (coat hanger, fish tail, end fed, etc.). Dieflow has written fast "mesh generators" for each of these die styles. We also have similar mesh generators for many other styles of dies including flat "dog bone" dies, some annular dies as well as a limited number of profile dies. While some other CFD programs available commercially may included some mesh generation capability, none have Dieflow's fast mesh generation for extrusion dies.
However, even Dieflow's mesh generating capabilities are not unlimited. For styles of dies in which no mesh generator presently exists, one must be written. This is the case most of the time since nearly every shape that profile extruders produce is new and unique.
There are three ways that a finite element mesh for these dies can be created:
In summary, with the mesh generation capabilities currently available, Dieflow usually recommends that if a satisfactory profile die can be achieved by traditional "cut and try" methods within in a few days, the use of computational methods is probably not justified. However, in those cases where arriving at a successful design may take weeks or longer, the use of 3-D CFD tools should be considered.
For flat dies, one usually wants to produce a film, sheet or coating of uniform thickness across its entire width. To do this, it is necessary for the die to produce a uniform volumetric flow rate across its exit. Although this is almost always the primary objective in flat die design, often there are secondary ones. For materials which can degrade (such as PVC), a secondary objective could be to eliminate any re-circulation or "dead zones" in the die. For other situations it may be to ensure that a maximum shear rate or shear stress is not exceed anywhere in the die in order to avoid "melt fracture", "shark skins" or other similar instability. For foam applications, the pressure profile is also important to ensure that "pre-foaming" does not occur before exiting the die.
For profile dies, usually the primary objective is to have a uniform average velocity at the exit of the die along the entire profile. Such dies are said to be "balanced". As with flat dies, there can secondary objectives as well.
What is involved in designing dies?
The performance of a particular die depends on the rheology of the material passing through it, its density and flow rate as well as the temperature of the die. In the past, arriving at an optimal design for a given situation required designing a die, machining it, performing trials to determine its performance, redesigning the die, and repeating the procedure until a satisfactory design was found. This can be an expensive and time-consuming procedure because of the many times this trial and error process has to be repeated.
The purpose of Dieflow is to eliminate the cost and time involved in die design by replacing the "cut and try" procure discussed above with highly accurate and dependable computer trials. Dieflow uses fully 3-D finite element flow algorithms to solve for flow everywhere in the die. The flow in the die is examined and the die design changed until one satisfying the design criteria of the die is obtained.
How is Dieflow different from other 3-D CFD (computational fluid dynamics) codes?
Most commercial CFD codes are "general purpose" codes, which means the same code is marketed to a wide range of industries to solve a wide range of flow problems. As a result they do not solve any one problem very efficiently. They also do not include efficient "mesh generators" so, as a result, one often has to spend days or even weeks just generating a single mesh. Because of this, other codes are not practical for design or optimization problems where the geometry (i.e. the mesh) must be changed over and over again. Dieflow is designed for that.
Dieflow contains only the necessary components needed to solve continuous extrusion problems in dies. As a result, it is much more efficient and faster than the other 3-D "general purpose" codes. We also include automatic mesh generators with the program depending on the style of die. These programs read a table of numbers (i.e. various dimensions of the die) and create a 3-D finite element mesh in seconds containing hundreds or thousands of nodes and elements. Again, since design of a die involves optimizing or changing geometry many times, fast mesh generation is a must. No other code comes with this capability.
How are dies for foams different from those for non-foaming materials?
In non-foaming dies, the pressure profile through the die is not important. But in foam dies it is. As a result, the internal geometry of a foam die is quite different. Foam dies are much more open before reaching the exit (lips) of the die where most of the pressure drop occurs.
Designing both non-foaming and foaming dies require an accurate rheology model (i.e. viscosity vs. shear rate) for the polymer or gel at the temperature at which the die will operate. Although, determining the rheology of foam gels can be much more involved than that for non-foaming polymers, the die design procedure is otherwise much the same.
What about the effect of die swell which occurs downstream of the die?
In the design of flat dies, it is usually assumed that a uniform flow across the exit of the die will produce an extrudate with a uniform thickness downstream. Die swell downstream of the die will influence the thickness of the flat extrudate but its effect on extrudate shape is usually limited to the extreme edges. Since the edges of a flat extrudate are usually trimmed and changing the speed of the line can control thickness, the effects of die swell are usually ignored in the design of flat dies. As a result, simulations for flat dies usually end at the exit of the die.
However, this is not necessarily the case with profile dies. Because of the effects of die swell, the final shape of an extrudate is often substantially different from that of the exit opening of the die. As a result, the design of profile dies producing complex shapes often involves more than just "balancing" the die but also involves compensating for die swell as well. In this case, one must calculate the shape of the extrudate downstream of the die, the flow within the extrudate AND the flow in the die.
A version of Dieflow with this capability is also available. It has the ability not only to determine the shape of the extrudate produced downstream of a die with a given shape but also can determine what the shape of the exit opening of the die should be in order to produce a profile with the specified shape downstream.
How practical is using 3-D CFD to design or optimize extremely complex profile dies?
The goal in profile die design is usually to produce a "balance" die, i.e. one that has a uniform average velocity at the exit of the die along the entire profile. Arriving at a satisfactory design requires designing a die, machining it, performing trials to determine its performance, redesigning (i.e. re machining) the die, and repeating the procedure until a satisfactory design is found.
The re-machining step is usually accomplished by one of three methods:
- Varying the land length along the profile
- Changing the shape of the transition region '"feeding" the land, or
- Changing the shape of the profile at the exit of the die to compensate for the effect of die swell downstream.
Because of the number of "cut and try" cycles that can be involved in the process, designing profile dies can take many days if not weeks. Dieflow uses these same methods (as well as combinations of them) to design and optimize profile dies. However, the "machining" of the die and the "trials" are replaced by fast and highly accurate 3-D computational simulations. Using these tools requires the repeated steps of making changes in the internal geometry of the die, recreating a 3-D finite element mesh of the internal flow volume incorporating these changes, solving the 3-D flow equations for the flow throughout the die and comparing the flow results against the design objective(s) of the die. The problem in using 3-D CFD to design complex dies is NOT in the die flow simulation but in the speed of creating (and re-creating) the 3-D finite element mesh on which the flow equations are solved. Without fast mesh generation capabilities, creating 3-D meshes on which flow equations are solved simply takes too long (also many days if not weeks) and cost too much. There is no program available anywhere which can be used to quickly generate a 3-D finite element mesh for every style of profile die. Each style of die requires its own mesh generator. Fortunately, there are relatively few die styles used in practice for the vast majority of flat dies (coat hanger, fish tail, end fed, etc.). Dieflow has written fast "mesh generators" for each of these die styles. We also have similar mesh generators for many other styles of dies including flat "dog bone" dies, some annular dies as well as a limited number of profile dies. While some other CFD programs available commercially may included some mesh generation capability, none have Dieflow's fast mesh generation for extrusion dies.
However, even Dieflow's mesh generating capabilities are not unlimited. For styles of dies in which no mesh generator presently exists, one must be written. This is the case most of the time since nearly every shape that profile extruders produce is new and unique.
There are three ways that a finite element mesh for these dies can be created:
- We can write an another automatic mesh generator which reads a table of numbers and quickly creates a FE mesh. This is only practical for dies whose internal geometry can be described by a relatively small number (i.e., 20-30) of parameters. Therefore, this is not recommended for complex profile dies.
- For profile dies that are made up of plates that are cut with a wire EDM between a limited number ( i.e., 5-7) of 2-D profiles shapes, Dieflow can read these 2-D profiles, add internal points and then create the volume (and mesh) between the profiles by a linear interpolation (much like a wire EDM does.) To do this, each of the profiles should be entirely represented by a "polyline" in a CAD *.dxf file. If not, adding the extra points "by hand" soon becomes time consuming and again impractical for more complex profiles.
- If a CAD drawing of the complete die exists in the form of a true 3-D IGES file, there are a few commercial software programs available that claim to be able to create a finite element mesh of the internal flow volume of the die directly from the IGES file. Unfortunately, these programs are fairly expensive and for the most part untested However, Dieflow sees the use of such methods as necessary for extremely complex profile dies in order to make the replacement of "cut and try" trials by computational methods practical . As far as we know, Dieflow is the only company exploring this option. However that capability is not available.
In summary, with the mesh generation capabilities currently available, Dieflow usually recommends that if a satisfactory profile die can be achieved by traditional "cut and try" methods within in a few days, the use of computational methods is probably not justified. However, in those cases where arriving at a successful design may take weeks or longer, the use of 3-D CFD tools should be considered.