# Calculating riser size, location via CAD models. (Technology in Progress).

A fundamental problem facing foundries is developing an adequate riser design for feeding a casting. The classical approach to riser size estimation is to calculate the volume and cooling surface area of various parts of the casting and use those measurements to derive the modulus. Regions of the casting which have the lowest modulus values solidify first, and those with the highest modulus values solidify last. In general, the modulus values govern the design of the riser in heavy-section castings; in "rangy" castings (those with large dimensions but thinner wall sections), the volume consideration often is the governing factor.

While these concepts are simple and straightforward, their implementation in casting design is not. This is due to the difficulty in calculating volumes and surface areas for complex, real-world castings. In real-world applications, the approach that has been taken by most foundry engineers is similar to that of weight calculation. The casting is arbitrarily broken into a number of pieces, and each of these pieces is identified as a simple geometric shape for which surface area and volume can be calculated.

In practice, however, this process is cumbersome and inexact; the arbitrariness of approximating a casting shape with a series of simple shapes reduces both repeatability and accuracy. Another intrinsic problem with this method is that it is based only on geometry. It does not directly take into account thermal effects such as specific properties of chilling or insulating materials and heat saturation of cores or various areas of the mold. While some factors have been proposed to correct these effects, such factors can increase the uncertainty surrounding the accuracy of results.

Expanding the Horizons of Casting Simulation

During the past few years, computer simulation of the casting process using accurate 3-D computer models has become increasingly widespread. Such simulations accurately predict the progressive solidification of the casting and its rigging system and the potential for casting defect formation. However, one drawback of casting simulation is that it requires an initial design to simulate.

It is ironic that many foundries, even those using the most advanced simulation tools, still use a traditional approach when developing the initial casting design for simulation. In general, this requires calculating approximate surface areas, volumes and modulus values through manual or software-based methods to break the casting into simple shapes.

Since most new casting designs are available today in 3-D CAD formats from the customer, a more accurate and more automated riser calculation method is achievable. The starting point for this development is the idea that the modulus approach is essentially an attempt to estimate the solidification times of various parts of the casting prior to attaching the rigging. However, using modern casting simulation software, the solidification time of every point within a casting can be calculated quickly without estimation.

These solidification times must be integrated with potential risers of various sizes that might be attached to the casting. At this point, having simulated only the casting with no risers, the solidification time of any arbitrary riser cannot be compared with the casting. Developing a calculation that could convert the solidification times in the casting to equivalent modulus values would allow direct comparison of a riser with the casting, since a modulus value for a riser generally can be calculated easily.

Also, given an array of modulus values within the casting, separate feeding areas within the casting must be recognized to determine how many risers are required and where they should be located. Pattern-recognition software accomplishes this by locating isolated areas of high modulus values--essentially "hot spots" in the casting that must be fed. In some rangy castings, there may be many such areas, so the system can discriminate between small areas that don't need feeding, and larger areas that do.

Once the individual feed areas are identified, an appropriate riser can be sized for each area. Knowing the maximum modulus value and the volume of each feeding area, it is simple to apply known rules to calculate the correct size riser for each area. Also, by plotting the location of the maximum modulus values within each feeding area, it is possible to pinpoint the required attachment point for each riser.

This methodology achieves an almost automatic calculation of required risering for a casting. The starting point is a 3-D model of the casting, which can be transmitted from a CAD system. Then, the alloy and mold material are selected and a simulation with no risers is run. The system then analyzes the results, calculates modulus values and suggests the number and location of required risers. The details of each riser are then provided by calculations that embody riser design rules based on modulus and volume requirements.

Such an approach provides more accuracy than traditional design calculations for several reasons. The accuracy of the 3-D model provides more exact geometry than usually is considered. In addition, by running a simulation, thermal effects such as heat saturation of cores are accounted for explicitly.

Riser Design in Action

As an example of how this approach might be applied, consider the casting model shown in Fig. 1, which has several relatively isolated heavy sections.

Designing the risering for this casting begins with selecting the casting alloy and mold material, in this case, AISI 1030 carbon steel and green sand. After a finite difference analysis mesh is generated, a simulation is run with no risers attached. This simulation produces a plot of solidification time throughout the casting (Fig. 2).

The next step is to apply the formula that converts the simulation data to modulus values. This operation calculates a modulus value for each point within the casting.

After this calculation is performed, the pattern-recognition algorithm is applied to the modulus values so that individual feeding areas within the casting can be identified. The number of suggested risers to produce this casting is displayed (in this case, two).

The maximum modulus and the volume of each of the feeding areas are displayed based on the results of the modulus and pattern calculations. From this, the dimensions of a riser for each area can be determined by entering the height-to-diameter ratio. The riser size calculations take into account the ratio of the riser modulus to the maximum casting modulus to ensure that the riser will solidify later than the casting. They also account for the effect of insulated or exothermic sleeves, which increase the modulus of a riser by blocking heat flow from the riser surface into the mold. The required volume of the riser is estimated from curves published in the Cast Metal Institute's Basic Principles of Gating and Risering that relate the volume ratio of the casting and riser to the modulus ratio. Entering the height-to-diameter ratio allows the software to calculate the height, diameter and actual-vs.-required volume for the riser.

Using the visualization tools of the simulation software, the various feed areas of the casting can be examined and seen clearly. The system has indicated that two distinct feeding areas of the casting exist (Fig. 3).

Further, the areas of highest modulus value can be viewed, so that the correct attachment points for the risers can be located (Fig. 4).

Using this information, the suggested risers were added to the casting model and a verification simulation was run (Fig. 5). This is an important step because the riser size calculations are approximations and cannot take into account all of the complex thermal interactions that occur in a fully rigged casting. The macroporosity prediction (the dark blue area) shows clearly that the volume provided by the risers was sufficient to feed the casting without the formation of internal shrinkage porosity (Fig. 5).

For a free copy of this article circle No. 344 on the Reader Action Card.

For More Information

Directional Solidification of Steel Castings, R. Wlodawer (1966). Available from the AFS Library at 800/537-4237.

About the Author

Lawrence Smiley has been involved with computer modeling of casting processes since 1985 and computer modeling of other heat transfer processes since 1975. He founded Finite Solutions, Inc., in 1993.

While these concepts are simple and straightforward, their implementation in casting design is not. This is due to the difficulty in calculating volumes and surface areas for complex, real-world castings. In real-world applications, the approach that has been taken by most foundry engineers is similar to that of weight calculation. The casting is arbitrarily broken into a number of pieces, and each of these pieces is identified as a simple geometric shape for which surface area and volume can be calculated.

In practice, however, this process is cumbersome and inexact; the arbitrariness of approximating a casting shape with a series of simple shapes reduces both repeatability and accuracy. Another intrinsic problem with this method is that it is based only on geometry. It does not directly take into account thermal effects such as specific properties of chilling or insulating materials and heat saturation of cores or various areas of the mold. While some factors have been proposed to correct these effects, such factors can increase the uncertainty surrounding the accuracy of results.

Expanding the Horizons of Casting Simulation

During the past few years, computer simulation of the casting process using accurate 3-D computer models has become increasingly widespread. Such simulations accurately predict the progressive solidification of the casting and its rigging system and the potential for casting defect formation. However, one drawback of casting simulation is that it requires an initial design to simulate.

It is ironic that many foundries, even those using the most advanced simulation tools, still use a traditional approach when developing the initial casting design for simulation. In general, this requires calculating approximate surface areas, volumes and modulus values through manual or software-based methods to break the casting into simple shapes.

Since most new casting designs are available today in 3-D CAD formats from the customer, a more accurate and more automated riser calculation method is achievable. The starting point for this development is the idea that the modulus approach is essentially an attempt to estimate the solidification times of various parts of the casting prior to attaching the rigging. However, using modern casting simulation software, the solidification time of every point within a casting can be calculated quickly without estimation.

These solidification times must be integrated with potential risers of various sizes that might be attached to the casting. At this point, having simulated only the casting with no risers, the solidification time of any arbitrary riser cannot be compared with the casting. Developing a calculation that could convert the solidification times in the casting to equivalent modulus values would allow direct comparison of a riser with the casting, since a modulus value for a riser generally can be calculated easily.

Also, given an array of modulus values within the casting, separate feeding areas within the casting must be recognized to determine how many risers are required and where they should be located. Pattern-recognition software accomplishes this by locating isolated areas of high modulus values--essentially "hot spots" in the casting that must be fed. In some rangy castings, there may be many such areas, so the system can discriminate between small areas that don't need feeding, and larger areas that do.

Once the individual feed areas are identified, an appropriate riser can be sized for each area. Knowing the maximum modulus value and the volume of each feeding area, it is simple to apply known rules to calculate the correct size riser for each area. Also, by plotting the location of the maximum modulus values within each feeding area, it is possible to pinpoint the required attachment point for each riser.

This methodology achieves an almost automatic calculation of required risering for a casting. The starting point is a 3-D model of the casting, which can be transmitted from a CAD system. Then, the alloy and mold material are selected and a simulation with no risers is run. The system then analyzes the results, calculates modulus values and suggests the number and location of required risers. The details of each riser are then provided by calculations that embody riser design rules based on modulus and volume requirements.

Such an approach provides more accuracy than traditional design calculations for several reasons. The accuracy of the 3-D model provides more exact geometry than usually is considered. In addition, by running a simulation, thermal effects such as heat saturation of cores are accounted for explicitly.

Riser Design in Action

As an example of how this approach might be applied, consider the casting model shown in Fig. 1, which has several relatively isolated heavy sections.

Designing the risering for this casting begins with selecting the casting alloy and mold material, in this case, AISI 1030 carbon steel and green sand. After a finite difference analysis mesh is generated, a simulation is run with no risers attached. This simulation produces a plot of solidification time throughout the casting (Fig. 2).

The next step is to apply the formula that converts the simulation data to modulus values. This operation calculates a modulus value for each point within the casting.

After this calculation is performed, the pattern-recognition algorithm is applied to the modulus values so that individual feeding areas within the casting can be identified. The number of suggested risers to produce this casting is displayed (in this case, two).

The maximum modulus and the volume of each of the feeding areas are displayed based on the results of the modulus and pattern calculations. From this, the dimensions of a riser for each area can be determined by entering the height-to-diameter ratio. The riser size calculations take into account the ratio of the riser modulus to the maximum casting modulus to ensure that the riser will solidify later than the casting. They also account for the effect of insulated or exothermic sleeves, which increase the modulus of a riser by blocking heat flow from the riser surface into the mold. The required volume of the riser is estimated from curves published in the Cast Metal Institute's Basic Principles of Gating and Risering that relate the volume ratio of the casting and riser to the modulus ratio. Entering the height-to-diameter ratio allows the software to calculate the height, diameter and actual-vs.-required volume for the riser.

Using the visualization tools of the simulation software, the various feed areas of the casting can be examined and seen clearly. The system has indicated that two distinct feeding areas of the casting exist (Fig. 3).

Further, the areas of highest modulus value can be viewed, so that the correct attachment points for the risers can be located (Fig. 4).

Using this information, the suggested risers were added to the casting model and a verification simulation was run (Fig. 5). This is an important step because the riser size calculations are approximations and cannot take into account all of the complex thermal interactions that occur in a fully rigged casting. The macroporosity prediction (the dark blue area) shows clearly that the volume provided by the risers was sufficient to feed the casting without the formation of internal shrinkage porosity (Fig. 5).

For a free copy of this article circle No. 344 on the Reader Action Card.

For More Information

Directional Solidification of Steel Castings, R. Wlodawer (1966). Available from the AFS Library at 800/537-4237.

About the Author

Lawrence Smiley has been involved with computer modeling of casting processes since 1985 and computer modeling of other heat transfer processes since 1975. He founded Finite Solutions, Inc., in 1993.

Printer friendly Cite/link Email Feedback | |

Comment: | Calculating riser size, location via CAD models. (Technology in Progress). |
---|---|

Author: | Smiley, Lawrence E. |

Publication: | Modern Casting |

Geographic Code: | 1USA |

Date: | Mar 1, 2002 |

Words: | 1340 |

Previous Article: | Filtering basics: Who, what, where, why & how. |

Next Article: | Casting advancements 'drive' golf industry: Casting technology has propelled the science of golf light-years ahead of the days when hickory shafts... |

Topics: |