In the discussion of molten metal filters several issues must be resolved. First is the mechanism of filtration. There are three types of filtration mentioned in literature and in our understanding.

The first is sieve technology. This is as obviously as its name. It is a simple accounting of the diameter of particulate not being able to pass through the diameter of the holes doing the filtering.

The second technique is called filter cake. This is where a large particle is captured at the surface of the filter and then as a result of chemical actions adsorbs or attracts other particles of smaller size much the same as you would see with a filter cake.

One must consider the morphology of the particles being captured. For example, Hafnium Oxide has a ‘stringer’ morphology. Thus, one can use a cell size larger than the particle because the stringer will create a net on the surface of the filter to trap finer contamination.

The third technique is of the most interest to us. This is a boundary layer or stagnant layer that occurs above the filter. This layer exists between the region of turbulent flow of the metal being poured and the laminar flow created in the filter. The mechanism while not totally understood is well documented. Post mortem polished cross-sections clearly indicate this layer. It is believed that this stagnant layer captures small particulate holding it in place while the vast majority of the metal passes directly through in a laminar fashion into the molds. This laminar flow versus turbulent flow also will also reduce casting shrink and wall splatter inclusions.

A Honeycomb filter inserted in a pour cup will significantly reduce the macro shrink in the pour cup that pipes down into the casting runner system. This may eliminate ceramic contamination from a cutoff wheel and steel shot from the cleaning process that may get trapped in the shrink contaminating the revert. Placing a honeycomb filter in the tundish at the top of a master metal ingot will have the same effect on the shrink pipe in the bar.

One of the issues with foam filters and the often touted tortuous path, is back pressure. As a result, there are two issues that occur far too frequently. One is freeze off as the back pressure is too great and the other is bypass where the metal floats the foam filter and simply flows around it. These are not a known issue with the cellular honeycomb.

Hafnium Oxide contained in many DS alloys has a chemical affinity for aluminum oxide.  This has been an accepted truth in investment casting for many years. Alumina balls could be put into the pour cup to grab the Hafnium Oxide. It works, but it is very expensive.  Foundries had scrap out a pound of DS alloy with every pour. Replacing alumina balls with a Versagrid ceramic filter would be a significant cost savings.

One could also use a ceramic filter to hold the alumina balls out of the pour cup metal. The molten metal would pass over the aluminum balls, but would not be contained in the pour cup revert.

Sizing of the filter is as much art as science. The flow required is determined by the alloy, the temperature, the casting size and the specific casting process. As the manufacturer here is Atlanta, the filter diameter (square and rectangular are also available), thickness and size of the cells (cpsi) are all adjustable and can be readily sampled for trials. The material we use is a 30 year development effort to achieve a high strength, thermal shock resistant body that has many years of success and a multitude of satisfied customers.

The ceramic honeycomb filters create laminar flow which is superior to turbulent flow. When the molten liquid metal passes through the ceramic honeycomb channel, the debris, the slag and other inclusions are removed. Using a ceramic honeycomb filter improves the quality of molten metal flow. A stable and fast laminar molten metal flow prevents air from getting into the molten metal flow, molten metal getting oxidized and splashing out.

Comparable foam filters have a wide variation in the amount of time it takes liquid to flow through them. Although the result of the designed random pattern creates the “torturous path”, it increases the time variability and the “splash” effect. Even from filter to filter in the same exact size and PPI the flow can vary considerably (+/- seconds), even in the same filter from pour to pour. This can create problems when trying to control the solidification of the crystals during casting.

The honeycomb filters have very consistent flow as well as generating laminar flow. Having more laminar flow prevents splashing and void inclusions in the castings.