Sand reclamation has become a necessity in no-bake operations, particularly given the current challenges regarding the cost and availability of new sand. Foundries can no longer afford to operate with new sand alone and most are exploring how best to economically integrate sand reclamation systems into their processes.
For a reclamation plant to realise the full value of a sand reclamation system, it must operate at optimal efficiency. Without the careful monitoring of outputs and control parameters in place, the foundry could risk an unacceptably high level of casting defects as well as lost production – and thus revenue.
What is ‘dry reclamation’
There are three main types of reclamation systems: Wet, thermal, and mechanical. For the purpose of this editorial, the focus is on mechanical reclamation of silica sand, specifically for ester-cured alkaline phenolic systems.
Dry reclamation is fast emerging as the most accepted way to reclaim no-bake systems. There are nearly as many dry reclamation processes as there are in¬dividual manufacturers of dry reclamation equipment. Operating and installation costs vary widely, depending on what equipment is used and the ton per hour reclaimed.
The most popular techniques include the use of dry pans, jaw crushers, pneumatic scrubbers, vibratory screens, shot blasts, and fluidised beds, and many combinations thereof. Crushing and scrubbing operations are generally followed by an exhaust step to remove fines. Then a mechanical device is used to separate large particles, metallics, and various tramp materials.
A word of caution: Although reclamation equipment may appear dusty or dirty, it is not a garbage disposal; it functions as a piece of processing equipment. Unfortunately, many foundry workers regard the sand heap as a convenient location to discard all sorts of trash. Reclaimers are not designed to remove garbage and this debris can result in clogged sand orifices, equipment malfunctions, and casting defects.
Factors to consider:
Quite simply, the lower the resin content, the easier it is to reclaim the sand. This is because the more resin there is to remove, the more difficult it is to clean the sand. Operating with the lowest binder percentage (that still assures breakage-free core and mould handling) results in minimum material costs, maximum reclaimer efficiency, and the best possible casting.
Sand to metal ratio
The more binder burned off the sand by the casting operation, the easier it will be for the re- claimer to operate at maximum efficiency. Sand is an excellent thermal insulator. It transfers heat at a surprisingly slow rate and establishes very wide thermal gradients in the resin-bonded sand core or mould. As a result, heat from the metal rarely penetrates the sand to decompose the binder more than a couple of centimetres from the mould-metal surface. It is not unusual for sand less than 50mm away from a heavy casting section to be completely intact at shakeout.
While a sand-to-metal ratio of about 2.7:1 is considered optimum, this might not always be achievable. Ideally, foundries should aim to use as little chemically bonded sand as possible to produce the casting.
High pouring temperatures (excessive superheat), types of metals that are slow to solidify, thick casting sections, and configurations that naturally produce a low sand to metal ratio all decompose the binder to a greater extent – thus assisting in the reclamation process.
Transverse and tensile testing
Type of sand
Certain sands reclaim better than others. For example, resin coatings are stripped more easily from round-grain sands than from angular grains. However, the scrubbing and grinding inherent to the reclamation process tends to round off sharper types of sands (e.g. angular silica, olivine, and chromite). The newly created, less angular grains have a reduced surface area – that in turn requires less binder.
As indicated in Table 1, fine grains (i.e. typically those smaller than 140 mesh) have far more grains per kilogram and will have a much larger surface area than a kilogram of coarse particles. This means additional resin will be required. By keeping fines to a minimum, foundries can reduce the possibility of gas defects, which are the result of a combination of issues including reduced permeability, increased resin content, and higher Loss of Ignition (LOI).
Most additives, including iron oxide, kaoline, clay, zircon, and olivine flours, serve a refractory function. But as they help break the resin coating away from the grain, they produce a dusty material, high in binder content as a result. This must also be removed, increasing the load on the reclaimer’s dust collector.
Type of binder systems
In order to examine how specific binder properties influence reclamation capabilities, it is useful to classify no-bake binder systems according to their chemical characteristics. The two main systems are acid-catalysed furans and ester-cured phenolics and phenolic urethanes (e.g. cold box and PepSet).
1. Acid-catalysed furans and phenolics
How easily the acid-catalysed furans and phenolics are to reclaim depends primarily on how much resin is used to bond the sand and how much resin is burned away by casting operations.
Experimental data and experience indicate that both furans and phenolics are quite thermally stable. Since phenolics tend to resist decomposition more than furans, they should be more difficult to reclaim. However, the phenolic coating is more brittle, so both the intact and partially burned binder chips away from the grains easier than the furans. The net result is that furans and phenolics dry reclaim equally well.
Phenolic-urethane resin systems are diluted (extended) with a solvent, which lowers viscosity. When used at their recommended levels, only a very minimal coating is placed on the sand grains. Since some of the solvent evaporates after coating, this reduces the already low percentage of resin that needs to be removed. This type of coating is also rather brittle, so it chips away easily from the sand grain surface. The combination of the extremely low binder content and the brittleness of the coating makes this group of noble resins extremely easy to reclaim via mechanical techniques.
Silica sand mixed with other sands
When chromite sand is reclaimed (or is present in the reclaimed silica sand system) it requires special attention.
Chromite sand is normally used in combination with silica sand to produce surface chill or to promote directional solidification. During the recycling steps of the reclamation process, the chromite and silica become thoroughly blended, unless some separation device is used.
The amount of chromite present in the silica system varies. Anything between <1% to more than 50% chromite are the normally encountered extremes found in chromite-silica mixtures. Research in South Africa several years ago reported that when fine particles of silica are mixed with chromite sand at silica levels from 1% to 7%, the refractoriness of the blend is lowered considerably and the defect of sand burn-in occurs on the casting surface. Fortunately, since chromite is the smaller portion of the usual system, the opposite situation (i.e. with low concentrations of fine particle silica) is unlikely to cause the silica-chromite
Elemental Iron (Fe) accounts for about 20% by weight of the chemical composition of foundry-type chromite sand. At elevated temperatures, and in an oxygen containing atmosphere, Fe oxides and chromite sand increase in weight. Because of this increase in weight during the LOI test, standard LOI testing cannot be run on chromite sand or chromite-silica mixtures. This change partially makes up for some of the weight loss because of the binder being burned off.
One final aspect of reclaiming mixtures of silica with chromite, zircon, or olivine is that these minerals are denser than the silica sand. This often results in the sand’s fines being more difficult to remove from the mass by standard exhaust techniques. The density of the mixture of non-silica minerals and silica sand goes up as the proportion of the non-silica sand to silica increases. This greater density mixture generally requires a lower resin percentage by weight to completely coat the grains.
What you should be measuring
Even though each foundry is unique in its operation, the types of castings used, and mould making procedures, the fundamentals are still the same. It is imperative that each foundry keeps track and records its unique reclaimed sand specifications. The quality of any new sand also plays an important role as ‘garbage in’ definitely equals ‘garbage out’.
The following parameters are all factors to consider during the reclamation process as they each impact on the reclaimed sand’s suitability in the production process.
Return sand temperature: Heat from the molten metal and the reclamation process itself increases sand temperature. Before sand can be coated again, this must be lowered to a reasonable level. Any temperature in excess of 35°C complicates the recoating operation significantly. As the temperature of the sand increases above the ideal operating temperature (as listed in Table 2) these problems become more and more serious. Many people will not accept a recoating temperature higher than 10°C degrees above ambient (or average room temperature).
Screen analysis: The screen distribution for reclaimed sand will usually show that the very coarse and the very fine particles have been removed. Ideally a plot of the screen distribution should appear similar to that shown in Figure 1, when compared to the original sand. Fines (grains less than 140 mesh) should be held to less than 1%. Agglomerated or compound grains (found on the 40 mesh or above screen) should be less than 3%.
Loss of Ignition (LOI): To determine loss of ignition, a small sample of sand (generally weighing 5-10g), is heated to between 870°C and 980°C for one hour. The combustible material that has been burned away represents the ‘loss on ignition’ or LOI. The lower the LOI the better. Just how low the LOI must be depends upon the type of metal being cast; less than 2.5% is usually satisfactory for iron while 2% or less is preferred for non-ferrous and steel. The larger the size of the casting, the lower the LOI must be in order to guard against excessive gas evolution.
pH: The pH (negative log of the hydrogen ion content) is a kind of chemical shorthand that uses numbers to express the acid, neutral, or basic state of a material. pH also indicates just how acidic or basic (alkaline) it is. As shown in Table 3 a pH of 7 is neutral. Less than 7 is acid: The lower the number the more acidic. More than 7 is basic or alkaline: The higher the number the more basic.
With the advent of reclamation, the term pH has gained special significance. This is because the surface chemistry of sand is drastically altered by the amount of the resin-catalyst coating remaining on sand grains after the reclamation process. These residual acid or basic components of the binder system now dictate the surface chemistry of the sand and must therefore be measured.
Determining the pH of sand, new or reclaimed, is a straightforward process. By placing a known amount of sand at a controlled temperature in deionised water the pH can be measured with a pH meter or a piece of pH paper. The presence of silicates, oil urethanes, and phenolic urethanes will render reclaimed sand basic, while furans and phenolics cause the reclaimed sand to be acidic.
Tensile strength: This is affected dramatically by the amount of high surface area carbon left on the sand after it has been exposed to the elevated temperature in the metal casting process. Tensile strength is also affected by the screen distribution, the grain shape of the sand (which is made rounder with each pass through the reclaimer), and many other factors. Tensile strength should not vary by more than 12% once the system is running smoothly.
The oft-asked question of ‘how much tensile strength’ generally depends on the strength necessary to handle the core or mould and withstand the Ferro static pressures necessary to produce acceptable castings. In ferrous castings, the acceptable minimum tensile strength is usually about 140 psi. However, this numerical value will be unique to each operation, machine, and operator.
Point impact penetration test, scratch hardness, and heat distortion tests: There are many tests that offer insight into how well the coated sand will hold up to the casting process. Tensile strength is the most common qualifier, as many other tests will correlate with tensile strength.
However, if the time is taken to run a proper battery of tests, they will generate a great deal of intelligence regarding the type of casting that can be obtained with the sand being tested and about the sand system itself.
Magnetics: The magnetic content for ferrous-type metals is determined by stirring a magnet through a known weight of dried sand and weighing the amount of material removed. A typical reclaimed sand magnetic value is about 0/2%. The maximum value should generally run about 25%, Good cores, moulds and castings have been achieved with a magnetic level of >10%.
Microscopic analysis: A great deal of information can be obtained by frequent microscopic examination of the sands in use – coated, raw, and reclaimed sand. A 40 to 100 power microscope is probably the best choice. Unfortunately, becoming familiar with what ‘good’ sand looks like requires a great deal of time before one is experienced enough to conduct a visual assessment of sand quality with confidence.
Reclaiming for quality and cost control
It’s important to treat reclamation as an on-going process. Things can – and do – go awry quickly. This makes monitoring and testing of reclaimed sand crucial to ensuring consistent quality. These tests also enable individual foundries to establish a unique ‘fingerprint’ of the reclaimed sand that works best for their specific operations.
For many foundries, the principal objection to any no-bake system is the comparatively high cost of coated sand. Reclamation can offer significant savings, especially when implementing a combination of the practices already suggested.
ChemSystems is well ahead of the curve when it comes to resin system stability. Its alkaline phenolic system is exceptionally robust when it comes to fluctuating sand conditions in terms of LOI.
ChemSystems also understands and actively promotes the value that comprehensive sand testing and monitoring can deliver to local foundries. It currently provides a comprehensive sand service to its customer base, conducted in its state-of-the-art sand lab. Here, all relevant tests can be performed as well as development for foundries needing assistance on different TBC additions, tensile, transverse, AFS, LOI, clay, work time, strip time, bend tests, cold-box core tests, conductivity tests and dog bones.
For further information contact ChemSystems on TEL: 011 922 1600 or visit www.chemsystems.co.za