Powered by

Expert Article

 Deepak Chowdhary

Deepak Chowdhary

Managing Director, MPM PVT LIMITED

Green Sand Casting - Part 2 Are You Overdosing or Underdosing?

Author / Editor: Deepak Chowdhary / Isabell Page

Most "block" foundries are subject to sand mergers or burn-on problems. The specific causality of this defect continues to annoy most foundries in this casting area with an almost regular periodicity.

Related Vendors

Data analysis can be an essential decision-making aid for foundries for an almost accurate dose-by-demand add-on program.
Data analysis can be an essential decision-making aid for foundries for an almost accurate dose-by-demand add-on program.
(Source: Pixabay / CC0 )

Each time the problem resurfaces, there is a scramble for a variety of reactive analysis. Some of the most common and typical reactive questions are: "Has the supplier changed the Lustrous Carbon additive?" or "Has the bentonite supplier changed his source or has the supplier been changed?". However, not many foundry process owners ask the following questions:

  • Has the GFN (Grain Finess Number) of the raw silica sand changed?
  • Has the GFN of the core sand changed? (since this is normally coarser than the GFN of the raw molding silica sand )
  • Has the production planning for specific type of blocks changed?
  • Is there more or less quantity or percentage of core sand infiltration?

Why are these questions of relevance to the incidence of sand fusion? The answer lies in a very basic study of Total Surface Area (TSA) done on the return/shake-out sand of specific block castings having varying core sand loads (see Picture 1).


Raw Core Sand GFN is normally coarser than that of the Raw Molding Sand. Figure 1 in the picture gallery shows that as the core sand influx increases, the TSA decreases and vice versa. Each Marker represents the return sand of a specific pattern. Because each pattern return sand will form a layer as it re-circulates into a storage hopper. The TSA of sand layer that will come through for the next molding batch mix can only be estimated The motivation is to examine the variability of the TSA of the sand before it goes into the mixer. The variability will decide how much make-up additives are required to achieve target properties. In absence of data analytics even variable additions can at best be approximate. Under-dosing or overdosing of the additives will also therefore, decide the rejection legacy.

Many foundries try and develop a return sand mapping logic. Many don’t. In both cases however, the additive dosing is often prophylactic or 'Fixed Dose’. In the absence of an advanced data analytic decision support system, which also takes into account among other complex logic, the complete sand mass balance; there is inevitably a limitation to the exact estimation.

If there is no data analytic decision support available, it is still a better idea to map the return sand and have varying additive dosage program rather than having a single fixed dose additive program. However, if there is in place data analytics installed, then the foundry can actually get decision support on a near precise dose-by-need additive program. The advantages of data analytics decision support is seen in Figure 2 and 3 in the picture gallery.

It’s not only the optimisation of the additive consumption that is of emphasis here, there is also the collateral benefit of reduced sand related defects when a data analytic decision support is followed. The motivation of determining TSA or graduating to data analytical decision support software is a variable “dose-by-need”addition advantage of prescriptive analytics:

  • To graduate from reactive decision making in targeting optimal sand properties in terms of controlling related rejections
  • To move towards process consistency reducing systems ‘noise’ by predicting and then optimizing the optimal target sand parameters
  • To move towards ‘dose-by-need’, variable additions of additives
  • Prevent overdosing and under-dosing the system sand
  • Thereby control, mitigate or reduce rejections

Given the molding process dynamics and the mass balance logic of the total sand loop, defects are inevitable. However, some ‘finger-printing’ to predict possible causality is possible by estimating the Total Surface Area of the Return sand and then its priority of discharge of each layer to the next mix. It's simple to state, but in a dynamic sand loop it can be practically very complex to predict by human intervention alone due to the shear multivariate nature of the molding process. The advantages of a data analytics decision support is priceless in this context.

Subscribe to the newsletter now

Don't Miss out on Our Best Content

By clicking on „Subscribe to Newsletter“ I agree to the processing and use of my data according to the consent form (please expand for details) and accept the Terms of Use. For more information, please see our Privacy Policy.

Unfold for details of your consent