A moist system whose components must work together
In green-sand moulding, sand forms the granular matrix, while bentonite and water contribute to the bonding system. “Green” describes the moulding condition rather than grain colour. The useful question is how to prepare a mixture able to fill the mould, retain its geometry and function in the manufacturing process concerned.
The institutional profile of Béni Mellal-Khénifra describes a region combining agricultural and food-processing activities with mineral resources 1. It does not identify foundry-sand consumption. This dossier therefore discusses a method applicable to a moulding project in the region, without attributing undocumented production or a foundry population to it.
Monitoring must concern the prepared mixture and mould. A sound virgin-sand sheet is insufficient if returns vary, bentonite is poorly distributed or water is added under different conditions. Materials, operations and results should be followed as one system, with clearly identified sampling locations.
1. Identify the granular matrix before additions
Retain the complete distribution of virgin and circulating sand. A fineness index alone does not show every proportion across size classes. Identify fines, coarse grains and any mixtures with core sands. Material actually sent to the mixer may differ from the virgin material described on its certificate.
Also request useful preparation information: washing, drying, storage and lots. These elements do not demonstrate mould performance by themselves, but make comparisons possible. Two preparations described merely as “the same sand” may actually involve different distributions or stock whose fines have changed.
Do not change grading to correct an anomaly before checking the other parameters. Modifying the matrix produces a new combination with binder and water. Our silica-sand page provides an introduction to general aggregate properties; here, the purpose is to organise interpretation of the moulding system.
2. Distinguish added bentonite from still-active bentonite
The mass of bentonite introduced into the mixer does not automatically represent all bonding capacity in the circulating sand. Returns carry their own thermal history and constituents. Distinguish fresh additions, existing fines and the binder fraction still functioning in the system.
Mrzygłód and colleagues describe the value of several prepared-sand properties and mention determining active bentonite to control additions 4. This distinction avoids treating a mass of fines as if it necessarily had the same role as fresh binder.
Select appropriate tests with the laboratory and retain their units, methods and conditions. When comparing a measurement with historical results, check that it is obtained in the same way. Changing the method can shift results without an equivalent change in material. The control sheet should make this change visible.
3. Understand why water is not merely a correction
Water contributes to activation of the clay system and to preparation. Simpson explains that its behaviour also depends on temperature, fines, additions and mulling 2. Adding the same amount of water therefore does not necessarily produce the same sand condition when return-stream conditions change.
Separate water already present in material from added water. Record masses or volumes on a defined basis. A control-panel setting does not replace checking the dosing device. When moisture is expressed as a percentage, identify the convention used so results and recipes can be compared correctly.
A sound decision restores the qualified mixture conditions rather than automatically increasing water when a mould appears fragile. An immediate correction can improve one observation while altering other properties. Connect the action with a measured result and monitor its effects after preparation, then at the moulding station.
4. Give mulling its place in qualification
Mulling is more than visually mixing materials. It must make preparation reproducible with the equipment concerned. Retain charge mass, addition order, times, settings and the condition of mixing components. Modifying these creates a different preparation even when the weighed recipe remains identical.
Do not transfer results from a small laboratory mixer directly to an industrial line. The laboratory compares materials under controlled conditions. The line must then confirm behaviour with its own preparation time, transfers and waiting conditions. The two levels complement each other.
When preparation drifts, also examine maintenance operations and changes in production rate. Less available time, irregular dosing or a worn component can modify the result. Virgin sand should not become the default explanation for a change originating in the preparation process.
5. Measure compactability using a consistent method
Compactability describes the reduction of a sand volume under a defined action. Simpson emphasises sample preparation, the method and equipment checks 3. A value must be compared with results obtained using a compatible protocol.
Specify where and when sand is sampled. A measurement at mixer discharge and one after waiting do not necessarily concern the same material condition. Retain the delay before testing and storage conditions for the portion. This avoids confusing evolution during waiting with a preparation error.
The target value belongs to a qualified system. It should not be extracted from a general range and imposed on every installation. Selection relies on moulding results, other tests and casting constraints. An indicator becomes useful when it triggers a documented action, rather than merely filling a box.
6. Read strength, permeability and friability together
The laboratory can examine several complementary properties. Simpson’s technical catalogue distinguishes strength, permeability and friability equipment 5. These measurements answer different questions. One improvement does not demonstrate an overall mould improvement.
Build a table connecting each measurement with its role in the process. Strength concerns a defined stress on the specimen; permeability describes gas passage under test conditions; friability examines surface integrity according to its method. Add specimen-manufacturing conditions so results remain interpretable.
Keep the preparation procedure consistent when comparing two formulations. Different compaction can modify several measurements simultaneously. If the purpose is to study compaction, identify it as a trial-plan variable. The intended effect must be distinguished from variation introduced by measurement itself.
7. Avoid contradictory corrections
A line can receive conflicting instructions when each department watches one indicator. One increases water to correct compactability; another increases an addition to restore strength; a third changes sand. Without a decision sequence, the actions make the system less readable and complicate cause identification.
Define a checking priority: sampling and method, charge mass, dosing, return material, then preparation and the parameters being studied. This sequence organises diagnosis rather than establishing a universal law of defects. It helps eliminate simple errors before modifying the qualified formulation.
After a correction, wait for a sufficiently representative circuit condition before concluding. A portion prepared immediately after addition does not always describe all circulating stock. Record the time, lot and production area concerned. The next decision should rely on material actually used for the moulds being evaluated.
8. Qualify trials useful to production
Choose a controlled reference and one main factor to study. Different water preparations can be compared with an identified matrix and binder, followed by a material change in a second stage. Do not multiply variables without a plan allowing their effects to be interpreted.
The cited university study models dependencies between properties for a defined experimental dataset 4. It does not supply an automatic recipe transferable to every foundry. A model calculated for particular materials and conditions needs examination before use at another installation.
For the local campaign, retain raw data and mould observations. Repeat selected preparations before introducing them on the line. An isolated better value can arise from test variability. Selection should be consistent with the other properties and observed operation.
9. Connect the laboratory with the actual mould
The mould contains areas with different filling and loading conditions. Select a trial geometry representing difficulties relevant to the decision: details, pockets, walls or handling. A laboratory cylinder is a measurement reference, rather than a complete reproduction of all these conditions.
Associate tests with the corresponding manufacture: material, preparation, machine, settings, delay and casting. Document defects using a consistent method and identified photographs. A description such as “worse” does not allow later trials to be compared. Define what actually changed and where the change appears.
If the casting has an anomaly, do not automatically conclude that bentonite or sand is defective. The dossier should also retain pouring conditions and other relevant process information. Mixture qualification reduces some uncertainty; it does not remove the other manufacturing factors.
10. An example diagnosis of variable returns
Suppose measured moisture stays near its reference while compactability and mould integrity vary. This is an interpretation example, rather than a measurement performed by Minerallum. Start by checking methods, delays and portion preparation. Apparently comparable results can concern different moments in the circuit.
If these elements are constant, examine return changes: core production series, virgin-sand proportion, temperature and fines distribution. Keep these as hypotheses until tested. Coincidence with a series change helps direct investigation but does not establish the cause.
Then prepare a controlled comparison using retained portions. Keep masses, equipment and conditions identical when the objective is to compare materials. If a preparation correction is investigated, identify it separately. This stage allows an action connected to an observation to be selected rather than a succession of compensating additions.
11. Write an actionable control sheet
The final sheet identifies matrix, additions, returns, sampling locations and methods. It distinguishes usual values, selected limits and associated actions. A limit without an action creates an unhelpful alert; an action without traceability prevents knowing whether the correction worked.
Also define events requiring reassessment: supplier, bentonite, new core stream, equipment or substantial production-rate change. The qualified system can evolve even when the sand name remains identical. Purchasing, laboratory and production should know the modifications and use a shared version of the sheet.
Green-sand quality ultimately rests on monitored preparation and combined interpretation of results. Sand, bentonite and water are not three independent purchases whose characteristics should each be maximised. They form a moulding system to qualify, maintain and reassess when its materials or operation change.
12. Compare sand at several circuit locations
A particularly useful campaign compares mixer discharge with arrival at moulding while identifying the same preparation. The objective is not to obtain two supposedly identical values at any cost, but to establish what happens during transfer and waiting. Record times, storage conditions and operations between the locations. Physical distance alone does not describe this journey.
Sampling should be arranged without arbitrarily changing material intended for each test. Prepare properly identified portions and retain the masses required for planned analyses. When measurements are performed in two laboratories or with two instruments, first check their comparability. An equipment difference can wrongly be attributed to quality loss during internal transport.
If sand is satisfactory at mixer discharge but different at moulding, first examine journey conditions. Is stock withdrawn in the intended order? Is it mixed with a previous preparation? Has the delay increased after a shutdown? Are temperature and moisture monitored at the appropriate location? Each question corresponds to a possible observation and should retain that status until the data resolve it.
An action on transfer or stock organisation can then be studied separately from a recipe change. For example, comparing two waiting times with the same sand addresses the delay question. Comparing two distributions with the same waiting time addresses the material question. Combining these trials makes the conclusion less precise, even when the final result appears better.
In the report, associate observations with the corresponding preparation and moulds. Shift averages can help monitor a trend but do not always identify the sand used for a particular casting. Retain detailed data when the campaign seeks a cause. It is then possible to decide which simplified indicators are suitable for routine monitoring.
This journey comparison complements raw-material monitoring. It shows why purchasing compliant sand and preparing a satisfactory mixture are both necessary stages, yet still insufficient to describe the condition at moulding. The dossier becomes more useful when it connects incoming inspection, preparation, transfer and the actual mould. That continuity supports maintenance of a qualified system without multiplying formulation corrections after every incident.
Before the campaign, check that equipment clocks and production records provide compatible time references. A mismatch between records can associate an analysis with the wrong mixture. Also arrange identification of transitions: shift start, restart after shutdown and production-series change. These periods should be recognisable in the data. Separating them allows regular operation to be compared with a transient situation, rather than building an average that combines two different conditions. The result remains an interpretation of the actual process, with its changes documented.