Grain function comes before the mineral name
A membrane manufacturer is not simply seeking clean sand. It qualifies a material for a precise position within a product: finishing a face, exposed protection or another function defined by its formulation. The word “aggregate” does not make these uses interchangeable.
The Poseidon PO 4000 sheet from BTM, a manufacturer based in Turkey, describes a membrane with sand on its upper face and polyethylene film on its lower face 1. This documents sand use within one identified product. The sheet does not provide a silica-sand specification applicable to the entire Turkish industry.
This dossier explains how to prepare qualification without confusing starting sand, potentially treated aggregate and the finished membrane. Its purpose is to connect material with its function and the manufacturer’s tests, rather than turn a roof photograph into a compatibility promise.
1. Identify the face and position within the system
Begin with a product diagram. Identify the upper face, lower face, overlap area and membrane position within the assembly. A grain applied to a face intended to be covered is not automatically exposed to the same stresses as a grain on a final layer.
IKO describes its P3 T/F Atelia base layer as having a sanded upper face 4. Keep that information attached to the particular reference. It does not establish that every base layer, or every lower face, carries the same sand.
The first purchasing question therefore becomes precise: which surface receives the material, at which stage and for which function? While those three points remain unknown, comparison based only on average diameter omits an essential part of qualification.
2. Distinguish sand from protective mineral granules
BTM’s Yonca shingle presentation distinguishes sand on the lower face from coloured mineral granules on the upper face 2. Two mineral materials can therefore coexist within a product in different positions.
A commercial designation such as “mineral granules” does not by itself establish that the material is quartz, unchanged natural sand or identical to the material on the other face. Request the identity and preparation of the aggregate retained in the technical dossier.
The decision is consequently not to replace every aggregate with silica sand. It is to verify whether a particular fraction meets a particular function, including any preparation. Validation for one face must retain that scope when another product is considered.
3. Separate ultraviolet protection from the mere presence of grains
IKO attributes additional protection against ultraviolet rays to the granules on its Armourplast Granular product 3. This claim concerns that membrane and its surfacing. It does not automatically give the same protection to a sand layer with different composition and distribution.
For an exposed application, define with the manufacturer which properties to check after ageing and the comparison conditions. Visible grains do not by themselves quantify protection of the bituminous layer.
A dossier must therefore distinguish mineral identity, achieved coverage and product performance. Tests can connect these levels; they should not be merged into a promise that every sand provides equivalent external protection.
4. Qualify the material actually applied
Retain a description of received material and material deposited on the membrane. Sand may be dried, classified or treated before incorporation. If preparation modifies its distribution or grain surface, that preparation belongs to the tested variant.
The dossier must identify who performs the operation and which controls define the final material. A successful trial using a prepared fraction does not establish the same result for sand delivered without that preparation.
Also retain removed fractions when examining them answers a useful question. A supplier comparison can become a preparation comparison if this distinction is lost. The manufacturer must be able to identify the material that actually produced the observed result.
5. Describe a curve rather than only an advertised range
A commercial range does not fully describe grain distribution. Request the useful control points, method and reporting basis. The coarse portion and fine fraction must be comparable between lots entering the trial.
Connect distribution with the deposition operation. The programme must examine whether the retained fraction can be applied consistently under defined conditions. A manually prepared sample producing an acceptable surface over a small area is insufficient on its own.
Retain stages where selection or segregation can modify the material. A receipt curve and a post-preparation curve are not necessarily identical. Qualification must show which distribution concerns the layer actually obtained.
6. Distinguish deposited mass from surface coverage
Two preparations can receive the same mass per unit area without exactly the same visible distribution. Mass measures quantity; coverage observation examines where that quantity lies. Specify which of these criteria governs comparison.
Define the observed area, sample preparation and inspection method. Photographs taken at different angles or under different lighting can complicate comparisons. Retain repeatable observation of the studied regions in the dossier.
Also examine regions not represented by a single average. Consistent deposition at the centre can coexist with a different edge region. If that region contributes to the product or its assembly, examine its behaviour in the plan.
7. Do not confuse initial loose material with grain retention
Distinguish loose material collected after manufacture from material detached by a defined test. Both quantities can be useful, but do not automatically represent the same phenomenon. State the stage at which each fraction was recovered.
The public ASTM D4977/D4977M notice describes an abrasion test of granule adhesion involving materials as manufactured 5. Its stated scope does not turn the test into universal evidence of retention after weather exposure.
For comparison, the laboratory must define the method actually used and the manufacturer its acceptance criteria. This dossier reproduces neither the complete standard procedure nor an acceptance threshold. The result must retain the scope of the product and test.
8. Direct analysis towards the relevant interface
When grains detach, examine what remains on the surface and what was collected. Lost mass alone does not show whether the question concerns the grain, its interface or part of the coating. Select with the laboratory observations capable of distinguishing these objects.
Retain an untested reference and comparable regions. Observation of received material can supplement analysis, but does not directly describe a manufactured interface. Product preparation must remain in the dossier.
Avoid turning colour or appearance into a certain cause. Identifying a particle or layer requires an appropriate method. Separate what is observed, what is compatible with a hypothesis and what a corrective trial can establish.
9. Connect ageing with finished-product control
The BTM sheet associates tests with several membrane characteristics, including artificial ageing and watertightness 1. These tests concern the membrane rather than a certificate for sand considered in isolation.
During material qualification, select final controls according to the product and system. Retain conditions before and after exposure within the programme. Better initial appearance does not automatically demonstrate a better response after ageing.
Examine a grain change alongside other necessary requirements. Reduced surfacing loss is insufficient if an essential product property becomes incompatible. The decision belongs to overall qualification using criteria defined before the campaign.
10. Examine overlaps as distinct regions
An overlap should not be represented solely by a measurement at the surface centre. Identify assembly regions and preparation specified by the manufacturer. Retain the distinction between an ordinary face and an area intended for a joint.
If a mineral variant changes an assembly region’s condition, examine it in the relevant configuration. Do not infer joint quality from a control performed only on loose grains. Evaluate product elements in their actual relationship.
This approach supplies no universal installation method. It allows the manufacturer to choose tests appropriate to its system and avoid a favourable deposit observation concealing an unstudied assembly question.
11. Build a comparison isolating the change
Prepare a reference and variant with identified materials and operations. When investigating the grain, retain relevant deposition and preparation conditions. If a setting must change, record it and interpret the result as belonging to that combination.
Manufacturing repeats provide different information from several measurements on one specimen. The former examine part of process consistency; the latter describe part of observation variability within that product.
Keep these levels in the report. An average without a defined experimental unit can imply robustness that the programme did not examine. Comparison should answer a precise technical question rather than multiply unrelated results.
12. Example of deposition per unit area
Imagine 12 kg of mineral material applied across 30 m² of product. The average quantity is 0.40 kg/m². This invented educational example describes neither a Turkish membrane nor a grade offered by Minerallum.
If another series uses 18 kg across 45 m², its average is also 0.40 kg/m². That equality establishes neither identical coverage, identical adhesion nor identical ageing. These properties require the observations or tests selected in the programme.
Retain the numerator’s meaning: prepared material, deposited material or material still retained after an operation. Treating these successive masses as the same deposit can make comparison misleading. Keep the area and measurement stage attached to the result.
13. Example of recovery after a defined stage
Imagine a surface portion receiving 100 g of an identified fraction, followed by an operation collecting 8 g of that fraction. If the balance contains no other loss or input, 92 g remain within the boundary. This invented calculation explains a mass basis rather than a roofing criterion.
The quotient 8/100 is 8%. It represents the collected fraction relative to the defined initial quantity. It does not automatically become a standardized index, service life or proof of incompatibility. The method and criterion must provide that interpretation.
If collected material also contains coating fragments or another material, do not equate its total mass with grain loss without analysis. Defining the object weighed is as important as calculating the percentage.
14. Describe a lot change and its scope
Qualification must produce usable receipt criteria: material identity, preparation, distribution and characteristics defined by the requirement. Every value must retain method, unit and lot. A vague designation such as “sand for bitumen” cannot replace this information.
Then separate material controls from membrane validations. The supplier can present defined material; the manufacturer qualifies its behaviour in the process. A favourable result does not justify removing controls needed for consistent deliveries.
Our silica-sand page helps prepare the mineral enquiry. For a Turkish application, specify the grain’s position and tested configuration. Do not attribute approval by the cited manufacturers or a finished-membrane offer to Minerallum.
15. Compare cost per accepted area
An economic comparison must preserve the selected technical boundary. Received-material cost, preparation cost and deposition operations can be tracked separately. Material cheaper per tonne may require more preparation or yield less accepted product. Make those differences visible rather than attribute them to one grain property.
In a fictional example, a campaign spends 600 euros on material and preparation to produce 1,000 m², of which 900 m² are accepted against defined criteria. Those cost items per accepted area equal 600/900, approximately 0.67 euros/m². Another campaign spends 660 euros for the same produced area and accepts 990 m²: the quotient is again approximately 0.67 euros/m². These invented amounts are neither Turkish prices nor Minerallum quotations.
Equal quotients do not establish complete equivalence between variants. Excluded items, manufacturing time, rework and final properties still require examination. Accepted area must also correspond to the same function and requirements in both campaigns.
Keep observed costs, assumptions and technical results on distinct lines. This balance helps select a repeatable variant without crediting sand with savings caused by a product change or less demanding inspection. It connects purchasing with qualified area rather than an isolated delivered quantity.
16. Decide according to the demonstrated function
A useful conclusion distinguishes three answers: defined material, compatible deposition and product meeting criteria. A favourable answer at the first level does not automatically validate the other two. Retain additional operations needed to achieve the result.
When several variants are possible, compare them within the same boundary: received material, preparation, deposition, losses and accepted product. A tonne’s price alone cannot compare qualified-area cost. Seek a repeatable option with the selected controls.
Silica sand can belong in a technical enquiry when its function is defined and tested. Qualification must remain attached to that function. This precision gives manufacturers a decision basis and suppliers a verifiable requirement without presenting every roofing aggregate as interchangeable.