Sieve analysis starts with a straightforward observation: mass retained at each physical sieve opening. Converting those masses into retained percentage, cumulative retained percentage and cumulative passing is deterministic.
Interpretation becomes more delicate when the curve is summarized by D10, D50 or D90. A percentile does not always fall exactly on a measured opening. A single value between openings is then an estimate that depends on an interpolation method.
Minerallum SIV-1.0.0 therefore separates values calculated directly from the entered data from estimates made between sieve points.
1. Recovered mass
For retained masses m_i and pan mass m_pan:
M = sum(m_i) + m_pan
Particle-size percentages are normalized to this recovered mass.
2. Retained percentage
For sieve i:
r_i = 100 × m_i / M
3. Cumulative retained
Sieves are ordered from largest opening to smallest:
R_i = sum of retained percentages through sieve i
4. Passing percentage
P_i = 100 - R_i
The result is a cumulative mass-based distribution based on the physical sieve openings actually entered.
5. What does D50 mean here?
In this tool, D50 is the point associated with 50% passing/finer by mass on the cumulative sieve-data curve.
It is a mass-basis percentile from sieve data. It is not automatically the same quantity as a laser-diffraction Dv50.
Particle shape and different measurement principles can produce significant differences among methods.
6. Three percentile situations
Exact
If 50% passing occurs exactly at a measured opening, D50 is directly located at that point.
Bracketed
If 50% lies between:
- 35% passing at 250 µm;
- 75% passing at 500 µm;
the result directly supported by the measurements is:
D50 is bracketed between 250 and 500 µm.
Not bracketed
If the target lies below the smallest measured opening range or above the measured range, the tool does not extrapolate it.
7. Why more than one interpolation is possible
Sieve data are discrete. Producing one scalar between two openings requires a model between the measured points.
SIV-1.0.0 offers, only when selected:
- log-size interpolation;
- linear-size interpolation.
The selected method is always displayed with the estimate.
The default is bracket-only, with no interpolation.
8. The pan is not a 0 µm sieve
The pan collects material that passed the smallest actual sieve. It contributes to the mass balance but is not a measurement at zero micrometres.
Using it as a zero-size coordinate would invent a measurement.
9. Mass closure
If initial test-portion mass is known, the tool can compare it with total recovered mass.
It reports the difference and recovery percentage, but it applies no universal compliance threshold.
Unaccounted mass cannot be automatically assigned to particle-size classes.
10. Measurement limitations
The calculator does not verify:
- sample representativeness;
- sieve condition/calibration;
- balance calibration;
- blinding;
- agglomeration;
- moisture;
- material loss;
- sieving time/efficiency;
- particle-shape effects.
It does not certify a material or replace a test method.
11. Why transparency matters
A particle-size result is not just a number. It depends on its distribution basis, measured points and - when estimation is needed - interpolation choice.
Minerallum proof mode exposes:
- masses;
- percentages;
- cumulative values;
- the percentile bracket;
- any selected interpolation formula.
Calculator technical values remain local in the browser.
Conclusion
The strongest part of sieve analysis is the cumulative distribution calculated from recovered masses. D10, D50 and D90 are then measured points, intervals or method-dependent estimates.
SIV-1.0.0 is designed to make that distinction visible instead of hiding it behind a single number.