A chip can pass the 20 mm sieve, sit on the 16 mm sieve exactly as the grading wants, and still be a sliver 6 mm thick. Sieves measure size. They say nothing about shape, and shape is what the flakiness and elongation index is for.
The test method below is from IS 2386 Part 1 and the limits are from the MoRTH Specifications for Road and Bridge Works, Fifth Revision (2013). Check the edition your contract cites before quoting a limit from here in a meeting.
What flakiness index and elongation index mean
The flakiness index of an aggregate is the percentage by weight of particles whose least dimension, the thickness, is less than three-fifths (0.6) of their mean dimension; the elongation index is the percentage by weight of particles whose greatest dimension, the length, is more than one and four-fifths (1.8) times their mean dimension.
Mean dimension is not measured on the particle. It is the average of the two sieves that bracket it: the one it passed and the one it was retained on. That is why the test starts with sieving, and why each sieve fraction has its own slot on the gauge. Neither test applies to material smaller than 6.3 mm.
Mean size = (20 mm + 16 mm) ÷ 2 = 18 mm
Thickness gauge = 0.6 × 18 mm = 10.8 mm
Length gauge = 1.8 × 18 mm = 32.4 mm
Thinner than 10.8 mm is flaky. Longer than 32.4 mm is elongated.
The flakiness index test and the combined index
IS 2386 Part 1 describes two separate tests on the same sample. MoRTH joins them, and the footnote under its physical requirement tables says how: the flaky stone is separated first, and only the non-flaky remainder is tried on the length gauge. Run that way, the test goes in this order.
- 01
Sieve the sample
The sample is split into the sieve fractions of IS 2386 Part 1, from 63 mm down to 6.3 mm. Each fraction that is tested needs at least 200 pieces.
- 02
Thickness gauge
Every piece of each fraction is tried through the slot cut for that fraction, 0.6 times its mean size wide. Whatever passes is flaky. It is set aside and weighed.
- 03
Flakiness index
The weight of flaky stone divided by the weight of the whole sample gauged, as a percentage.
- 04
Length gauge, on the non-flaky stone only
The stone that did not pass the thickness gauge is offered, piece by piece, between the pins of the length gauge, set 1.8 times the mean size apart. Whatever will not go between the pins is elongated. It is weighed.
- 05
Elongation index, then add
The weight of elongated stone divided by the weight of the non-flaky stone, as a percentage. The combined index is the two percentages added together.
Sample gauged = 2,000 g
Passing the thickness gauges = 360 g
Flakiness index = 360 ÷ 2,000 = 18.0 %
Non-flaky stone = 2,000 − 360 = 1,640 g
Retained on the length gauges = 246 g
Elongation index = 246 ÷ 1,640 = 15.0 %
Combined index = 18.0 + 15.0 = 33 per cent, inside a 35 per cent limit
Two things follow from that sequence. A particle that is both thin and long is counted once, as flaky. And the elongation figure is a percentage of the non-flaky stone, not of the whole sample, so it is not the number a laboratory gets by running the two IS tests independently. When a plant lab and a site lab are a few points apart on the same stockpile, ask both which denominator they used before anyone argues about the stone.
Flakiness and elongation index limits as per MoRTH
| Layer | Clause and table | Combined index |
|---|---|---|
| Water bound macadam | Clause 404, Table 400-8 | 35 per cent maximum |
| Wet mix macadam | Clause 406, Table 400-12 | 35 per cent maximum |
| Crusher-run macadam base | Clause 407, Table 400-15 | 35 per cent maximum |
| Bituminous macadam | Clause 504, Table 500-6 | 35 per cent maximum |
| Dense bituminous macadam | Clause 505, Table 500-8 | 35 per cent maximum |
| Bituminous concrete | Clause 507, Table 500-16 | 35 per cent maximum |
| Stone matrix asphalt | Clause 515, Table 500-35 | Less than 30 per cent |
| Pavement quality concrete | Clause 602.2.6.2 | 35 per cent maximum |
Granular sub-base under Clause 401 carries no flakiness or elongation requirement in Table 400-2. For water bound macadam the limit is enforced only on crushed or broken stone and crushed slag.
One figure for nearly every layer is easy to remember and easy to misread. It is an acceptance ceiling, not a target. A plant reporting 33 or 34 on every sample is one worn liner away from a rejected stockpile, and the test is not a rare one: Section 900 asks for it once per 500 cubic metres of aggregate on wet mix macadam and once per 350 cubic metres per source on DBM and bituminous concrete. The rest of the physical requirements sit alongside it in aggregate tests and their MoRTH limits.
Why flaky stone is a problem in a pavement
- It breaks. A thin particle bridging two others is a small beam, and it snaps under the roller or under traffic. In a granular layer that makes new fines after the grading was approved. In a bituminous layer it opens uncoated faces inside the mix.
- It lies flat. Flaky particles turn onto their broad faces during compaction and stack like tiles, so the layer has planes it can shear along instead of stone locked against stone.
- It will not pack. For the same weight, flaky and elongated stone leaves more voids and has more surface. A granular layer is harder to bring to density, and a bituminous mix wants more binder to coat it.
- It moves the mix design. A job mix proved on cubical stone does not behave the same when the next month's chips are flakier, even if every sieve still passes.
What makes a plant produce flaky stone
Shape is decided by how the rock was broken, and most of the causes are settings and habits, not geology.
- Jaw-only product. A jaw breaks large lumps in a single layer between two plates, and rock loaded that way splits into slabs. Chips screened straight from jaw product are the flakiest stone a plant can sell. What each stage is for is set out in jaw, cone and VSI.
- Too much reduction in one stage. A cone closed right down to make 10mm from coarse feed in a single pass is working past its ratio. It slices instead of crushing.
- A cone that is not choke fed. With the chamber full, stone breaks against stone and corners come off. Fed in a trickle, each piece is nipped once between steel faces and leaves as it broke.
- A setting far from the product size. Shape is best in the stone closest to the closed side setting. Chips much smaller than the setting have fallen through the chamber almost untouched, so 10mm taken off a cone set for 40mm is poorly shaped by design.
- Worn liners. As a mantle and concave wear out of profile the chamber stops gripping the stone the way it was cut to. Shape can drift before the tonnage does, which is one of the signals covered in what decides wear life.
- The rock. Bedded sandstone, slaty and schistose rock, and flow-banded basalt all have a plane they prefer to split on. A plant can improve them. It cannot make them granite.
How a plant brings the index down
- Put the chips through a tertiary stage. A short-head cone in closed circuit, or a VSI, exists for this. A VSI is the strongest correction available, because it breaks particles against each other at speed and a flake does not survive that.
- Keep the chamber full. A surge bin and a controlled feeder ahead of each cone, and an operator who slows the feeder instead of letting the cone run half empty.
- Share the reduction. Open the secondary a little and let the tertiary do more, so that no stage is slicing.
- Send more round again. A slightly tighter top deck returns more stone to the tertiary crusher, and stone that has been through twice is more cubical. It costs saleable tonnes per hour, which is the trade-off worked through in crusher plant capacity in TPH.
- Change liners on shape, not on thickness. If the index climbs across a liner's life, the set is finished for chip production even with metal left on it.
What to ask a supplier
Ask for the combined index on each single size separately, 20mm and 10mm, not one figure for a blend, and ask which crushing stage each size came off. Then ask for the last three results, not the best one. Our plants run a primary jaw, a secondary cone and a tertiary stage feeding the screen decks, with samples pulled at commissioning and through the run, because flaky aggregate off a badly set crusher fails the index however clean the grading looks. The sizes are listed on materials and the plants on stone crushing.