IS 2430 : 1986Methods for Sampling of Aggregates for Concrete

IS 2430:1986 specifies the methods for sampling of aggregates for concrete — how a *representative* sample of fine and coarse aggregate is drawn from stockpiles, bins, conveyors, wagons or trucks before any grading or quality test. It is the always-first step: the IS 383 grading/quality limits and the IS 2386 test results only describe the consignment if the sample was drawn per IS 2430.

It sits in the aggregate-acceptance stack:

• IS 383 — aggregate grading, deleterious limits (the spec the result is judged against)
• IS 2386 (Parts 1–8) — methods of test for aggregates (run on the IS 2430 sample) · IS 456 — concrete (the structure that depends on it)
• IS 10262 — mix design (uses the graded sample's properties)

Aggregate is the most variable concrete ingredient: stockpiles segregate (coarse rolls to the toe, fines stay at the apex), get contaminated at the base, and vary load-to-load. IS 2430 fixes:

• Increment sampling at multiple points — for stockpiles, increments from top/middle/bottom and around the pile (never one scoop off the surface); for flowing streams/conveyors, full cross-stream cuts at intervals; for bins/wagons, sampling through the depth
• Number and size of increments related to the aggregate's nominal maximum size — a 40 mm aggregate needs a far larger gross sample than sand to be representative
• Compositing then reduction by quartering or a riffle box down to the test quantity, so the laboratory portion still represents the consignment

The engineering point: grading and silt/clay content drive water demand and strength — and they vary across a single pile by more than the IS 383 tolerance. A non-representative sample certifies a scoop, not the stockpile feeding the mixer.

Scenario: acceptance of a 20 mm coarse aggregate stockpile for M30.

Step 1 — gross sample: take increments from several locations and depths of the pile (dig in past the surface segregated layer), at least the IS 2430 number/size for 20 mm aggregate.

Step 2 — composite: combine all increments on a clean hard surface and mix thoroughly.

Step 3 — reduce: quarter (or riffle) the heap down to the laboratory test quantity, keeping a retained portion.

Step 4 — test & judge: run IS 2386 grading (sieve analysis), flakiness/elongation, and deleterious content, and judge the stockpile against IS 383 — accept/reject the consignment, not the scoop.

Skip the multi-point increments and you either reject good aggregate on a fines-rich apex sample, or pass a silty/segregated pile that drives up water demand and cracks the IS 456 concrete.

1. One scoop off the pile surface. The surface is the segregated, weathered, contaminated zone — never representative.

2. Too few increments / sample too small for the aggregate size. A handful for 40 mm aggregate cannot represent the grading — increment count and size scale with nominal max size.

3. No proper reduction. Sub-sampling the gross heap by grabbing the convenient part re-introduces the bias quartering/riffling is meant to remove.

4. Sampling only one load and inferring the source. Loads vary; sample per the plan, per consignment.

5. Sampling divorced from acceptance. The plan exists *so that* IS 383 limits and IS 2386 results are valid — sample, test and accept as one chain.

• IS 383 — coarse & fine aggregate specification (grading, deleterious limits)
• IS 2386 Parts 1–8 — methods of test for aggregates (particle size, deleterious, mechanical)
• IS 456 — plain & reinforced concrete (acceptance dependency)
• IS 10262 — mix design (graded-sample properties feed the design)
• IS 3535 — sampling of hydraulic cement (parallel sampling logic)

IS 2430 is reaffirmed and invisible on site, yet aggregate is the *most* variable thing in the mix and the cheapest to get wrong: a silty or badly-graded pile inflates water demand, drops strength and cracks slabs, and the lab can only catch it if the sample was honest. Disputed grading/silt failures very often unwind not to the quarry but to the sample — a convenience scoop off the apex. The fix is procedural and costs nothing: multi-point increments through the pile depth, composite, quarter down, judge the consignment. Witness it — if the engineer doesn't watch the sampling, the 'representative' sample is whatever was easy to reach. Same discipline as IS 3535 for cement: a test only certifies a lot if the sample genuinely came from the lot.