Splitting a Field Sample That Won't Cooperate
A technician collects a 48 lb production sample of 3/4 in. crushed limestone from a stockpile conveyor belt. The lab needs a 22 lb sample for gradation and a 5 lb sample for LA abrasion. The technician uses a 12 in. wide riffle splitter with 3/4 in. chutes. After the first pass, the two halves weigh 31 lb and 17 lb. The technician recombines and splits again, getting 29 lb and 19 lb. A colleague suggests quartering on the floor instead, but the floor is damp and there is a floor drain nearby. The technician also notices several flat, elongated particles bridging the chute openings.
1. Evaluate whether the riffle splitter is the correct reduction method here and identify the likely cause of the unequal splits.
Read the worked answer
A riffle splitter is appropriate for a free-flowing material, but the chute width must be at least 3x the nominal maximum particle size for reliable splitting. For 3/4 in. (19 mm) material, chutes should be roughly 2.25 in. wide or larger; 3/4 in. chutes are too narrow and cause bridging and segregation, especially with flat/elongated particles. The unequal masses (31/17, then 29/19) confirm a systematic bias, not random variation. The correct action is to switch to a larger splitter (or a mechanical splitter with proper chute width) rather than repeat the failing pass. Recombining and re-splitting with the same undersized device will not correct the bias.
2. Assess the quartering alternative given the floor conditions, and state what you would do instead.
Read the worked answer
Quartering requires a clean, dry, hard, level surface and a mixing procedure (forming a cone, flattening, quartering). A damp floor near a drain introduces contamination and moisture pickup, and any spill toward the drain loses fines-both bias the reduced sample. Do not quarter on this floor. Options: use a larger riffle splitter; use a mechanical sample divider; or, if quartering is the only option, use a clean steel plate or tarp on a level surface away from the drain. Document the deviation and the method actually used.
3. The technician wants to take the 5 lb abrasion sample from the first split half. Is that acceptable?
Read the worked answer
No. The first split half came from a biased, non-representative split (unequal masses, bridging). A test sample must be reduced from a properly mixed, representative parent sample. Take the abrasion sample only after a valid reduction step produces a representative subsample. Also verify the 5 lb mass meets the minimum for the specified abrasion gradation; if the required grading needs more mass, the 5 lb target is itself wrong and must be recalculated from the test method.