You do a proctor in a mold, get some value at some moisture content. The ASTM defines four methods and if there's a lot of coarse-grained materials, you'd be using method D. That's based on minus 3/4-size material. Let's say there is actually another 20 percent of material that's greater than the 3/4-size.
What you'd then have in the field is 80 percent of your material with some idealized maximum dry density - let's say that value is 120 pcf. You'd also have 20 percent of the material that'd have a greater density - some density based on the specific gravity of the soil grain. Well, we can know the contribution of these particles to the overall density 'cause these particles are not affected by compaction moisture content. If the specific gravity of the soil grains is 2.7, that'd mean the 20 percent of oversize material wants to be at 168.5 pcf. So, in the idealized state of 100 percent relative compaction, you'd have 80 percent at 120 and 20 percent at 168.5, which would return a maximum dry density of 129.7. This would relate to the condition of compaction at the optimum water content. Now the procedure is a bit more elegant that that, 'cause there'd also be some adjustment to the optimum water content. Refer to D698 (ASTM) for more context and referenced standards.
f-d
ípapß gordo ainÆt no madre flaca!