Pilotho,
Correct.
Does that mean if the leg only had one wheel, it would be exerting a load of 45T on the ground? However, if there were 4 wheels, there would only be a load of 20T on the ground from each wheel = 4 * 20T = 80T.
The load spreads down from the top layers to the subgrade through load transfer by the layers through a cone of about 45 degrees as shown in this
figure.
If there are 4 wheels each of 20 tonnes, there is a total load of 80 tonnes on the ground. Immediately under a single wheel, where only the effect of that single wheel can be felt by the soil particles, the load on a soil particle is 20 tonnes. This is the same effect as if the pavement was loaded with a single wheel of 20 tonnes. Hence at the surface, the ESWL is 20 tonnes, even though there is an 80 tonnes load on the ground spread over 4 wheels.
As you go deeper, the effect of the other wheels starts to become felt as the load 'cones' start to overlap. Eventually you get deep enough so that all the load cones are overlapping. At that point, you have the effect of all 4 wheels of 20 tonnes each (= 80 tonnes) on the soil particles, and this is the same effect as if the pavement was loaded with a single wheel of 80 tonnes. The ESWL is 80 tonnes.
In between, the ESWL varies between 20 tonnes and 80 tonnes, depending on how deep you are and the extent to which the load cones overlap. The typical pavement is say 750mm deep, and at that depth, the overlap is such the vertical load on the soil particles is the same as would be induced by an ESWL of 45 tonnes.
Of course as you go deeper, the load(s) spread or dissipate, and while the load seems to increase as load cones overlap, it is decreasing rather more quickly because the load(s) are being spread over a bigger area. What an awkward clunky workaround the ESWL concept was, IMHO. Almost as awkward as PCN, but don't get me started on that.