[I’m gonna try this response for the 3rd time today - shortest yet this time so in case it disappears I’m not losing that much effort . . . [%-)] ]
Only about 3/4 of an inch ! Just a little deeper than the standard point where the gage is measured - which is at 5/8" below the top - for a new rail, and not even as deep as the nominal wheel flange depth of 1 inch. So, as the gage face wears away (from curve wear, typically) there’s not much depth of head left below that/ underneath to support the wheel and keep it from falling in as the gage gets wide. Hence RWM’s cynical/ sarcastic (and on point) comment above about it being “gage free” rail !
For more info, see Caltrains Standard Drawing SD-2201 - “DATA FOR STANDARD RAIL SECTIONS” at:
http://www.caltrain.com/engineeringstandards/SD2201.pdf
The “HEAD-FREE SECTION” detail is just to the upper right of the center of this sheet. Note that the “undercut” angle is steeper than I indicated in my post yesterday.
The pertinent dimension is “D-1” - “DEPTH OF HEAD - AT CORNER” (lower), in the chart or tabulation at the upper left corner of this sheet. Note that it is 25/32 inch (3/4 inch, practically) for the 132-Lb. Head Free section, as compared with 1-9/16 inches (1-1/2 inches, practically) for the standard 132-Lb. RE “Tee” rail section = only half as deep !
For more on the head-free theory, see paragraph "b. Head-Free Joint Ba