The pipe sits back a bit to allow bypass air to be sucked in the front of the thrust tube. An article in RCJI, about half a year ago, showed that, in the particular installation studied, the nominal longitudinal gap between thrust tube front and turbine rear was, I believe, .25".