The number is deliberately close but not the same as pi/2.

See here: https://www.hpmuseum.org/cgi-sys/cgiwrap/hpmuseum/archv017.c...

And you can compare to wolfram alphas result.