Why? Big question, lots of people have been asking that for the last 10 years.  I am sorry I have only just spotted your question and hope this answer is in time for you.  I think the simple answer is the best.  My simple answer is that interactions diagrams  are derived empirically from test data.  ASME and AFCEN  each had a different set of test data and therefore they proposed different interaction diagrams.  

ASME may have taken their interaction diagram from Ruggles and Ogata 1994 (see 2nd link Figure 8.1). 

AFCEN would have taken their interaction diagram from c-f tests at CEA (see 3rd link).

Then comes a second question why have ASME not changed their interaction diagram?  This one is more difficult to answer so I suggest you ask them.  

Of  course as a fan of ductility exhaustion and the UK life assessment procedure R5 I would simply not use either ASME or RCC-MR (see the 4th link for an approach better than time fraction).  And a 5th link to a review paper on R5.  

http://www.afcen.com/en/publications/rcc-mrx/70/rcc-mr-2007

http://www.osti.gov/scitech/servlets/purl/10131895

http://proceedings.asmedigitalcollection.asme.org/proceeding.aspx?articleID=1600698

http://www.maneyonline.com/doi/abs/10.1179/174328006X79463

Article Advanced Ductility Exhaustion Methods for the Calculation of...

Agree with Michael William Spindler ....Following...