A37575,
I think that you might have provided a significant clue in your initial post when you stated "when No 2 engine was failed at VR", the aircraft is still on the ground. The 3 subsequent respondants (ASFKAP, LEM, and BOAC), presumably current B737 operators, have correctly indicated that this is a normal event after a power supply change on the ground.
I last flew the B737-300 some 16 years ago, and well remember this "quirk", but it is supposed to be inhibited after the aircraft has commenced taxy. Regretably, 16 years has dulled my memory of the other logical operand within the system which inhibits this self test after the aircraft has begun moving under it's own power. It would appear that the personnel programming the simulator have skipped this "gate", and it could be well worthwhile to investigate it further with the technicians responsible for the simulator's operating system.
Be assured, it will not happen in the real aircraft after it has begun moving under it's own power.
Regards,
Old Smokey