problems for the flow over flat wall
 

hi guys ,i need your help for the problems on the flow over flat wall.below is my problem:
it is DNS for the flow over flat wall,the zone is square,1x1 in dimensionless(it can get smaller unit net size by adaptive code),
RE=10000(base on the velocity of the uniform flow of the upper flow domain),Grid is square,the min grid=0.0002,
when x<0.05,i put the uniform flow there.i used the FVM for this problem,
so the process begin.
when i check the result ,i find the velocity outside the bounday layer is larger than the velocity that i set x<0.05,but the velocity in the boundary layer is very good after compared with the value simulated by Blasius,
in one sentence,the velocity out of the layer is larger than the value of the Blasius,it can be 10%larger ,when reach the top of the domain,it still be 5%larger ,
my workmate used the LBM(lattic boltzman method )for the same problem(his RE=50),he have the same problem with me too.
it is strange!
maybe you can help me.thank you ! i am waiting!