Thanks alot my dear brother

Ok, our Re number is about 33 million that is a large number. For this case can we discard the BL effects and calculate the derivatives with about 2% error that you say?

I suggest looking at the International Towing Tank Conference (ITTC) publications on the topic which are available free on line and can be found with a simple google search. In most cases towing tank work is only able to deal with the Froude Number scaling and not the Reynolds number problem. To some degree the Correlation allowance is set up to correct for the difference between full scale and model scale. There were a bunch of tests done back in the last century (1900 to 2000) where the same hull form was tow tested with different size models down to quite small and I believe that the ITTC results reflect those results. Good Luck