I think first of all we have to join R and T then (R join T) join S which will give 50 tuples but in between ((R join T) join S) and leftover U attribute D and G is common which is not key of any of the schema hence in this case number of tuples should be 50*10= 500 as it will behave like Cross product.