Considering the array or cell array is best way rather than assigning new variable in each iteration.
Fs=zeros(1,35);
for i=1:35
Fs(i)=L(i)/tmax(i); % assuming L and tmax both are scalars
end
% Assuming L and tmax both are scalars, if not then
Fs=cell(1,35);
for i=1:35
Fs{i}=L(i)/tmax(i);
end
