Have you ever considered two-way bending with grating? The tripping point is: you have the moment of inertia/section modulus for the direction it spans but not the other direction. Thoughts?
The only reason I could see doing this is because you are having making a section of grating work when considering one-direction spanning. I've designed some industrial facilities that have required some extraordinarily hefty grating and I've always been able to find a section that would work, even if I had to cut the span down to a really small number.
To echo Mike, why spend design budget on something that should take 5 minutes?