Several references indicate that the torsional stiffness of concrete members is significantly less than the full gross section properties. Two of the references indicate that in the concrete structure will behave based on the reinforcing provided, and if torsion stiffness is assumed to be small the analysis will redistribute the forces and the engineer will design for these redistributed forces. In essence the structure will behave the way it was assumed in the analysis (if we reduce torsion reinforcing it will crack and torsion forces will be redistributed).

For cases where torsion ensures equilibrium (i.e., no redistribution is possible) then forces cannot be redistributed and the beam section will need to be reinforced to ensure that the calculated torsion force can be resisted without excessive deformation.

The references include:

• Reinforced Concrete Structures, Park and Paulay, John Wiley, 1975.

  The authors indicate for most situations the assumption of zero torsional stiffness can be made. They do indicate that it is still important to provide at minimum torsion reinforcing to prevent excessive service load cracking.

• Reinforced Concrete Fundamentals, 4th Edition. Ferguson. John Wiley, 1979.

  The author indicates that while flexural stiffness decreases maybe 50 percent from cracking, torsional stiffness drops down to 5 or 10 percent its uncracked value. The author also mentioned that the consideration of the torque to be used in the design is very complex due to the cracking effect. Thus it is always better to neglect the rigidity of the members for torsion and to consider them fully cracked.

• Hormigón Armado (Spanish). Motoya, Meseguer, and Moran. 12th Edition. Ed. Gili, Madrid-Spain, 1988 .

  The authors mention that the structure will behave exactly in the same way as it was idealized in the analysis (cracked or uncracked). They suggest modeling the reinforced concrete structures with a very low torsion rigidity i.e. assume it is significantly cracked.