# The SOLID65 element Tension-Tension-Compression Domain results do not appear to match the theory. Is this a known limitation?

 Defect 19719 was filed to investigate this issue. It appears that the SOLID65 concrete crushing results do not match the theory equations for a portion of the Tension-Tension-Compression loading domain. For the attached input, which consists of a single element, strange results occur for tensile Sigma1 and Sigma2 between 0.1 Pascals and 500,000 Pascals. Different tensile loads are applied (S1=S2) and the compressive load (S3) is increased until element failure. A summary of some of the output data follows:= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = The criterion for failure of concrete is: F/fc - S >= 0 For the Tension - Tension - Compression Domain, you get: F = F3 = Sigma_1 and Sigma_2 (where Sigma_1 = Sigma_2 in this case) S = S3 = (ft/fc)*(1.0 + (Sigma_3/fc)) Substituting the latter two equations into the first: (Sigma_1/fc) - (ft/fc)*(1.0 + (Sigma_3/fc)) >= 0 Solving for Sigma_3 AT FAILURE (i.e., drop the ">" sign): Sigma_3F = fc*((Sigma_1/ft)-1)= = = = = = = = = = = = = = = = = == = = = = = = = = = = = = = = = = = = = = =ANSYS 8.1 SUMMARY for various press_x values and NSUBST,200,2000,200==================================================================== press_x (_z) Sigma_1 (_2) Sigma_3 Sigma_3F============================================================== 1,000,000.0 999,933.4 unconv (4.99e6) -5,000,665.6 900,000.0 899,936.2 -5,991,396.73 -6,000,638.8 800,000.0 799,939.7 -6,991,878.49 -7,000,603.1 700,000.0 699,944.1 -7,992,431.37 -8,000,558.8 600,000.0 599,949.4 -8,998,070.72 -9,000,505.6 500,000.0 499,956.7 -9,703,553.96 -10,000,434.7 400,000.0 399,966.6 -9,703,539.60 -11,000,335.9 300,000.0 299,975.8 -9,703,525.24 -12,000,243.1 200,000.0 199,984.6 -9,603,439.12 -13,000,156.1 100,000.0 99,992.5 -9,703,496.51 -14,000,075.1 10,000.0 9,999.3 -9,703,483.58 -14,900,007.2 1,000.0 999.9 -9,703,482.29 -14,990,000.7 100.0 100.0 -9,703,482.16 -14,999,000.1 10.0 10.0 -9,703,482.15 -14,999,900.0 1.0 1.0 -9,703,482.15 -14,999,990.0 0.1 0.1 -9,703,482.15 -14,999,999.0 0.05 0.0 -14,908,219.5 -15,000,000.0 0.01 0.0 -14,908,219.5 -15,000,000.0 0.001 0.0 -14,908,219.5 -15,000,000.0= = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = =Development closed this defect report out with the following note: The failure surface for SOLID65 is a complex, multifaceted set of functions that for certain stress states may exhibit anomalous response. While for homogeneous stress states this would lead to obvious inaccurate results, for more typical non-homogenous stress states these anomalous responses would be highly localized with little affect on the global response. Because of these anomalies, there are long term plans to update this failure surface at a later release.Correction of this anomaly will most likely involve upgrading the new generationelements (18X) to accept concrete material behavior. We regret any inconveniencethat this current limitation causes..