01-04-2017 02:19 PM
Help,
I am running a nonlinear analysis. Sol 106. I define a material curve, with the last value for stress 400 MPa. The problema is that I get stresses of 600 MPa (Von Misses)!!!!
Could you please help me understand?
many thanks
Ionut
01-05-2017 03:37 AM
HI @Lambrescu
Did you choose the type of nonlinearity during the material definition? If not, please try.
Also remember that not all elements support non-linear material.
Give us some screenshot of SOL106 setup and material definition so it will be easier to help.
Have a nice day
01-05-2017 05:30 AM
01-05-2017 09:38 AM
Hello Ionut,
getting stress values beyond stress limits can also depend on extrapolation of internal integration points onto grid points. The effects is larger if your elements are larger. I think the limitation criterion is used at integration points.
The correct limitation of stress values can also depend on yield function criterion. If you select "Mises" sometimes your component stress values can be larger than mises value.
If you uses shell elements mixing top and bottom side can also destroy the relationship between integration points and extrapolated grid point results.
One last point: Defining a material with non linear behavior does not mean that you also use that definition in your solution. Its only a definition of material parameters.
Integation scheme, Integration network can have an influence of results, too.
Best wishes,
Michael
01-05-2017 12:30 PM
Dear Michael,
Thank you for your support. Just a few questions:
1. When you say that "component stress values can be larger than mises value" you mean 10% larger, or even much, much more? I experience such a case1
2. What "Integation scheme, Integration network" settings do you have in mind as influencing the result?
Many thanks
Ionut
01-05-2017 06:00 PM
Lonut,
1. I do not know your model so I can't say how big is the extrapolation effect, but if you model e.g. a beam with only one solid element over thickness and loaded with a bending moment it could produce every extrapolation rate for exeeding stress limit.
2. integration network for solid elements can control i.g. the number of integration gauss points. I believe that it also has an influence of extrapolation of stress values onto grid points.
.../nxnastran/10/help/#uid:id499736
But I'm not sure that its the reason of our problem.
What's your model?
Best regards,
Michael
01-06-2017 11:14 AM
Dear Michael,
Please find attached two images.
- One depicts one model: two solids, one steel, on composite. The two a glued.
- The second depicts the other one: only the steel part.
The element size is 1 mm. The model size is 60 x 40 x 6 (for the steel part, the bigger one, with a hole), and 40 x 38 x 4 mm fof the composite (the rectangular solid).
Many thaks
Ionut
01-09-2017 04:13 AM - edited 01-09-2017 04:15 AM
Hello Iounut,
I think your meshing should not be the problem.
Please check that link or familiar information:
Restrictions and Limitations in SOL 106
Having component stress values beyond 400 as ultimate tensile strength can occur if you define Mises as Yield function criterion, I guess.
See also MATS1
In the first step, if I were you, I would only try to discover the model without laminate, if I .
Best wishes,
Michael
01-09-2017 06:43 AM
Thank you MIchael,
I will make some other tests.
Kind regards
Ionut
01-12-2017 09:08 AM
Did you solve the problem ?
IT's quite interesting how to avoid / control such situations ...