I have modeled a RC beam-column connection in Ansys APDL. Them beam tip displacement in the experimental tests is 75 mm, but in my model it finally reached 6 mm. What are the options that may be effective to happen this? I tested displacement and pressure in loading. results were the same. when I get concrete plot, the third cracks never appear, but the first and second cracks do.
waiting for your suggestions,
regards.
Dear Arash
can you post a figure of comparison. Also descripe more details about material plasticity , boundary conditions and how did apply incremental load.
we may be able to help you
Simulation Governance - you have a mathematical model of the problem. The FE model should match this mathematical model (verification) and the experimental analysis should also match this mathematical model (validation) - see:
http://www.ramsay-maunder.co.uk/simulation-governance/raising-the-game/
Dear Raafat E S Ismail
thank you for your care. I use multilinear isotropic hardening and concrete for solid65 and use bilinear isotropic hardening and Link180 for steel reiforcement. A hing support for buttom of the column and simple support for top pf the column in the direction of x.
Dear Angus Ramsay
thank you. I read it. I want to validate and verify my model and I have some software problem.
I don't have much info re your experimental set up, but I would check the boundary conditions. For example, if it is modeled as clamped at one end but there is minimal rotation in actuality (as perfectly clamped is difficult to effect experimentally), you might see if a rigid body rotation or a rotational stiffness at the fixed end, applied to your numerical model, produces better correlation.
Okay Arash. I would first suggest you consider a very simple problem that has a theoretical solution. You can then check that your FE modelling is bringing you close to the true solution. For example, you could try a simple beam like problem and check that everything is correct. You might find you need to perform mesh refinement to converge on the exact solution. Anyway, this sort of exercise should provide some confidence in the FE approach before you move off to consider more complex materials like RC concrete etc... Hope that you find this suggestion useful.
Dear Arash
sometimes there are local deformations/yielding occur at the supports and applying load, to avoid these we describe small parts of the model at these locations as elastic with a little bit large modulus of elasticity.
then to know the source of your problem, I suggest to try changing the concrete part by steel material and plot load deflection curve if it goes well up to large value of displacement then we know the problem is due to the concrete material plasticity model.
in this case, you try to calibrate concrete material parameter. also size of mesh affect the accuracy (take size 20mm). try also change the load increment
one more trial is to increase the cross section area of reinforcement just to see what they affect the load deflection curve.
sometimes experimental results are not accurate due to the presence of initial cracks caused by moving the specimen prior of experiment. the supporting frame in experiment must be rigid enough.
after you tried and still you couldn't solve it we can discuss more
note that your problem needs patient
Dear Raafat E S Ismail
I substitute solid65 with solid 185, and steel properties for that beam, and you were right. It reached the 75 mm displacement easily. After that I tried to change concrete plasticity , but every time I set open shear coeficient a none-zero number, the ansys closed the program automatically. I don't know why.
sorry I stopped using Ansys long time ago and now I am using ABAQUS.
i am also facing this deflection problem for rc beam analysis in ansys mechanical apdl. How can I overcome it ?
Waiting for your suggestions.
regards.
Not sure if the experimental boundary conditions are as you believe them to be. e.g. it is very hard to model a clamped condition, so some experimenters will use a beam of double the length with symmetric loading about a knife edge to guarantee a zero slope at the knife edge.
Dear Ojalvo
I have just modeled a simply supported rc beam by using ansys apdl command . I am using solid65 for concrete and link180 for steel(rebar ,support plate , load plate). The beam could carry the actual load what i got the experimental result but in Ansys apdl in this case the beam not to be deformed like what deflection value got in the experimental . I have got the deflection value ten times less in Ansys from the experimental deflection value . I check the boundary conditions like concrete nonlinear property , modulus elasticity of concrete as well as check the steel properties modulus of elasticity , poisson's ratio etc. I couldn't get the problem what occurs in deflection value. What can i do to get the accurate deflection value in ansys what got in experiment ?
Waiting for suggestions .
Thank You.
Regards.
- Badges
- Science topic