Hi,
It seems that ccx solver included in PPM doesn’t support ABAQUSMCWD material.
It will be great if it could.
Regards,
Tomek
As a new member i can,t upload .pmx.
hi, you can do replacement of the solver with officially distributed from CalculiX website.
some notes, is a Modified MC with tension cut-off not original ones.
The solver used inside PrePoMax can be set using Settings->Calculix.
1 Like
only want to confirm and update, latest version (v2.21) of CalculiX has officially supported these material type
*MOHR COULOMB
Friction angle (degrees), Dilation angle (degrees), Temperature (optional)
*MOHR COULOMB HARDENING
Cohesion yield stress (c), Equivalent plastic strain, Temperature (optional)
below an example, define new material in menus and add parameter using Keyword editor. Modulus of elasticity of concrete usually taken from compressive strength e.g Ec=4700*Sqrt(f’c) and cohesion is in range of 20% f’c
*MATERIAL,NAME=ConcreteFc25
*ELASTIC
23500.0,0.2
*MOHR COULOMB
37.5,36.0
*MOHR COULOMB HARDENING
5.0,0.0
5.0,1.0
1 Like
Hello everyone,
Unless I’m mistaken, I understood that the MC material model was already implemented in Calculix, introduced through the keyword editor. If I’m correct, this is a way to simulate brittle materials that primarily work under compression, such as concrete or masonry walls.
I’ve conducted tests by introducing the plasticity parameter, and my model converges. However, after reading this post, I’m beginning to doubt whether I might have a conceptual error and my model is incorrect, which is likely due to my limited knowledge in FEM calculations.
The keyword introduction used is similar to the following:
plaintext
*MATERIAL, NAME=MCTerreno
*USER MATERIAL, CONSTANTS=5
18300000000, 0.28, 1000, 35, 5
*ELASTIC
**ELASTIC should appear in a second keyboard. It allows plastic calculation in a third keyboard.
**E: Elastic modulus
** v: Poisson’s ratio
** C: Cohesion
** ph1: Internal friction angle [degrees]; cannot be zero
** ps1: Dilatation angle [degrees]; cannot be zero
It would be greatly appreciated if you could confirm whether my model might be suitable or if, on the contrary, I’m on the wrong path. Attached is a file with the model.
Thank you very much in advance for your help.
Best regards,
Damián
The MC material was added to the official CalculiX version last year (with the release of ccx 2.21). You don’t need the *USER MATERIAL keyword anymore to use it. There are dedicated keywords for this material model now: *MOHR COULOMB and *MOHR COULOMB HARDENING. Check the syntax in the documentation and give it a try.
I just saw it a moment ago; I was unaware of that syntax. Sure, I will try it out. Thank you very much for your help.
Best reguards; Damián
hi,
i opening your file and it seems using earl implementataion of advanced ones, it’s Modified Mohr-Coulomb which i previously mentioned. Also, few of parameter is uncommon for normal concrete: cohesion is 20MPa (too high) and Dilation angle is 2deg only (too low). However, it can be justified by experiment result.
indeed, standard MC material officially supported and available can have dependent on plastic strain for cohesion. It can represent tension softening in concrete, thermal dependent also possible. However, the original author reported a Modified MC is simpler and robust for most case, interesting to do benchmark of both material models. All of them are available in CalculiX distribution, but undocumented ones.
1 Like
Unless I’m mistaken, in this case, I intended for the cohesion to be zero; it concerns a ashlar vault that should not have significant tensile stresses. 0.20 is the Poisson’s ratio.
As for the dilation angle, as a reference, I’ve read that it should be between 25 and 30 degrees less than the friction angle, but I haven’t found more information in that regard. I’ve simply performed the operation. If you have any information on this matter that I could consult, it would be greatly appreciated and tremendously helpful.
Thank you very much for your assistance and expertise.
Best reguards; Damián
sorry i miss, there is no separation. As i understand, cohesion can not be zero for Modified MC models. But it’s good news when solver has no problems.
You don’t have to apologize, you help me a lot and I’m grateful. I set the value to zero out of ignorance, without knowing it could lead to an error. Thank you very much.
thanks for appreciating, help someone can help myself and anyone else. Too low in cohesion value in Modified MC reported in numerical stability problems. The original author benchmark running inside Abaqus is limited to 7.5MPa lowest. However, it seems not applicable in CalculiX, previously i have no problem with 3.0MPa values.
i do simple test of plain concrete, both shown to have the same number of iteration and identical in results. Probably, it needed to investigate further in triaxiality and confinement condition due to steel reinforcement (stirrups and main bars)
Unless I am mistaken, the cohesion value is assimilated to the tensile strength that the material can withstand. The tensile strength value of concrete, in my opinion, seems a bit high. According to the Eurocode, the tensile strength value of concrete can be extracted from the following table; typically, about 10% of the compressive strength is considered as a magnitude order:
In the case of masonry walls, Professor Heyman talks about values around 2.5% of the compressive strength [I am speaking from memory]. A tensile strength value for a traditional stone wall with a compressive strength of 5 MPa would have tensile strengths of 0.125 MPa. 7.5 MPa, in my opinion, implies a high value of tensile strength, both for concrete and masonry elements.
In this regard, I don’t have enough knowledge of FEM calculations to assess the difficulty that may arise from using lower cohesion values [tensile strength], close to zero or even zero. As soon as I have a moment, I will test the model with the correct syntax, and if you agree, I will share the model in case it may be of interest to you.
I greatly appreciate your help and your knowledge. Thank you very much.
indeed, cohesion parameter have relation to concrete tensile strength, but actually it’s a friction resistance on how it began to separate (sliding). Masonry wall can be an exception case since the value is a mixed or homogeneity of stone block and mortar filler, and 2.5% tensile strength seems for later parameter in dominant.
lower cohesion value probably is not a problem in condition of model with enough confinement (stable formed). However, it can lead to numerical stability for real case e.g reinforced beam, column or wall. Previous report on Modified MC also confirmed by Abaqus CDP model, value of 7.5MPa seems too high for normal concrete and given overestimate in results.
below example result for value is half of previously (5.0MPa to 2.5MPa), both material model convergence fast. *Edited (result are identical)
I have tried the model with the syntax for the modified Mohr-Coulomb material; indeed, in this case, it does not allow for zero cohesion. In addition to this limitation, by setting a cohesion value lower than the resulting stresses in the model, no plastic deformations are monitored in the output results.
Probably due to a misunderstanding on my part, I understand that if the stresses exceed the introduced cohesion value, they should be monitored in the output as plastic deformations; am I correct? Could I be doing something wrong? Thank you for your help.
well, it seems something uneven when i tested again externally. Modified MC material in CalculiX v.2.21 is not working normally. The keyword card is readable but treated as elastic, required to investigate further.
please ignore my previous comment, keyword precedence also not working properly. My model returned to the initial standard MC material, an overrides being ignored by the solver. This leads me to the wrong conclusion, both have the same iteration and identical results, which actually can not.
model attached seems too large (148Mb) i’m not download yet, but standard MC material is capable regarding this output request:
*MOHR COULOMB HARDENING
0.10,0.0
0.12,1.0
0.13,2.0
I think your plastic Strain is way to large. If that works as in Plasticity 1 means 100% and 2 =200%
Try
*MOHR COULOMB HARDENING
0.10E6,0.0
0.12E6,0.1
0.13E6,0.2
Could you draw the Mohr coulom surface for any of your elements considering sigma1 sigma 3 friction angle and cohesion to see how close or far are your from the yield surface?
i can confirm, even keyword card is readable and run without any error message, but Modified MC material in latest distribution (v2.21) is not working normally.
below result using older versions (v2.15), number of iteration is a little bit longer, result is similar, equivalent plastic strain output is not available.
probably, it should be no problem since the actual value will be interpolated by the solver. Also, maximum 20% in example given is too high for brittle material like concrete.
What is not working normally?. Is yours an old sintax? There is no plastic strain definition.
I didn’t enter into the hardening details yet for this material model but looks like the Yield surface is pushed keeping Friction angle constant right?. That would mean the intersection of Yield surface with ordinate (cohesion) increases as the Mohr Coulomb Circle grows and Yield surface is pushed.
Manual: " The basic idea is that the shear stress at which plastic flow occurs increases with increasing pressure"
My results without hardening where as expected when testing.