Reinforced concrete pole bending failure test
Hello there, I am a civil engineering student working on my thesis project.
I am trying to model reinforced concrete poles for transmission lines. My goal is to obtain results with different rebar configuration and types of concrete. I have been reading a lot, I tried using SOLID65 and LINK180, thanks to the forum I have seen that Ansys 2020 R2 have the possibility of asigning "Reinforcement" to line bodies solving my biggest problem yet.
I will use this propertie, and for the concrete MenetryWilliam material model.
Here is a picture of the type of test I am trying to recreate.
As you can see there is a support in the middle just to avoid the effect of the own weight.
1) Usually this posts embedding lenght is consider around H/10 , H being the total length. Would it be ok to generate the post from this section, giving it the fixed support to the base surface.
2) Also this bending test as you can see allows a big deformation, around 50 cm at the top in a 9 m long pole. Is this big enough to turn "Large Deflection" ON ?
3) Which would be the best way to apply loading, considering the fact that I want to determine the structure resistance? In the test loading is applied slowly at 20 cm from the top. Should a slice in the pole surface be enough? also I would be great to determine permanent deformation after unloading (without reaching failure).
4) How could I determine and plot cracking pattern or the beginning of cracking
*5) Regarding the weight, considering it is a big structure, should I add gravitational force and limit displacement of the body in the vertical axis as a way of recreating test conditions?
Regarding geometry I have been wondering how to cut a line body using a plane, some rebars don't go all the way and I thought modelling full length using point pattern and line from points would be easier to do, the only thing is that I have not been able to cut the line bodies.
If you reached this far thanks for the patience, I have seen many tutorials and surfed along the forum, papers, and google trying to learn this amazing software.
have fun with one of my failed attempts in ANSYS2019 R3
Best Answer

peteroznewman Member
Thank you for a very detailed post. Sorry you had to wait so long for an answer.
1) You can initially have a fixed support at the base. The angle of taper will determine where the failure will first start to occur. With little taper it will fail at the base, with more taper, it will fail near the middle.
2) Yes, you must have Large Deflection On.
3) Yes, slice it off at the load point. Apply a Remote Displacement to pull on that cut face. Type=Rigid.
4) Do some research on MenetreyWillam failure model and Microplane failure model. Cracking is problematic because the results are highly sensitive to the mesh.
5) Don't include gravity
It is good you used Symmetry. Since you should be using an applied displacement, when you request the reaction force, you double that to get the force for the full pole.
Convergence can be very challenging. You must put in many substeps. When you post a reply to ask a convergence question, include the NewtonRaphson Force Convergence Plot. Click on the Solution Information folder and add 3 NR Force Residual Plots. Include one of those plots. These are helpful for the "Failed to Converge" errors.
Good Luck!
Peter
Answers
Update:
Im using a simmetry and using half pole, I applied as shown fixed support at the base, set Displacement in Y axis = 0 and applied the load at the edge, at 20 cm from the top in direction Z = 1000 N. As I use simmetry should I consider half the load?
I still can't manage to arrive at a solution with displacements in the magnitud the should be. Any advice on meshing or on analysis settings? The actual mesh has around 9000 elements which is really less than previous attempts.
Thank you for a very detailed post. Sorry you had to wait so long for an answer.
1) You can initially have a fixed support at the base. The angle of taper will determine where the failure will first start to occur. With little taper it will fail at the base, with more taper, it will fail near the middle.
2) Yes, you must have Large Deflection On.
3) Yes, slice it off at the load point. Apply a Remote Displacement to pull on that cut face. Type=Rigid.
4) Do some research on MenetreyWillam failure model and Microplane failure model. Cracking is problematic because the results are highly sensitive to the mesh.
5) Don't include gravity
It is good you used Symmetry. Since you should be using an applied displacement, when you request the reaction force, you double that to get the force for the full pole.
Convergence can be very challenging. You must put in many substeps. When you post a reply to ask a convergence question, include the NewtonRaphson Force Convergence Plot. Click on the Solution Information folder and add 3 NR Force Residual Plots. Include one of those plots. These are helpful for the "Failed to Converge" errors.
Good Luck!
Peter
Thanks a lot for your answer Peter, I don't mind waiting for someone who takes the time to answer everyone in such a complete way, so nothing you should be sorry about. Also sorry for the repetitive messages.
Great I will try applying remote displacement to the split face, would that generate two solids and a contact region or the slice should be done by imprint faces? I am sorry I'm still learning things related to Geoemetry design.
I was able to apply a 500 mm displacement to the top surface, with around 20 steps,(took around 11 hours) and arrived at a converged result (its the first time I reach so far), I believe the structure could take more displacement. I retrieve the Force reaction and doubled it but it wouldn't match the test load as it should be apply 20 cm lower.
There have been many attempts and modification between the original post and this, I used the NR Residual Plots, and modify some material parameters. I am going to investigate about the Microplane failure model, but I sense that for a structure this big, to obtain detailed crack plots would take a huge amount of elements and computing time. I tried to obtain some crack plots using the PLCRACK command I took from another post, but as I said the structure is to big, and I dont know how to zoom with code to obtain a closer picture.
The force Convergence plot for this case:
How I applied the displacement :
I added this displacement 0 normal to the simmetry plane too
Here is how I applied the displacement by steps, I used Automatic Stepping ON and set Initial to 1500 Minimum to 500 and Maximum 2000
And here is the plot of Force Reaction vs Displacement:
I thought maybe checking plastic strain at certain times would be at least an indicator of the cracking zones, would you agree?
Here is the Equivalent Stress for the concrete at the end of the simulation, probably all the blue side is already cracked and the compressed side haven't reach 30 yet which is the concrete specified resistance
Should I increase steps to reach maybe 700 mm ? Should probaly do it applying first your boundary and loading advice. The axis are rotated because I rotated the simmetry plane in order not to cut any reinforce element.
I attach a pdf file with the model I am using, I model the whole post and apply simmetry and suppress the line bodies below the simmetry plane.
Reinforcement and mesh of the complete model. Would it be more accurate to work with the whole thing?
Thanks a lot for your help, sorry for the long post.
One more doubt, what Output control settings do you recommend for the analysis? In order to have a smaller table data for forces and displacements but still representative. I tried with " Store Results at Equally spaced points" Value 50 and got around 1500 rows of data
Forgot to attach the picture of the "Crack and Crushing"
I manage to apply the load as you said, Remote displacement type rigid, I apply it to the circular edge I got by imprinting faces of a cut at 20 cm from the top and then uniting the bodies.
I am working know with the full post, no simmetry and manage to reduce the ammount of elements to around 6000.
I haven´t reach failure yet, but the model converge with relative big displacements for now, and not so many steps.
Hello @peteroznewman , I have read about MenetreyWillam model, in fact that is what I'm using for the concrete. I still can´t manage to find a way to determine at least the load/displacement that produces the first crack, or anything crack related besides the equivalent plastic strain plot as I show above.
If you happen to know or can point me to the right direction I would be very greatful.
Thanks!