I am a intermediate user of FEM software in the area of Geotechnics and relatively new to Prepomax (PPM).
I am starting a new learning journey on the use of PPM and working on a project to simulate concrete pavement with dowel. In fact, there is a small gap or clearance between the dowel and concrete holes. I have no idea on setting up configuration on the contact issue with/without gap.
I have initiated this tutorial file by learning on examples and tutorials on Youtube.
For experienced users, I hope you can help on the configuration for this simulation?
I attached necessary files for your perusal.
Thank you.
You may also need contact between the layers because they may intersect during the deformation. And consider if you want tie constraint (permanent bonding) or contact (possible sliding and separation) between the dowel and the holes in concrete.
I would check some research papers on this topic too. Also for BCs because yours seem to be based on my tutorial about pure bending of RC beam. For example:
Thank you for swift response to the issue raised herewith.
I made simple modeling on the issue, and I am cautiously aware of the issues you mentioned and suggested.
The set-up configuration is fundamental and if you could help improve or rectify the setting/configuration (especially on the contact boundary (permanent bonding/sliding-separation), it would be beneficial for educational/practical purposes if you could demonstrate using PPM.
I am still gaining information and knowledge to improve the simulation.
Yeah, now there’s no interference between the slabs anymore. So let’s move on to boundary conditions. You should consider whether simply-supported BCs match your desired setup. The Surface spring constraint already supports the model in the Y (vertical) direction. You could also cut the model in half in the YZ plane to utilize symmetry - just fix UX for the faces of the cut. Then only the Z direction will be unconstrained, and you can pick some nodes where no displacement in that direction will occur to eliminate all rigid body motions.
Of course, the Surface spring stiffness should correspond to the stiffness of the foundation underneath the slabs. You can find some values in literature. You could also model a larger segment along with the foundation, like in the first article I shared. Then BCs are easier to define because you can just fix the bottom of the foundation far from the slabs - thanks to Saint-Venant’s principle, the results should be good for them.
Dear @FEAnalyst
It was a fruitful input from you to improve this configuration setting.
Valuable lessons learned from executing this file are as follows:
Two material properties
Add-in material properties using Calculix keyword editor
COMPRESSION_ONLY
Boundary conditions
Spring stiffness
Contact surface/pairs
Symmetry analysis
I will apply the lessons to more challenging real world applications.
Thank you for the direction.
P.S. If anyone can find locations to improve for this concrete pavement issue, let’s explore and discuss together as I personally think that this is a practical issue and useful for geotechnical/pavement engineers.
For such honeycomb structures, you can analyze periodic segments / RVEs or unit cells. You will find many examples for various hexagonal honeycombs (also reinforced in different ways) in literature.
Thank you for the information.
In this post, I attached a small scale panel pavement with dowel.
Kindly have a look at the attached file, if feasible within your knowledge, please help suggest or give opinion/comment on the mesh setup (e.g., Extrude mesh, Sweep Mesh, etc).
Boundary conditions are unrestrained all sides.
The center panel will act as a loading panel from wheel load (simple loading for now, later on I will try to have a dynamic or moving wheel load on the pavement).
While the bottom panels will be spring boundary condition as you mentioned earlier.
Small clearance (0.2mm) between dowel surface and concrete hole surface, and between each panel side.
Don’t you want to try with a unit cell model first ? Either way, you could get rid of the chamfer to simplify the geometry using the built-in Geometry → CAD Part → Defeature tool. However, due to the holes, you can’t use extruded/sweep meshing for the hexagonal cells. Some holes even have conical drill points - can you make them flat ?
To obtain a hex mesh, you would likely have to split the cell into volumes with 5/6 faces having 3/4 edges each and use a transfinite meshing algorithm. However, this would be rather problematic and maybe you could just use a tetrahedral mesh for now.
Dear FEAanalyst,
Very much appreciated for pinpointing the issue regarding the panel setup.
Regarding the hex panel issue, I could not find relevant settings from your tutorial videos on Youtube or in this forum, especially the types of mesh and geometry relationship.
I am glad that you confirmed about extrude mesh and few suggestions to advance to next setup.
I am attempting to set up this analysis to confirm the capability of Prepomax and I was lost and wandering what to set up next (e.g. Boundary conditions).
If your specialty can help, I will be happy to learn from your examples.
Besides FEAnalyst, if anyone interests, please help give comments/suggestions.
I attached herewith two ppm files: Unit Panel and Group Panel for your perusal.
Actually, it doesn’t have to be either full pavement structure or a single hexagonal cell. You could model a Representative Volume Element (RVE) - a segment of the structure that represents its periodicity.
As an exercise, you can cut and paste this region around the pavement to see if it fits the whole pattern:
hi, a dowel in concrete slab on ground have one side embedded and the other side is sliding. Common problem is in modeling anchorage due to simplification by plain bars, an approximation can be represented by headed bolt or plate end cap. Also, models need large enough e.g quarter of slab dimension with actual axle load in areas. Stiffness of soil modeling as spring from subgrade modulus, this value corelates with CBR value tested. Regarding to sizing of dowel requirement and minimum distance to surface by hand calculation, some code guidance’s available. Nonlinear material for concrete (MC) and steel (VM) also required for such as simulation.
