Hi All,
I have some upcoming projects that might require some fairly complex plate models - in anticipation of that I'm running through practice models to get a feel for some of the issues I might face. Specifically, I'm looking at modeling closed rib orthotropic bridge deck. The model will consist of a steel road plate, longitudinal ribs, transverse floor beams, and some form of supporting structure that will likely consist of standard line members.
I have identified some issues I might have and I'm looking for some general tips on how to deal with them:
[ol 1]
[li]I'm using an excel spreadsheet to generate my node coordinates and define the plates for my deck plate, ribs, and floor beams. Excel compiles all of the nodes and plates into a couple clean tables and then I can copy and paste the entire structure into RISA. This works quite well, but with a complex structure this can turn into a lot of copy and pasting - is there anyway to directly link an excel sheet to RISA so RISA automatically updates when I adjust the spreadsheet? I've done some reading on this and it doesn't look like this is an option, but I'd love to hear otherwise.[/li]
I've played around with some options for this on a test model - I can model say a 4x4 plate (units don't matter) and a smaller 2x2 plate centered on the larger one, load the 2x2 plate, submesh both and the 2x2 plate will load the 4x4 plate. The issue with this is that my 2x2 "load plate" is reinforcing my 4x4 plate locally. To overcome this, I can create my load patches with a new material that has no stiffness or density. The main issue I found with this in my test model is that if I want to add in a new load after I submesh my deck plate I then have to submesh again locally to integrate my load plate...this can get messy fairly quickly and lead to some ugly looking mesh. I may have to just be very careful in making sure I properly define all of my wheel loads before submeshing - even on my small test model, adding in partial area loads after submeshing is a real pain.[/li]
[/ol]
I have some upcoming projects that might require some fairly complex plate models - in anticipation of that I'm running through practice models to get a feel for some of the issues I might face. Specifically, I'm looking at modeling closed rib orthotropic bridge deck. The model will consist of a steel road plate, longitudinal ribs, transverse floor beams, and some form of supporting structure that will likely consist of standard line members.
I have identified some issues I might have and I'm looking for some general tips on how to deal with them:
[ol 1]
[li]I'm using an excel spreadsheet to generate my node coordinates and define the plates for my deck plate, ribs, and floor beams. Excel compiles all of the nodes and plates into a couple clean tables and then I can copy and paste the entire structure into RISA. This works quite well, but with a complex structure this can turn into a lot of copy and pasting - is there anyway to directly link an excel sheet to RISA so RISA automatically updates when I adjust the spreadsheet? I've done some reading on this and it doesn't look like this is an option, but I'd love to hear otherwise.[/li]
[li]Any tips on managing the selection of plates? I'd like to be able to isolate certain parts of the model to review results. I know you can do this with the Criteria Selection tool, but even that can get a bit sloppy if I have lots of plates (that I don't want selected together) that are both parallel with a plane, or same thickness, or too close together to easily select separately. I have saved selections of plates before submeshing in hopes that RISA would maintain that group of plates as a saved selection after submeshing, but that does not work. I'm considering maybe creating additional materials like steel_deck and steel_floorbeam - then I could easily filter different plate elements on the basis of material. Any suggestions on how to handle a large amount of plates in a model would be great.[/li]
[li]Applying loads seems like it will be the biggest challenge. I need to apply uniform loads over the entire deck area as well as point loads from the wheels. For the wheel loads I would like to apply a proper load over the contact area of the wheel rather than a discrete point load - considering the size of the wheel contact area and the typical spacing between ribs I don't think discrete point loads would be appropriate. In order to define the wheel loads in a patch, I need to first create a plate to cover the contact area. I can create a specific deck plate under the wheel, but then I'd have to manually model in the rest of the deck plate around my "wheel plates". Considering the number of wheels on a single design truck and the fact I want to check multiple truck positions, I could end up with hundreds of little patch loads and modeling the main deck plate around these smaller plates will be too time consuming.I've played around with some options for this on a test model - I can model say a 4x4 plate (units don't matter) and a smaller 2x2 plate centered on the larger one, load the 2x2 plate, submesh both and the 2x2 plate will load the 4x4 plate. The issue with this is that my 2x2 "load plate" is reinforcing my 4x4 plate locally. To overcome this, I can create my load patches with a new material that has no stiffness or density. The main issue I found with this in my test model is that if I want to add in a new load after I submesh my deck plate I then have to submesh again locally to integrate my load plate...this can get messy fairly quickly and lead to some ugly looking mesh. I may have to just be very careful in making sure I properly define all of my wheel loads before submeshing - even on my small test model, adding in partial area loads after submeshing is a real pain.[/li]
[li]Moving loads - There isn't a way to run a moving load across a plate, but are there any clean work arounds to this? I've considered putting in a dummy member along the wheel path and using a standard moving load along the dummy member - but as I mentioned above, I'd like to capture the effect of the contact area of the wheel so I can get an accurate look at the local stresses around the ribs. The best I came up with is to use excel to define a series of wheel positions and surface loads and integrate these patch loads as described above. This leads to potential issues noted above about integrating these plates, and it also means I need a separate load combination for each truck position - I'm looking at fairly short span structures so this might not be too big of an issue. I may have to revisit the dummy member concept and put a few parallel members in close together, each carrying a fraction of the wheel load, to get closer to the results from using area loads. Again, any tips or tricks for this would be greatly appreciated.[/li][/ol]
