Jump to content
GMS, SMS, and WMS User Forum
NickV

SRH-2D HY-8 Culvert Modeling Troubleshooting

Recommended Posts

I am attempting to model a new driveway across a creek to determine the structure required to have a "no-rise" impact on the base flood elevation of the creek. In doing so, I am trying multiple options. I have successfully completed an existing "no-bridge" model and multiple box culvert and bridge options, I am having trouble with the models with HY-8 multi-barrel RCP culvert options. I copied and modified the existing mesh to create a berm to the dimensions of the driveway across the creek, added hy-8 culvert BC elements and ran the model. The model will say 100% complete, but will only complete about 1 hour of a 2 hour simulation. I have increased and decreased the time step to attempt to resolve this issue. In some cases it helps, others it doesn't. In either case, the results are very sporadic, changing with every time step. Is it normal to see such varying depths and velocities throughout the entire simulation when modeling culverts? Is there anything I can check that I might have over looked that might cause these results? I am using SMS 13.0.3. I have patch elements for both my channel and proposed driveway surface. Equilateral elements between the BC lines. I do have a material assigned to the roadway surface between the BC lines, I read in the newer versions it was not necessary to leave this area unassigned anymore for culverts. I do have near vertical elements on each side of the berm, would this area need to be deleted to represent a vertical wall? I have the inverts of my culverts 0.01 feet above the stream bottom as they will be placed in the creek at existing elevation and grade. I do not know what else to try as I am not getting any errors and the model will run but the results are not reliable. I would appreciate any help or advice. Thanks!

Share this post


Link to post
Share on other sites

First, I would recommend updating to the latest version of SMS, 13.0.5. Second, yes you should use voids to represent your culvert walls. Vertical, or near vertical walls have always given me trouble. The voids only present a problem if the structure is overtopped.

The model might say 100% complete, but you need to look at the tab of the run to verify the model ran to completion and didn't crash, which is what it sounds like it is doing. If these suggestions do not help, let us know. Also, at a couple of screen shots and that might get you a few more ideas to try.

Share this post


Link to post
Share on other sites

The culvert option will over top at the 100-yr flow. The idea would be for it not to over top until the 25-year storm event but let it act as a weir during a higher event. I believe ultimately the culverts will not work but I need to show the client why it will not work. Does the culvert BC need to snap (Shift-Q) to the bottom of the berm or the top? Currently it is to the bottom. I will try deleting the elements along the face of the berm and see if that works for the 10-year flow.

Any suggestions of screen shots that would help with this? I can provide screen shots but I am unsure of what might actually be of benefit to show.

Thanks for the response.

Share this post


Link to post
Share on other sites

The culvert should snap to the mesh where the invert of the culvert is. As for pics, just helps to see if there are any other problems that might stand out that you are not aware of. I have had better luck with quads  for culverts than triangles.

Share this post


Link to post
Share on other sites

I've attached a plan view and a view of the downstream side of the berm. This is before deleting elements to help the vertical wall issue. Hopefully this helps.

PlanView.JPG

DownstreamFaceOfBerm.JPG

PlanView2.JPG

Share this post


Link to post
Share on other sites

First of all, I misspoke. You use the vertical walls when modeling a culvert as a pressure flow bridge. I see why you are getting a crash here. Your elements are nearly on top of each other. Seeing the pics helps a lot. You need refine you mesh more near the culvert itself. Those near vertical walls at the face of your culvert are unnecessary. I would try 2 things:

1. Refine your mesh at the culvert with more slope going up to the road.

2. Do a 2D culvert with voids on the side and make your culvert BC into a pressure flow BC.

Either way, I would at least double the nodes in the vicinity of the culvert. It shouldn't add much to your run time.

Hope this helps!

Share this post


Link to post
Share on other sites

Yes, the vertical walls are seeming to make the model crash quicker. So even though the proposed crossing will have vertical walls in the form of retaining wall blocks, the model does not need to represent that? Is this taken care of in the HY-8 settings when assigning culvert attributes? Using the 2D culvert with pressure flow BC would more accurately represent a Box Culvert or a span bridge, correct? I am using this mesh to model my pressure structures for a 3 barrel box culvert and a span bridge (the holes in the mesh are only for the multi-barrel box culvert).

For doubling the nodes you mean instead of having 6 elements across the creek at this point, I should use 12 like in the pressure images below? I only added more elements to this mesh so I could delete the elements for the 8-inch piers of the box culvert.

PressurePlan.JPG

PressureDownStream.JPG

Share this post


Link to post
Share on other sites

The model does not need to see the vertical walls using HY8 because once the water goes into the culvert, it will come out the other side of the culvert. Therefore, no flow will be between your two BC arcs and the vertical face is irrelevant. 

Yes, the pressure flow option is for box culverts and box bridges. When using the pressure flow option, your outside vertical walls can be represented as voids as well.

The number of elements in this channel looks good, but I would add one or two to the driveway. Also, use the mesh quality feature to check out the elements. That will help the fidelity of your mesh.

Share this post


Link to post
Share on other sites

Thanks for all the advice and help. Always learning with the software. I modified the slopes and the model is running now. I thought I had a good grip on Pressure BC models, but I got a curve ball with these RCP culverts. Thanks again for all the help. I will apply all the techniques and I will get back to you if there is still issues. It appears to be running the full simulation now.

Share this post


Link to post
Share on other sites

I have successfully ran the model several times with various crossing configurations. I have a few reservations about the accuracy of the results.

First the RCP culvert options. I made the changes we discussed and the model successfully runs. I created two simulations using the same mesh, material, and monitor coverages. I duplicated the BC I used for the four 48" RCP culverts and modified the BC to use five 48" RCP culverts. The results show an increase in water elevation upstream, even with the extra culvert. The only thing I changed was the number of culverts in the HY-8 window. Is there more to this than what I have done? These results do not match what I believe would make sense. More pipes = more flow which should lower the upstream water elevation or at the very least not cause an increase. Also, I know this option will overtop at the 100 year flow. Does it matter that the crossing is over topped, but due to the culvert assignment, the elements over the culvert display as dry? Will this impact my results in anyway? (see the screenshot for clarity)

For the pressure culvert (box culverts) options. I believe I know this answer, based on my results, but will I need to create a new mesh and modify it to fit the dimensions of each culvert? I.E. to model a 3 barrel at 10'Wx9'H and at 10'Wx10'H, I created one mesh and two BC coverages and modified the ceiling elevations accordingly (I set the crest for both at an elevation that would not overtop). These two models created the same water elevations upstream. Which isn't a stretch with just one foot higher ceiling. To test my theory, I used the same mesh and duplicated the BC of one and changed the length across the weir from 30' (3x10'W barrels) to 48' (3x16'W barrels) to model a 3 barrel 16'Wx9'H culvert. The results yielded no change in the upstream water surface elevation. So before I made a new mesh for every option of width, I wanted to confirm that I will have to do this and adjust the location of my vertical walls and locations of the piers to accurately represent the various widths of box culverts to model.

CulvertPlan.JPG

BoxPlan.JPG

3x16x9Settings.JPG

3x10x9Settings.JPG

Share this post


Link to post
Share on other sites

For the culvert option, check the culvert.dat file and compare them. The overtopping should be in there too if I remember right. You are correct about the pressure flow I think. You will need to adjust the walls to make it the right size. Also, remember that the velocity goes to 0 by the voids. That will affect your flow. Try to make a small element next to it to mitigate this effect.

 

Share this post


Link to post
Share on other sites

I have made new meshes for each option and it appears the results are more consistent with what would be expected with the various culvert box sizes. For the culverts, would it be better to represent my vertical piers with obstructions in the BC or to add the small element along the sides? Thanks again for all the tips and advice in this thread.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...