## Friday, May 31, 2013

### Skew Angle for Bridges

FIGURE 1

If a bridge is not oriented perpendicular to the flow lines going through the bridge, HEC-RAS has an option for skewing the bridge deck and/or piers.  (A user can also skew cross sections in a similar manner)  When skewing a bridge, the user reduces the bridge’s deck/roadway stationing by multiplying the values by cos Î˜ *b.  The new bridge opening width becomes the equivalent length that is perpendicular to the flow lines passing through the bridge.  Figure 1 shows an example of a bridge to be skewed.

The HEC-RAS Reference manual makes a note that the skew angle (Î˜) should be based on the angle of the flow path as it goes through the bridge compared with a line perpendicular to the cross sections bounding the bridge.  It should not be based on angle upstream of the bridge as shown in Figure 1 with the gray lines, which would overestimate the skew angle.  As the water approaches a bridge that is highly skewed, it’s common that the flow lines will turn before going through the bridge.   A field visit is very important to visualize the flow pattern at the bridge and to help estimate the skew angle.

In determining a bridge skew angle, a modeler should also consider the question, “Will I be modeling large flows or small flows and will the flow paths through the bridge vary between the two?”  If you are modeling a dam breach, it’s more likely that the flow path during the peak of the flood won’t turn as much through the bridge, and therefore the hydraulic skew should match the geometric skew of the bridge.   If you are only modeling lower flows, the HEC-RAS Reference Manual suggests that skew angles below 20Âº to 30Âº do not appreciably affect the flow patterns through a bridge – the reason being that during lower flows the water/flow lines will be able to turn or meander more easily through the bridge opening than during large flows.

Users should also remember to apply the same skew angle to the bounding bridge cross sections.  If for some reason you don’t want to skew the bounding cross sections, you may have to manually alter either the cross section or bridge deck stationing to ensure that the bridge opening correctly aligns with the cross section.

1. Do the cross-sections have to be perpendicular to the flow for Computation. My Surveyor got the XS data skewed at an angle ranging from 0-30. Entering the data does not represent a true picture of the river. What should be the corrective action?

1. If the skew is significant, you can provide a skew angle to any cross section. I would certainly do that for skew angles greater than 10 degrees, plus or minus.

2. Will my geometry file show the angle change? I entered the skew for the bridge and the geo file did not change, despite recognizing that I had entered a skew.

1. It will not look any different in the geometry schematic. You'll notice that x scaling change in the cross section plot though.

Chris

3. HEC-RAS allows for skew angles between 0 and 45 degrees. What if the skew angle is 60 degrees? I have a case where flow is parallel to the roadway upstream and downstream of the bridge and passes beneath the bridge at a severe 60 degree skew. The length of the bridge along the centerline is 86 ft and the normal bridge opening (perpendicular to flow) is 43 ft.

1. Once you get skews that high, it is common that the flow lines will bend and the actual skew is still less than 45 degree. IN other words, from an aerial photo the skew may look like 60 degrees, but if you measure it off of the flow lines it is much less.

4. Have you ever come across needing to set a skew angle for just an entrance or exit? See W Calaveras Blvd Bridge, Milpitas, CA. Lat: 37°25'48.18"N, Long: 121°54'33.27"W.

5. With respect to the X axis scaling change after applying a skew, would cuts and fills to the skewed XS be representative of actual real world cuts and fills given that the X stations are now shorter in the skewed XS

6. They should be. Yes. But double check it.

7. Chris, should circular piers be skewed? I guess if you look at a circular pier from all angles, the width (Diameter) doesnt change and thus should not require a skew in HEC-RAS. What do you think.

1. Good question. I suppose if it is just a single circular pier (not a column of piers) then I'd say you are correct.

8. If a small stream bends differently u/s from d/s of the bridge, would one skew the u/s and d/s internal sections differently? Should it use the angle of the flow d/s determined as described for u/s or just use the u/s skew for both as would normally be done?

1. I usually pick an average skew angle and use it for both. However, it might not be a bad idea to try a few different ways and use the more conservative result.

9. Chris, what about the case where the flow through the bridge is not skewed, but the bridge structure itself is skewed above the channel? See the image hosted at https://drive.google.com/open?id=0BzWjEONrpQsEWEVSa0ItX1ZvSms. Thick red lines show the bridge, flow is from right to left. It is also worth noting that the bridge needs to be modeled under pressure flow. How would you suggested orienting cross sections here? Should the skew option you discuss here be used at all? It seems like this option is for when flow lines skew through the structure, which isn't the case with the project from my image. Any help would be appreciated! Thank you.

1. In this case, put the bounding cross sections (2 and 3) parallel to the bridge. Then skew the bridge deck and the bounding cross sections, but not the pier. You should skew any time flow lines are NOT perpendicular to the bridge deck or any cross section.

2. Thank you, Chris. From the Hydraulic Reference manual and the dialogue box for bridge skew it looks like it's not generally recommended to go above 30 degrees for the skew. Measuring mine how you've explained in this post shows a skew of 44 degrees. Do you think I would be better off changing cross sections 2 and 3 to be at 30 degrees, or leaving them parallel and using 44 degrees in the skew dialogue window?

3. I'd still leave them parallel. The 30 degree limit is just a guideline and is based on the idea that much more than 30 degrees skew, and the flowlines will typically bend to get through the opening or over the deck. It's very subjective. I recommend trying a range of skews and build that into your sensitivity analysis.

10. Chris, We are performing a study in which we have a number of bridges that are just like the September 29, 2016 post. The abutments and the piers are aligned with the stream flow, but the bridge is skewed relative to the stream.

We have been having a discussion in the office as to how to best lay out the top of road and bounding cross-sections to make our work proceed as quickly and accurately as possible. We know that the top of road and bounding cross-section should be the same length, but the question is where to start our “0” station on the left bank.

Should we have our “0” stations parallel to the stream centerline so the center line of the bounding cross-sections are in the correct locations, OR, should we have our “0” station for those three cross-sections be located perpendicular to the road and then adjust by hand the station location of the stream center line at both the upstream and downstream face bridge face? This may be more of a HECGEORAS question…

1. Not quite sure I understand what you mean by "0" stations being parallel to the stream centerline or perpendicular to the road. In general, your bounding cross sections (sections 2 and 3) should be parallel to the bridge and roadway approaches, even if they are skewed. Then you skew those bounding cross sections as well as the bridge deck. All other cross sections should be perpendicular to flowlines. Not sure if this answered your question...

11. For a pre and post bridge comparison, as I understand it, the cross sections would need to be the same in both models. I have a skewed bridge and have the modified locations of the cross sections upstream and downstream of the bridge to be as close to parallel as possible given my the bridge has a slight s curve.
Please tell me if my thought process is correct.
I would need to apply a skew to these cross sections in both the pre and the post bridge models and the bridge in the post bridge model. Also, to verify I am understanding, after the skew, the terrain I originally cut my cross sections from will no longer match.
Lastly, I know my piers are going to be placed parallel to the flow and are cylindrical, so I am wondering if I would need to skew them as well.
I am also realigning the channel.

12. Hi Chris!
I want to create a new cross section in existing HEC-RAS model. The situation is like there is existing road crossing the river and now new road joins the existing road just before the bridge. The new road after some distance runs parallel to the river. So I want to take new cross section perpendicular to the river till the road and along the road then after. If I just take the stations and elevations of my proposed cross section, will HEC-RAS take it straight line or it will be along the road? If it takes straight, then is there any way to bend the the line?

1. You can bend the cross section cut line by cutting it the way you want in ArcGIS and then importing using HEC-GeoRAS.

2. Unfortunately, I have only base HEC-RAS model and contour map (without surface). I added some sections manually in base HEC-RAS model using contour map and spreadsheet to get revised model. In order to include proposed roadway, some station and elevation are added in revised model. In addition, levees option is used at top of pavement to make the roadway act as barrier. So, I want to know the functionality of culvert if it is provided across the barrier(roadway). Will barrier allow water to flow at defined culvert location through it? If yes, can we see that in "View Cross sections" tab?

3. You can only remove water from a river at a cross section by using a lateral inflow (outflow with negative values) boundary condition. If you want to use a culvert, then you'd have to simulate the barrier with a lateral structure.

13. Is the skew angle that is entered for the bridge cross sections and the US and DS sections the same skew angle entered for the angle of attack of the flow on the abutments on the abutment scour tab under the Hydraulic Design Function? The manual states that 90 degrees should be entered under angle of attack for abutment that are perpendicular to the flow (normal condition), but I don't see how that would agree with the figure presented here. Can you please explain? Thanks! Great Blog!

14. Hi Chris,
When adding skew angle to bridge/culverts, looks like HECRAS adjusts the stationing of bank locations but does nothing to levee and culvert (I have multiple culverts and using multiple opening option) locations. Do we have to adjust them manually or it does not matter? Thanks.

15. Let's say the bridge in Figure 1 above runs east to west. If you were to orientate the bridge to run north to south, would your skew angle then exceed 45 degrees?

1. The skew angle is always the angle between the bridge alignment and a line perpendicular to the flow lines at the bridge. HEC recommends not exceeding 45 degrees because typically the flow patterns will adjust to flow more directly through the bridge opening for overly-skewed bridges.

16. Chris, do you think that a small residential bridge that is 15-ft wide, 80-ft long with 4 sets of 4 piers that are 1-ft in diameter and not continuous need to have a skew angle? Is there a way to model piers that are not continuous?

17. What are the situations that make us use the skew angle ??? Is there skew in the bridge or in the pier ???

18. I am creating a Corrected Effective Model for a bridge on skew of 17 degrees. The bridge is supported by 4 sets of beams supported by 4 separated piles at a skew to the flow. It doesn’t seem correct to model the 4 piles under each beam as continuous on a skew as the impediment to flow is larger than reality. The bridge is to be replaced by a bridge supported by two sets of beams supported by two piers. I am wondering how to deal with non-continuous piers on a skew

1. unfortunately, that's a limitation in RAS. Piers are continuous. You might experiment with less skew on the piers or a smaller drag coefficient to see if that can help correct the error.

19. Hello...
Does anyone guide me how to set the skew angle left or right...please?
Thanks...