Jump to content
TUFLOW Forum

cfrk

Inactive members
  • Content Count

    64
  • Joined

  • Last visited

Everything posted by cfrk

  1. Hi, My first thought would be to check that no WLL's are snapped to either end of the culvert. If it's still giving you problems you can specify Meshparts == ON in the .tcf file. Then when using Tuflow_to_GIS use the switch -rmp "<meshpart>" ("<meshpart>" being your 1d_nwk layer, e.g. "1d_nwk_culvert001") to remove that culvert from the output. This is explained in the manual on page 11-8. Hope this helps Cheers Richard
  2. Hi Bill, I've just encountered the problem noted in the manual where the folder name is too long to create a copy of my model. So I'm having to first copy my Tuflow folder structure to a point higher up the folder tree and then run the -c command. It would be good to have the option to specify the location the copied files will go to, to avoid this problem. Also, as I have several runs to copy, which use many of the same layers, it would be nice if I could batch these, and have all runs copied to the same folder to avoid duplicating layers, especially on larger models. Cheers Richard
  3. Thanks for your comments - the technical manual will be a great help. A colleague has since found a relationship between K and C, the coefficient of discharge for submerged upstream face of the bridge, where K = 1 / C^2. The standard value of C for most bridges of around 0.8 gives the magic value of K = 1.5625, which indicates that Estry is converting to orifice flow calculations. This value also appears to be adjusted using the relative approach and departure velocities. However, as you point out, the transitional phase is particularly problematic, and I guess even more so when trying to compare between two modelling programs. It also seems that Hec begins the transition to orifice flow when the energy grade hits the soffit, whereas Estry uses the downstream water level based on user-input BG tables.
  4. Hi Paul, If you add a -grid[number] switch to the command line you can specify the size of the grid on which to output vector arrows: C:\Progra~1\Tuflow\utilities\TUFLOW_to_GIS.exe -b -mif -vector -sf1 -t99999 -grid5 rob_021_V.dat Hope that helps Cheers Richard
  5. Hi, I am attemping to calibrate losses through bridge structures with a Hec-Ras model. Originally I set K values based on Hydraulics of Bridge Waterways, but found that backwater was underestimated by Estry. I increased the K values to a point where I couldn't justify any further increase, and backwater was still being underestimated. I changed Structure Losses == Fix which enabled me to line up backwater with Hec-Ras results. I understand that Structure losses == adjust is preferred over fix, for clear-spanning bridges. Unfortunately Tuflow won't allow me to specify BWF or BWA channels, so I have to fix losses at all bridges. I imagine that calibrating afflux by fixing structure losses, and then running using == adjust would be pointless. How exactly does Estry calculate flow and backwater through a structure, i.e. what formulas are used? Does Estry calculate orifice flow once the soffit is reached? I ask because Estry and Hec-Ras predict a different peak flow through a submerged bridge. Do I need to calibrate K values for backwater above the soffit to Hec-Ras results? The equations in Hydraulics of Bridge Waterways for a submerged bridge do not appear to rely on a K value (instead there is a coefficient of discharge), and yet adjusting K above the soffit level in Estry appears to change the backwater calculated. I'm not sure if I should be changing K so that the results match Hec-Ras (sometimes as low as 0.5 and as high as 3) or to just use 1.56. In the absence of other cailbration data, should Hec-Ras be trusted above Estry (i.e. fixing bridge losses for calibration)? Thanks for any advice Richard
  6. Hi Bill, Just a note regarding interpolation of loss coefficients. I am calibrating bridge losses in an older model using the 2007 Tuflow build, and using structure losses == fix. Checking the backwater at the bridge, there definitely doesn't seem to be any interpolation of losses between those values specified. There are obvious jumps in backwater when water level at the downstream node reaches changes in the loss table (when these changes are significant enough), rather than a smooth transition as you would expect from interpolation. It looks like the value is fixed until the next value is reached. The 08 update doesn't mention anything about fixing this so the bug may still be in the latest build too. Cheers Richard
  7. cfrk

    Removing old tools

    Cheers Mick, that's sorted it. Thanks Richard
  8. Similar to the previous post on the forum... I am having the same problem with HQ boundaries. I've checked through Bill's comments, but my elevations range by 2m, and every flow value is calculated as 0 cumecs, not just the first few. Any thoughts why? I've specified slopes of 0.01 at my boundaries. Thanks Richard
  9. cfrk

    Removing old tools

    Minor problem - I've installed the newest version of miTools to c:\BMT_WBM\miTools\ and changed the startup.wor file to reference this instead of the old version which was in .....\professional\Tools\WBM. When I open MapInfo it loads the new version with a few new buttons, but then the old version gets loaded on top removing them again. I've tried moving the old folder, but MapInfo keeps trying to find the old version. Is there any other place that the old files could be getting referenced from?
  10. cfrk

    I culverts...

    On initial set up of an 750m long culvert network, consisting of R and I culverts, flow appears to stop at the first I culvert in the system, causing the model to go unstable. I have searched through all check files and made sure everything is set up properly. I have noticed in the check files that the u/s and d/s inverts of the I culverts are changed by Estry to the average of the inverts that I input, creating a flat channel. I therefore wonder how my HW table will tie in to this, given that it is based on the shape of the culvert at the upstream end, so the bed level is the upstream invert of the channel (which is now higher than the bed level that Estry is working with). Has anyone got a successful model using an I culvert working? Even if I can fix this, I'm not sure if the averaging of inverts is going to work in this instance, given the long sections I am using (up to 65m lengths, about 6m wide and 2m high). Thanks Richard
  11. Hi, you don't need to worry about fixing this, this post explains it: http://www.tuflow.com/forum/index.php?showtopic=269 Cheers Richard
  12. cfrk

    Type I culverts

    Hi, At the Tuflow 08 workshop it was mentioned that there were still problems with I culverts. Is this still the case or are they safe to use? Cheers Richard
  13. Apparently this is what happens when you run out of server disk space... Better delete those old check files
  14. Has anyone else had any random occurences of Tuflow just closing itself? I set a run going over the weekend, only to come back and there was no "finished!" window, and I couldn't load the results. Just now I've seen a simulation window just dissapear with no messages two attempts in a row.... Mysterious Tuflow bug? A bit hard to diagnose, the last line in the .tlf file is "writing output at time...."
  15. To add to the last post, I've got a second culvert with adverse slope, this time only 0.02m over 10m. As with the last problem, the 900mm culvert flows fine at just over 1 cumec, until the upstream head reaches a certain point, then flow drops right down to about 0.1 cumecs, even though there is a head drop of about 1.5m across the culvert. I've noticed that the culvert starts flowing again as the upstream head drops back down on the other side of the hydrograph, and at about the same upstream head that flow stopped, flow starts again...
  16. Hi all, I'm running a model with a culvert with an adverse slope (about 0.3m over 25m). The model runs fine with flow going through culvert, until the upstream head reaches approximately 2m above the invert of the pipe (which is 1.3m diameter). Here flow abruptly stops flowing through the pipe, and water builds upstream until the model goes unstable. If I drop the downstream invert to below the upstream invert, the model runs through. Any suggestions why the pipe with adverse slope might be flowing fine and then suddenly stop, seemingly due to a head level? Cheers Richard
  17. Sorry not true - I was looking at the _MB .csv file which was output to the 2d directory. I am stuck though in that the .mif check files for 1D show mostly 0's. I imagine this is due to my flows being relatively small... Therefore thematically mapping 0's isn't particularly helping in identifying the problem nodes...
  18. Hi Bill, I'm using the new Read MI Z HX Line command, and for previous runs had the RIDGE option specified in my .tgc file. For one section of reach I want to remove the RIDGE option. I've used a separate points layer for the HX line. In the manual it says you can specify an "R" flag for type ZP points (when using points in the same layer as the HX line) to use the ridge option. Because I've used a separate points layer with only an elevation field, I don't believe I can use this option? Would it be possible to be able to specify the "R" flag for the HX line itself rather than at the elevation points? Also it would be good to have a similar flag to say "Don't use the Ridge option" which overrides the .tgc file flag, as it would be quicker for me to specify RIDGE in the .tgc file and flag the one or two HX lines for which I don't want to use the ridge option. Alternatively any useful tips other than having to combine my _pts layer with the HX layer would be appreciated! Cheers Richard
  19. Hi, I'm looking into reducing the mass error of a model and after seeing the second column of % error stack up and the third stay at 0%, I concluded that it was a problem in the 1D domain. I've checked the 1D MB files, which contradict the console screen, indicating a good mass error of close to 0%. Looking at the 2D mass error outputs indicates that this is the source of the error. I am wondering if the console window and .tlf file are actually showing total error, then 2D error and finally 1D error? Cheers Richard
  20. cfrk

    1D network pits

    Thanks David, Yes the u flag I imagine would only let flow in, what I meant to ask was is there a way to have the opposite to a u flag. I.e. only have flow in the negative direction of the channel. Using a u flag on pits I'm guessing I'd have to manually digitise every pit channel which I'd rather avoid... Im at a loss with regards to the losses (sorry), there doesn't seem to be any standard losses for flow surcharging from a manhole, not surprisingly! Cheers Richard
  21. Hi, I'm looking at connecting my 1D pipe system to the 2d domain using pits. I am a little confused about the pit network attributes. I see that an R pit channel is a rectangular (in the vertical) pit channel. The manhole covers are predominantly rectangular, with dimensions of 675 x 675mm. The manhole itself is 2100mm diameter (circular). In the 1D nwk attributes should I be putting the height and width to 675mm as an R pit or as a 2100mm diameter circular pit? What kind of contraction coefficients and entry exit losses can be expected when discharging from a vertical culvert into the 2d domain? I would ideally only allow flow to surcharge from the pipes into 2d (i.e. a manhole cover popping), not the other way around (at least until the manhole has been surcharged hence removing the cover). Can a U flag be applied to a pit? Any suggestions much appreciated. Cheers Richard
  22. Hi Bill, I'm having the same issue with the latest version. With the 2007 version it was working fine specifying the actual depth of the bridge deck (not the deck invert in mAOD). It seems that possibly in the latest version the depth is being read as an invert level, as the initial water level at the node directly upstream of the bridge in question is being set (in my model) at 0.268m (the deck depth is 0.27m, however the channel invert is around 3.0mAOD). Cheers Richard
  23. I'm setting up a model near the coast where there is a significant area of 2d domain (and several 1D pipe outfalls) with negative elevations. As I understand, Tuflow generally sets an IWL for each cell at 0.002m above the ZC level. However for cells with a negative ZC, Tuflow sets IWL to 0. This then causes a wall of water to appear in the model at the start time at those cells. I am having to manually digitise a polygon around cells with negative ZC and specifying a suitably low water level. I can't find anything in the manual about negative elevations, has anyone had similar issues and found a simpler work around? Thanks Richard
  24. In a pipe model I am specifying losses at pits in a pipe network. Do these losses get adjusted by Estry when Structure Losses == Adjust is set (default)?
  25. Hi, I'm trying to load the zsh_zpt_check.mif file, but when I try to import it to mapinfo I get an error: "Found [point] while searching for [)]". I have split my z shape file so that all points are in a _pts layer. Wondering if this is causing problems with the check file. Any ideas how to correct this error? Cheers Richard
×
×
  • Create New...