Jump to content
TUFLOW Forum

Sam Marginson

Members
  • Content Count

    50
  • Joined

  • Last visited

Community Reputation

0 Neutral

About Sam Marginson

  • Rank
    Advanced Member

Recent Profile Visitors

1419 profile views
  1. Hi TUFLOW people 🙂 suddenly I am unsure of something as basic as what grid size to select! Looking at the first chart in section 3.2.1 of the release notes, I'm left wondering whether the seeming lack of impact of grid size on results is limited to Quadtree users or whether the results would be similar for standard HPC users. Normally I'd try to select a grid size so that channels & bridges are represented by at least 5 grid cells, but if I'm dealing with a relatively wide floodplain that's a bit demanding. No doubt the usual rule of "you should satisfy yourself that the results are reasonable" applies, but do you have any experiences you can share in that regard?
  2. note that this is the total form loss applied using Method A (no longer the default) as the bridge loss approach in 1D. Section 5.7.2.3 of the 2018-03 version of the manual states, "once the downstream water level is within 10% of the flow depth under the bridge, a bridge deck submergence factor is phased in by applying a correction for submerged decking using a minimum value of 1.5625 (if the specified loss coefficient is greater than 1.5625, this value is applied)". A derivation of that value follows in the manual in the past I have subtracted the layer 1 FLC from 1.5625 to get a value for layer 2 when using the "cumulate" method. However, the value applied will not reach 1.5625 until the top of layer 2 due to the weighting of the FLCs shown in the equation in section 6.12.2.2 of the manual. It is similarly impossible to select a value for layer 2 that produces a FLC of 1.5625 from the bottom of layer 2 as the value you need to apply to achieve that FLC becomes infinitely large as the depth above the underside of the bridge approaches zero. I think that if you wanted to reproduce the 1D Method A approach, you'd need layer 2 to represent that part of the flow area "within 10% of the flow depth under the bridge", so that the loss is equal to 1.5625 at the point where the water hits the underside of the bridge deck
  3. Hi Bill & Nick I'm trying to do this with a pipe that's connected to the next pipe downstream by an "X" channel, but I'm getting the following error "ERROR 1243 - More than one channel snapped to start of connector - check direction of connector." the model runs when I don't read in the layer that contains the single channel I'm overwriting, which makes me think it's not recognising that the channel is being overwritten when it's doing the check for connectors that have more than one channel upstream. Have you had this before? thanks Sam
  4. Hi! I'm having some trouble getting it to rain in a model I've converted to HPC. It worked fine in classic, but when I add the lines below to the TCF, no water enters the model... Solution Scheme == HPC Hardware == GPU in the TLF it shows that it's still reading the hyetographs correctly, but for some reason the Vi column is full of zeroes. Any ideas? I thought maybe it didn't like the Global Rainfall BC command & tried it with a 2D_RF polygon but no luck. Similarly, I tried changing the PO layer to one without QS lines but still no luck not sure if it's related but the DOS window also contains a bunch of errors along the lines of "CUDA driver API error 0400". They occur all the way through but the run doesn't stop please help! Sam
  5. Hi! I'm having some trouble getting it to rain in a model I've converted to HPC. It worked fine in classic, but when I add the lines below to the TCF, no water enters the model... Solution Scheme == HPC Hardware == GPU in the TLF it shows that it's still reading the hyetographs correctly, but for some reason the Vi column is full of zeroes. Any ideas? I thought maybe it didn't like the Global Rainfall BC command & tried it with a 2D_RF polygon but no luck. Similarly, I tried changing the PO layer to one without QS lines but still no luck not sure if it's related but the DOS window also contains a bunch of errors along the lines of "CUDA driver API error 0400". They occur all the way through but the run doesn't stop please help! Sam Gin_09_05PC_2160m_post_HPC.tlf Gin_09_05PC_2160m_post_HPC.hpc.tlf
  6. is there somewhere that the maximum flows across each QS line are written? The maximum value in the POMM.csv appears to be the 2D flow or do I have to take the maximum of the time series written in the SQ.csv file?
  7. am I correct in saying that, when you only write maximum results as map output (Map Output Interval == 0), Write PO Online == ON essentially does nothing? if so, any chance we could get it changed so that it writes the PO results at the Time Series Output Interval please? Sorry if this has been changed in the 2017 version - I'm still out of date
  8. Hi all what about infiltration? Does it still work if you use the Global Rainfall BC command? It certainly seems to but just thought I should check cheers Sam
  9. Hi Bill any more details about this? I flicked through the manual & release notes but the only things I thought this could be a reference to was the Set Variable command or Set IWL == Auto cheers Sam
  10. actually I take that back. I'm getting similar results at the D/S end of my results in the XMDF as well
  11. actually I'm wondering if this is some sort of TUFLOW_to_GIS artifact. When I look at the XMDF in QGIS it looks OK...
  12. x2 on this question, except I have a slightly different problem - I don't want to apply an HQ boundary, I want to apply a constant water level all around the model perimeter However, as I'm running climate change simulations & don't want a different active area for each climate change scenario, in some locations in some runs the level's below ground will this turn the whole line into an HQ boundary, or only the cells that are below the level? It looks like the whole thing's turning into an HQ boundary
  13. anybody else have this problem? I work with mixed models a lot & I find it a pain to convert everything to the one format to work in
  14. I've got multiple runs going on my computer, one of which is a GPU run. It seems to have a drastic effect on the run times of my CPU models. Has anyone else experienced this?
  15. can I have values >1? I can't seem to using the GPU version I tried to check what was going into the model when I have a value <1 using the RFR output type & I get 0, again using GPU does this output work with GPU?
×
×
  • Create New...